University of Southern California Dissertations and Theses

Transplantation and genetics of human pancreatic islets in diabetes: Approaches in translational medicine and statistics

(USC Thesis Other)

Transplantation and genetics of human pancreatic islets in diabetes: Approaches in translational medicine and statistics

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content TRANSPLANTATION AND GENETICS OF HUMAN PANCREATIC ISLETS IN DIABETES: APPROACHES IN TRANSLATIONAL MEDICINE AND STATISTICS by John S. Kaddis A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (SYSTEMS BIOLOGY AND DISEASE) May 2011 Copyright 2011 John S. Kaddis ii ACKNOWLEDGEMENTS I am indebted to my advisors, colleagues, collaborators, and funding sources for their contributions to the research described in this dissertation. Much of their work and advice is noted in the acknowledgement section below and/or in each of the chapters therein. I am grateful to my lovely wife, who provided me with a great deal of support and guidance, both scientific and personal. If our children ever read this, they should know that their unpredictability was a joy, especially during early morning or late night writing sessions. Lastly, I am thankful to my parents and siblings, on both sides of the aisle, for their unwavering love and encouragement. Chapter 2 includes a weblink to the journal article, Kaddis JS, Olack BJ, Sowinski J, Cravens J, Contreras JL, Niland JC. Human pancreatic islets and diabetes research. Journal of the American Medical Association. 2009; 301(15):1580- 1587. Authorization was granted for inclusion of this document in electronic format as a weblink only. This manuscript, http://jama.ama- assn.org/content/301/15/1580.full.pdf+html, is hosted on the journal web site and appears with permission of the American Medical Association. Copyright © (2009) American Medical Association. All rights reserved. Chapter 3, Section 1 is a reprint of Kaddis JS, Danobeitia JS, Niland JC, Stiller T, Fernandez LA. Multicenter analysis of novel and established variables iii associated with successful human islet isolation outcomes. American Journal of Transplantation. 2010;10(3):646-656 and reproduced with permission from John Wiley and Sons via the Copyright Clearance Center. Copyright © (2010) John Wiley and Sons. All rights reserved. Chapter 3, Section 2, Part 1 entitled “Influence of RNA Labeling on High Throughput Expression Profiling of MicroRNAs” is a manuscript in preparation for journal submission. Significant contributions were made by co-authors, Daniel H. Wai, Jessica Bowers, Dr. Nicole Hartmann, Lukas Baeriswyl, Dr. Sheetal Bajaj, Dr. Michael J. Anderson, Dr. Robert C. Getts, and Dr. Timothy J. Triche. This work is a joint collaboration. The research conducted at the Departments of Pathology and Laboratory Medicine, Children's Hospital Los Angeles Saban Research Institute and Keck School of Medicine, University of Southern California (USC), Los Angeles, California, USA, was done so under the direction of Dr. Timothy Triche, with which John S. Kaddis, Daniel H. Wai, Dr. Sheetal Bajaj, and Dr. Michael J. Anderson are also associated. Jessica Bowers and Dr. Robert C. Getts are employed by Genisphere LLC, Hatfield, Pennsylvania, USA. Dr. Nicole Hartmann and Lukas Baeriswyl are employed by the Novartis Institutes for BioMedical Research, Novartis, Basel, Switzerland. Chapter 3, Section 2, Part 2 entitled “Identification of Predictive Genetic Markers Associated with Favorable Human Islet Transplantation Outcomes” is a study iv currently under investigation. Significant contributions were made by co-authors, Drs. Jose Oberholzer, Jan Jensen, and Joyce C. Niland. This work is a joint collaboration. The research conducted at the City of Hope was done so under the direction of Dr. Joyce C. Niland. Dr. Jose Oberholzer resides at the University of Illinois, Chicago and is the principal investigator of this project. Dr. Jan Jensen is affiliated with the Cleveland Clinic. Chapter 4 entitled “Physiological Genetic Modeling of Human Beta Cells In Diabetes: A New Framework for Simulating Complex Disease Genetics” is a study currently under investigation. Significant contributions were made by co- authors, Drs. David V. Conti and Richard Watanabe. This work is a joint collaboration. The research conduced at USC was done so under the direction of Dr. David Conti. Dr. Richard Watanabe also is affiliated with USC. Chapter 5, Section 1 is a reprint of Qian D, Kaddis J, Niland JC. A matching algorithm for the distribution of human pancreatic islets. Computational Statistics and Data Analysis. 2007;51(12):5494-5506 and is reproduced with permission from Elsevier via the Copyright Clearance Center. Copyright © (2007) Elsevier. All rights reserved. Chapter 5, Section 2 is a reprint of Niland JC, Stiller T, Cravens J, Sowinski J, Kaddis J, Qian D. Effectiveness of a web-based automated cell distribution v system. Cell Transplantation. 2010;19(9):1133-42 and is reproduced with permission of the Cognizant Communication Corporation via the Copyright Clearance Center. Copyright © (2010) Cell transplantation by the Cell Transplant Society. All rights reserved. Chapter 5, Section 3 entitled “Standardized Transportation of Human Pancreatic Islets: An Islet Cell Resource (ICR) Center Study of Over 3000 Shipments” is a manuscript in preparation for journal submission. Significant contributions were made by co-authors, James Cravens, Dr. Dajun Qian, Barbara Olack, Martha Antler, Dr. Matthew Hanson, Dr. Klearchos K. Papas, and Dr. Joyce C. Niland. This work is a joint collaboration. The research conducted at City of Hope was done so under the direction of Dr. Joyce C. Niland, with which John S. Kaddis, James Cravens, Dr. Dajun Qian, Barbara Olack, and Martha Antler are also affiliated. Dr. Matthew Hanson performed the research while at the University of Wisconsin, Madison, but has since moved to the Promega Corporation. Dr. Klearchos Papas is located at the University of Minnesota. vi TABLE OF CONTENTS Acknowledgements....................................................................................... ii List of Tables ................................................................................................ ix List of Figures............................................................................................... xi Abstract........................................................................................................xiii Chapter 1: Introduction...............................................................................1 Chapter 1 References .................................................................9 Chapter 2: Human Pancreatic Islets as a Tool in Diabetes Research ..................................................................................19 Chapter 3: Predicting Human Islet Isolation and Transplantation Outcomes in Type 1 Diabetes Section 1: Multicenter Analysis of Novel and Established Variables Associated with Successful Human Islet Isolation Outcomes.........................................................20 Chapter 3.1 Abstract..................................................................20 Chapter 3.1 Introduction............................................................20 Chapter 3.1 Materials and Methods ..........................................21 Chapter 3.1 Results...................................................................22 Chapter 3.1 Discussion .............................................................25 Chapter 3.1 Acknowledgments..................................................28 Chapter 3.1 References ............................................................28 Section 2: Optimization of Microarray Analysis -Techniques: Influence of RNA Labeling on High- Throughput Expression Profiling of MicroRNAs..................31 Chapter 3.2.1 Abstract...............................................................31 Chapter 3.2.1 Introduction.........................................................32 Chapter 3.2.1 Materials and Methods........................................36 Chapter 3.2.1 Results................................................................41 Chapter 3.2.1 Discussion ..........................................................45 Chapter 3.2.1 Conclusion..........................................................48 Chapter 3.2.1 Acknowledgments...............................................49 Chapter 3.2.1 References .........................................................50 -Application: Identification of Predictive Genetic Markers Associated with Favorable Human Islet Transplantation Outcomes .....................................................68 Chapter 3.2.2 Abstract...............................................................68 Chapter 3.2.2 Introduction.........................................................69 vii Chapter 3.2.2 Materials and Methods........................................70 Chapter 3.2.2 Preliminary Results.............................................74 Chapter 3.2.2 Discussion and Future Work...............................76 Chapter 3.2.2 Conclusion..........................................................78 Chapter 3.2.2 Acknowledgements.............................................78 Chapter 3.2.2 References .........................................................80 Chapter 4: Physiological Genetic Modeling of Human Beta Cells In Diabetes: A New Framework for Simulating Complex Disease Genetics .................................90 Chapter 4 Introduction...............................................................90 Chapter 4 Materials and Methods .............................................93 Chapter 4 Preliminary Results...................................................95 Chapter 4 Discussion and Future Work.....................................96 Chapter 4 Conclusion................................................................99 Chapter 4 References .............................................................101 Chapter 5: Improving Availability of Islet Transplantation in Type 1 Diabetes Section 1: A Matching Algorithm for the Distribution of Human Pancreatic Islets.......................................................109 Chapter 5.1 Abstract................................................................109 Chapter 5.1 Introduction..........................................................109 Chapter 5.1 Methods...............................................................111 Chapter 5.1 Results.................................................................116 Chapter 5.1 Discussion ...........................................................119 Chapter 5.1 Acknowledgments................................................120 Chapter 5.1 References ..........................................................120 Section 2: Effectiveness of a Web-Based Automated Cell Distribution System...............................................................122 Chapter 5.2 Abstract................................................................122 Chapter 5.2 Introduction..........................................................122 Chapter 5.2 Materials and Methods ........................................123 Chapter 5.2 Results.................................................................127 Chapter 5.2 Discussion ...........................................................129 Chapter 5.2 Acknowledgments................................................130 Chapter 5.2 References ..........................................................131 Section 3: Standardized Transportation of Human Pancreatic Islets: An Islet Cell Resource (ICR) Center Study of Over 3000 Shipments ................................132 Chapter 5.3 Abstract................................................................132 Chapter 5.3 Introduction..........................................................133 Chapter 5.3 Materials and Methods ........................................136 viii Chapter 5.3 Results.................................................................140 Chapter 5.3 Discussion ...........................................................145 Chapter 5.3 Acknowledgments................................................149 Chapter 5.3 References ..........................................................151 Chapter 6: Conclusion.............................................................................168 Chapter 6 References .............................................................171 Bibliography ..............................................................................................172 ix LIST OF TABLES Table 3.1.1: Organ Donor Demographic Variables Affecting Islet Isolation Success 23 Table 3.1.2: Influence of Donor Medications and Organ Function on Islet Isolation Outcome 24 Table 3.1.3: Pancreas Factors Tested for Contribution to Islet Isolation Outcome 25 Table 3.1.4: Influence of Pancreas Processing Factors on Isolation Outcome 26 Table 3.1.5: Multivariable Logistic Regression Model of Factors Influencing Islet Isolation Success 27 Table 3.2.1.1: Correlation of Replicate Chip Signal Intensity Values 60 Table 3.2.1.S1: Brain Sample Correlations of Processed Signal Intensity Values 61 Table 3.2.1.S2: Lung Sample Correlations of Processed Signal Intensity Values 62 Table 3.2.1.2: Gene List Agreement within and between RNA Labeling Kits 64 Table 3.2.2.1: Characteristics of 26 Human Islet Preparations Used in This Study 86 Table 4.S1: Assumed Model Parameters 106 Table 5.1.1: Actual and MAID-derived Results for the ICR Consortium Data from January–September 2005 116 Table 5.1.2: MAID-derived Results by Rejection Rate for Simulated Data 118 Table 5.2.1: Summary of Pre- and Post-Web Deployment of the Matching Algorithm for Islet Distribution (MAID) 128 Table 5.2.2: Frequency of Received-to-Requested Islet Ratios by Pre- Versus Post-MAID for all Approved Studies 129 x Table 5.2.3: Evaluation of Factors Associated With Received-to- Requested Islet Ratio Postimplementation of MAID 129 Table 5.3.S1: Reproducibility and Precision of Temperature Monitors Used in Study 156 Table 5.3.1: Description of Shipment Data Obtained 158 Table 5.3.S2: Temperature During Packing of Islet Shipping Containers 155 Table 5.3.2: Changes in Shipping Parameters During Transportation of Islet Containers 159 Table 5.3.S3: Two-Way Agreement Between Temperature Monitoring Methods 162 Table 5.3.S4: Effects of Packing Islet Shipment Containers at Different Temperatures 163 Table 5.3.3: Univariate Logistic Regression Analysis of Shipping Conditions Associated with a Temperature Drop <15°C Inside the Container 164 Table 5.3.4: Islet Quality Rankings in Over 3000 Shipments of Human Pancreatic Islets 166 Table 5.3.5: Temperature Control Improvements in Over 1500 Shipments of Human Pancreatic Islets 167 xi LIST OF FIGURES Figure 3.1.1: Distribution of Post Purification IEQs in 1023 Human Islet Isolations 21 Figure 3.2.1.S1: Concordance of Detection Calls within and between RNA Labeling Kits 58 Figure 3.2.1.1: Comparison of Detected miRNA Genes 59 Figure 3.2.1.2: Coefficient of Variation within and between RNA Labeling Kits 63 Figure 3.2.1.3: Evaluation of Statistically Significant Differentially- Expressed Genes Using Biotin-HSR Labeled RNA 65 Figure 3.2.1.4: Correlation of Microarray Data Using TaqMan MicroRNA Assays 66 Figure 3.2.1.5: Dynamic Range of miRNA Gene Chip Using Biotin- HSR Labeling Kit 67 Figure 3.2.2.1: Assessing Human Islet Preparations Using Multivariable Logistic Regression Analysis 87 Figure 3.2.2.2: Agreement Between Composite Donor Score and Human Islet Isolation Outcomes 88 Figure 3.2.2.3: PCA Mapping of 26 High Quality Human Islet Samples 89 Figure 4.1: OGTT Simulation Results 107 Figure 4.2: Variation of Selected Population Parameters from the Glucoregulatory Compartmental Model 108 Figure 5.1.1: Algorithm Schema of MAID 111 Figure 5.1.2: Geographical Locations of 8 Producers and 62 Requesters in an ICR Consortium Study from January–September 2005 114 xii Figure 5.1.3: Distribution of 68 Islet Isolations for the ICR Consortium Data from January–September 2005 using the LDM Model and After Applying MAID 128 Figure 5.1.4: Simulation Results by Islet Supply Versus Demand Ratio 118 Figure 5.1.5: Simulation Results by Rejection Rate 115 Figure 5.2.1: Online Web-Based Islet Offer Broadcasts and Notification Widget 125 Figure 5.2.2: Islet Allocation Process via Targeted and Open Islet Offers 125 Figure 5.2.3: Possible Scenarios for Calculation of the Received- to-Requested Ratio Pre- and Postimplementation of Matching Algorithm for Islet Distribution (MAID) 127 Figure 5.2.4: Distribution of Received-to-Requested Islet Ratios of Pre- Versus Post-MAID for All Approved Studies 125 Figure 5.3.S1: Temperature Profiles During Cross Country Shipments Using Standardized Islet Shipping Containers 160 Figure 5.3.S2: Optimization of Internal Temperature Control During Mock 18-Hour Islet Shipments 165 xiii ABSTRACT Destruction or dysfunction in the human pancreatic islet affects at least 23.6 million affected individuals in the U.S. alone. The inability to regulate insulin production and maintain glucose homeostasis leads to a variety of severe diabetic complications at an estimated 2007 US health care cost of $174 billion dollars. Although medical management, lifestyle changes, and pharmacological agents are successful treatment tools for some, they are less effective, and have failed, in those with unstable diabetes, indicating that an urgent need for alternative therapies exist. Pancreatic islet transplantation is a form of cellular replacement therapy that has been shown to restore glycometabolic control and render some patients insulin independent. Our long term goal is to therefore improve human islet survival and transplantation success rates by understanding the factors affecting cell function in-vitro and in-vivo both in the native pancreas and transplant environments. A survey of the challenges and relevance of human pancreatic islets as a tool in diabetes research is first undertaken using descriptive statistics. Univariate and multivariable logistic regression (MLR) analysis was next used to show that a number of novel and established organ donor, pancreas processing, and islet isolation factors improve the odds of obtaining successful human islet isolation yields. This is an important finding because low islet yield is often the rate- limiting factor for wait-listed transplant recipients. Upon optimization of xiv microarray processing and analysis methods, in a subset of transplant quality human islet preparations, we further demonstrate that genomic variability also exists between samples. This suggests the presence of donor-specific intrinsic islet factors that are distinct from those in our MLR model and is also noteworthy because an understanding of genetic variability in islet quality may help to improve inconsistent graft behavior post-transplant. Using differential equations to physiologically model human islet beta cell function, tools are then developed to assess genetic variability in the normal and diabetic state. Lastly, we demonstrate the feasibility of a national islet matching program by generating a set of mathematical formulas, based on factors previously identified, that are evaluated using experimental and simulated data. This is vital because single- center islet allocation procedures are inefficient and lead to limited availability. The algorithm is then implemented through a web-based automated cell distribution system and further examined prospectively using mixed statistical methods. A descriptive study to develop a standardized islet transportation protocol was then performed to ensure that matched islets could be shipped remotely to clinical research laboratories. When taken together, these studies provide insight into factors affecting human pancreatic islets that may be used to improve therapeutic options for diabetes patients and expand the availability of the procedure. CHAPTER 1: INTRODUCTION Epidemiology of Unstable Diabetes The global prevalence of diabetes is on the rise, from 171 million people in 2000 to 285 million in 2010 1,2 . Of growing concern is a small group of individuals that have unstable diabetes, loosely defined as those who experience frequent episodes of hypoglycemic unawareness 3 , diabetic ketoacidosis (DKA) 4,5 , and/or exhibit poor glycemic control requiring hospitalization 6,7 . Epidemiological evaluations of the number affected are difficult to perform because standardized definitions have been historically controversial 8,9 and remain absent from the literature 10 . Gill et al conducted a survey in the United Kingdom and estimated the brittle diabetes prevalence to be 1.2 in 1000 total diabetics or 2.9 in 1000 insulin treated diabetics, defined as insulin-dependent diabetes mellitus with blood sugar fluctuations of any type leading to life disruption and recurrent and/or prolonged hospitalizations 11 . Islet Transplantation A Cellular Replacement Strategy. When compared to non-diabetics or those with stable disease, brittle or advanced diabetes results in premature mortality and a poorer quality of life 12-14 . This can be explained, in part, by the fact that conventional clinical therapies 15 have been shown to inadequately control unstable diabetes, leading to long-term complications 16 and the need for additional strategies to treat the disease. Whole pancreas replacement is 1 preferred when end-stage renal disease is present, but technically complex and invasive as a solitary procedure 17 . Islet regenerative strategies, reviewed in chapter 2 18 , are promising, but not yet clinical available 19,20 . Islet transplantation is a form of cellular replacement therapy experimentally being used as a minimally invasive clinical procedure where pancreas-extracted islets are infused through the portal vein and delivered to the liver to produce and release the islet hormones necessary for glycemic control 21,22 . The procedure is primarily restricted to a subset with type 1 diabetes (T1D), but is also used in cases where chronic pancreatitis requires removal of the entire organ 23 . It has also been used to treat type 2 diabetes in one cohort undergoing liver transplantation 24 and in a soldier following traumatic wartime injury 25 . The benefits of tight glycemic control have been well documented 16 . Translational Barriers to Widespread Availability. Two recent reports have shown that from 1997 to 2007, 378 individuals in North America received human pancreatic islet transplants as a treatment for T1D 26,27 , compared with only 241 from 1893 through 2000 28 . Clearly, the procedure is gaining clinical utility, as is the use of human islets in basic science laboratory research 18 . These reports, the most comprehensive collection available to date, also demonstrated that in 75% of those achieving insulin independence at least once post-transplant, graft function remained up to 40 months post follow-up. However, as reviewed in chapter 2 18 , insulin independence rates have varied from historic lows of 10% to 2 highs of 80-90% when the procedure was performed at established centers. While now the long-term insulin independence success rates are estimated at around 50% 29 , a monumental advance over earlier eras, addressing inconsistent graft behavior is of critical importance in improving overall transplant performance. The factors affecting graft behavior can be divided into pre and post-transplantation barriers. While interplay between the two is almost certain, the variables prior to transplant are described here in order to explain the relevance of this thesis. Post transplantation challenges, such as transplantation techniques, activation of innate and adaptive immunity, and early loss of transplanted cells have been described elsewhere in detail 30-33 . Limited Donor Supply: With the exception of 1 series of reports that used porcine pancreata 34-36 , living 37 or deceased human organ donations have been the only source of islets used clinically. As discussed in chapter 2 18 , data from the Organ Procurement and Transplantation Network show that cadaveric human pancreata was only recovered in 27% of eligible organ donors from January 1, 2001 through November 30, 2008. Moreover, of those, 74% were used for whole pancreas transplantation. Because deceased organ donors are relied upon almost exclusively for transplant-grade human islets, the small percentage of recoveries are problematic given that nearly 75% of transplant recipients will require greater than 1 islet infusion 26 . In chapter 2 we also showed that the cost of recovering a pancreas ranges from a low of $600 to a high $39,800 and further 3 limits the availability of this organ for research or transplantation purposes. Taken together with the heterogeneous organ donor population 38 , it is not surprising that alternatives to using human pancreata are being developed. Exciting work, reviewed in chapter 2 18 , utilizing surrogates for human beta cells 19,20 , cell lines 39,40 , and nonhuman islets 41 promises to usher in a new era of research, but is currently in the experimental phase. Difficulty and Expense of the Manufacturing Process: Once a pancreas is in hand, liberation of the islets from its native endogenous surroundings is achieved using a process equivalently termed islet isolation or pancreas processing 42 . The technical complexity of the procedure can be visually represented 43,44 but also underscored by the detailed studies performed on all of the following steps of islet isolation: a) organ preservation 45,46 , b) trimming and cleaning of the pancreas 47,48 , c) cannulation for enzymatic perfusion 49-51 , d) digestion and dilution 52 , e) pancreas purification 53,54 , f) islet characterization 55 , and g) culture 56 . Of the many factors involved in isolation, one of the few consistent findings is that extended cold ischemia time detrimentally impacts islet isolation outcome prior to transplantation 38,57-59 . Increasing evidence is also now pointing to the importance in composition and choice of enzyme selection during the perfusion process 60 . An appreciation is also emerging for the experience required 61 by laboratory personal to successfully perform islet isolations for transplantation within the required regulatory framework 62 . Finally, the alarming cost of processing a 4 pancreas 63 has been estimated at $21,000 per pancreas 64 . This does not include the cost of the pancreas, nor does it factor into account facility maintenance costs, estimated to be between $0.8 to $3 million US annually 65 . Such daunting requirements have led to the need for core pancreas processing facilities able to transport islet preparations to distant clinical transplant centers and basic science laboratories 61 . Lack of Predictive Pre-Transplant Assessment Criteria: The “holy grail” of islet characterization is a reliable, validated and easy to perform real-time potency assay, capable of quickly predicting islet mass and function in patients prior to transplantation. The following established clinical transplantation lot release criteria 66 fails to predict or ensure adequate transplant performance but is required as a necessary regulatory condition: a) purity >30%, b) viability >70%, c) stimulation index, from glucose stimulated insulin release assay, > 1, d) islet yield > 5000 islet equivalent (IEQs)/kg bodyweight for initial infusion or > 3000/kg if being re-transplanted, and 5) contamination of < 5 EU/kg. The currently accepted “gold standard” assay to determine in vivo graft function and survival is the post-transplant ability of human pancreatic islets to reverse diabetes in diabetes-induced immuno-incompetent mice 66 . Studies have shown that results from this assay correlate well with clinical outcomes in humans immediately following transplant 67 . However, one of the major limitations of this model relate to the time required for assay completion, typically 28 or more days, which 5 effectively limits the application of this method as a retrospective tool 68 . Therefore, a number of surrogate approaches have been introduced to monitor beta-cell function in vivo or predict graft performance pre-transplant 69-74 . Reviewed in detail 66 , a number of these assays will require additional testing prior to widespread acceptance and implementation. This is of critical importance, as a lack of predictive tools continues to be major hurdle in improving graft performance. Limited Placement and Transportation: The process of organ donation is gaining international attention 75 , as systematic procedures in the US continue to evolve 76 . Anyone in the US who wishes to receive an organ transplant, including islet transplant recipients, must be placed on the United Network for Organ Sharing (UNOS) waiting list 77 . In this way, if a wait-listed transplant recipient is matched to an available organ but, for whatever reason, cannot accept the offer, the organ can continue to be offered to other qualified individuals. This ensures maximization of the limited availability of human organ donations. However, this is not the case with islet transplant recipients. If an islet preparation becomes available but, for whatever reason, cannot be transplanted into the pre- designated local recipient, there is no way to quickly determine if the islets can be used by another islet transplant recipient, either locally or nationally. 6 However, even if this were not the case, strategies for transportation of human pancreatic islets from the manufacturing laboratory to the clinical transplant center are not standardized. Groups in Minneapolis/Portland 78,79 , Miami/Houston 61,80 , Los Angeles/San Francisco 81 , Giessen/Huddinge 82 , Geneva/Budapest 83,84 , the Groupe Rhône-Alpes, Rhin et Geneve pour la transplantation d’Ilots Langerhans (GRAGIL) consortium 85-87 , and the Nordic Network 88 have all reported using an islet processing center at one location for clinical transplantation at another. Of the 7 clinical collaborations listed, none provided any descriptive data on in-transit temperatures; however, one group stated that islets were shipped on ice 79 , another that an insulated cooler was used with gel packs and a temperature monitor 61 , and another that islets were shipped at room temperature in transfer bags with a temperature monitoring unit 86,87 . Second, the methods of islet transportation was described in only 4 collaborations, and included commercial carrier 79,83,84 , charter jet 61,80 , and ambulance 86,87 . Third, the time in transit from the islet laboratory to the transplant center, reported by 2 groups, ranged from a minimum of 1 hour to a maximum of 10 61,85-87 . One additional group combined islet processing and transit time, and reported a maximum of 24 hours 79 . Complementary to time in transit was distance traveled, reported by 3 groups, and ranged from a minimum of 500 kilometers to a maximum of 1500 miles 78,79,83,88 . Finally, assessments of islet sterility, purity, viability, and/or function was provided by 1 group both before and after islet shipment 61,80 , while another included only post-delivery 7 evaluation 83,84 . To-date, the only complete and detailed islet-shipping-for- transplantation protocol was published by Ichii and colleagues, who extended the initial reports from the Miami/Houston experience 77 . In summary, addressing pre and post transplantation barriers offers an optimal solution to improving islet graft function and expanding the availability of the procedure in qualified patients suffering from unstable diabetes. There is a strong body of evidence to suggest that maximizing the limited donor supply, minimizing the variability in manufacturing, improving islet characterization and prediction tools, and expanding placement and transportation strategies may result in long-term benefits to islet transplant recipients. The lessons learned from this work can be applied to other forms of cell replacement and regenerative strategies in diabetes 89 and beyond 90-93 . 8 CHAPTER 1 REFERENCES 1. Wild S, Roglic G, Green A, Sicree R, King H. Global Prevalence of Diabetes. Diabetes Care. May 1, 2004;27(5):1047-1053. 2. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Research and Clinical Practice. 2010;87(1):4-14. 3. Cryer PE, Davis SN, Shamoon H. Hypoglycemia in Diabetes. Diabetes Care. June 1, 2003;26(6):1902-1912. 4. Sivanandan S, Sinha A, Jain V, Lodha R. Management of Diabetic Ketoacidosis. Indian Journal of Pediatrics. 2010:1-9. 5. Koul PB. Diabetic Ketoacidosis: A Current Appraisal of Pathophysiology and Management. Clinical Pediatrics. March 1, 2009;48(2):135-144. 6. Tattersall R. Brittle diabetes. Clin Endocrinol Metab. Jul 1977;6(2):403- 419. 7. Gill GV. The spectrum of brittle diabetes. J R Soc Med. May 1992;85(5):259-261. 8. Amiel SA. "Brittle" diabetes. Bmj. Aug 3 1991;303(6797):260-261. 9. Williams G, Gill GV, Pickup JC. 'Brittle' diabetes. Bmj. Sep 21 1991;303(6804):714. 10. Bertuzzi F, Verzaro R, Provenzano V, Ricordi C. Brittle type 1 diabetes mellitus. Curr Med Chem. 2007;14(16):1739-1744. 11. Gill GV, Lucas S, Kent LA. Prevalence and characteristics of brittle diabetes in Britain. QJM. Nov 1996;89(11):839-843. 9 12. Kent LA, Gill GV, Williams G. Mortality and outcome of patients with brittle diabetes and recurrent ketoacidosis. Lancet. Sep 17 1994;344(8925):778- 781. 13. Secrest AM, Becker DJ, Kelsey SF, LaPorte RE, Orchard TJ. All-Cause Mortality Trends in a Large Population-Based Cohort With Long-Standing Childhood-Onset Type 1 Diabetes. Diabetes Care. December 1, 2010;33(12):2573-2579. 14. Cartwright A, Wallymahmed M, Macfarlane IA, Wallymahmed A, Williams G, Gill GV. The outcome of brittle type 1 diabetes - a 20 year study. QJM. 2011. 15. American Diabetes A. Standards of Medical Care in Diabetes - 2011. Diabetes Care. January 1, 2011;34(Supplement 1):S11-S61. 16. The Diabetes Control and Complications Trial Research Group. The Effect of Intensive Treatment of Diabetes on the Development and Progression of Long-Term Complications in Insulin-Dependent Diabetes Mellitus. New England Journal of Medicine. 1993;329(14):977-986. 17. White SA, Shaw JA, Sutherland DE. Pancreas transplantation. Lancet. May 23 2009;373(9677):1808-1817. 18. Kaddis JS, Olack BJ, Sowinski J, Cravens J, Contreras JL, Niland JC. Human Pancreatic Islets and Diabetes Research. JAMA: The Journal of the American Medical Association. 04/15 2009;301(15):1580-1587. 19. Aguayo-Mazzucato C, Bonner-Weir S. Stem cell therapy for type 1 diabetes mellitus. Nat Rev Endocrinol. 2010;6(3):139-148. 20. Halban PA, German MS, Kahn SE, Weir GC. Current Status of Islet Cell Replacement and Regeneration Therapy. J Clin Endocrinol Metab. March 1, 2010;95(3):1034-1043. 21. Ichii H, Ricordi C. Current status of islet cell transplantation. J Hepatobiliary Pancreat Surg. 2009;16(2):101-112. 10 22. Lakey JR, Mirbolooki M, Shapiro AM. Current status of clinical islet cell transplantation. Methods Mol Biol. 2006;333:47-104. 23. Blondet JJ, Carlson AM, Kobayashi T, et al. The role of total pancreatectomy and islet autotransplantation for chronic pancreatitis. Surg Clin North Am. Dec 2007;87(6):1477-1501, x. 24. Ricordi C, Alejandro R, Angelico MC, et al. Human islet allografts in patients with type 2 diabetes undergoing liver transplantation. Transplantation. Feb 15 1997;63(3):473-475. 25. Jindal RM, Ricordi C, Shriver CD. Autologous Pancreatic Islet Transplantation for Severe Trauma. New England Journal of Medicine. 2010;362(16):1550-1550. 26. 2007 update on allogeneic islet transplantation from the Collaborative Islet Transplant Registry (CITR). Cell Transplant. 2009;18(7):753-767. 27. Alejandro R, Barton FB, Hering BJ, Wease S. 2008 Update from the Collaborative Islet Transplant Registry. Transplantation. Dec 27 2008;86(12):1783-1788. 28. Brendel MD, Hering BJ, Schultz AO, Bretzel RG. International Islet Transplant Registry, Newsletter 9, Vol 8 (No 1), June 2001. Accessed online December 16, 2010 at: http://www.med.uni- giessen.de/itr/newsletter/no_9/news_9.html. 29. Robertson RP. Islet Transplantation a Decade Later and Strategies for Filling a Half-Full Glass. Diabetes. 06/01 2010;59(6):1285-1291. 30. Barshes NR, Wyllie S, Goss JA. Inflammation-mediated dysfunction and apoptosis in pancreatic islet transplantation: implications for intrahepatic grafts. J Leukoc Biol. May 1, 2005;77(5):587-597. 31. Huang X, Moore DJ, Ketchum RJ, et al. Resolving the Conundrum of Islet Transplantation by Linking Metabolic Dysregulation, Inflammation, and Immune Regulation. Endocr Rev. August 1, 2008;29(5):603-630. 11 32. Korsgren O, Lundgren T, Felldin M, et al. Optimising islet engraftment is critical for successful clinical islet transplantation. Diabetologia. Feb 2008;51(2):227-232. 33. Westermark GT, Westermark P, Berne C, Korsgren O, the Nordic Network for Clinical Islet T. Widespread Amyloid Deposition in Transplanted Human Pancreatic Islets. N Engl J Med. August 28, 2008 2008;359(9):977-979. 34. Elliott RB, Escobar L, Garkavenko O, et al. No evidence of infection with porcine endogenous retrovirus in recipients of encapsulated porcine islet xenografts. Cell Transplant. Nov-Dec 2000;9(6):895-901. 35. Groth CG, Korsgren O, Tibell A, et al. Transplantation of porcine fetal pancreas to diabetic patients. Lancet. Nov 19 1994;344(8934):1402-1404. 36. Valdes-Gonzalez RA, Dorantes LM, Garibay GN, et al. Xenotransplantation of porcine neonatal islets of Langerhans and Sertoli cells: a 4-year study. European journal of endocrinology / European Federation of Endocrine Societies. Sep 2005;153(3):419-427. 37. Reynoso J, Gruessner C, Sutherland D, Gruessner R. Short- and long- term outcome for living pancreas donors. Journal of Hepato-Biliary- Pancreatic Sciences. 2010;17(2):92-96. 38. Lakey JR, Warnock GL, Rajotte RV, et al. Variables in organ donors that affect the recovery of human islets of Langerhans. Transplantation. Apr 15 1996;61(7):1047-1053. 39. Hohmeier HE, Newgard CB. Cell lines derived from pancreatic islets. Molecular and cellular endocrinology. Dec 30 2004;228(1-2):121-128. 40. Narushima M, Kobayashi N, Okitsu T, et al. A human beta-cell line for transplantation therapy to control type 1 diabetes. Nature biotechnology. Oct 2005;23(10):1274-1282. 12 41. Kirchhof N, Shibata S, Wijkstrom M, et al. Reversal of diabetes in non- immunosuppressed rhesus macaques by intraportal porcine islet xenografts precedes acute cellular rejection. Xenotransplantation. 2004;11(5):396-407. 42. Lakey JRT, Burridge PW, Shapiro AMJ. Technical aspects of islet preparation and transplantation. Transplant International. 2003;16(9):613- 632. 43. Qi M, Barbaro B, Wang S, Wang Y, Hansen M, Oberholzer J. Human pancreatic islet isolation: Part I: digestion and collection of pancreatic tissue. J Vis Exp. 2009(27). 44. Qi M, Barbaro B, Wang S, Wang Y, Hansen M, Oberholzer J. Human pancreatic islet isolation: Part II: purification and culture of human islets. J Vis Exp. 2009(27). 45. Salehi P, Hansen MA, Avila JG, et al. Human islet isolation outcomes from pancreata preserved with Histidine-Tryptophan Ketoglutarate versus University of Wisconsin solution. Transplantation. Oct 15 2006;82(7):983- 985. 46. Qin H, Matsumoto S, Klintmalm GB, De Vol EB. A Meta-analysis for Comparison of the Two-Layer and University of Wisconsin Pancreata Preservation Methods in Islet Transplantation. Cell Transplant. Nov 19 2010. 47. Kin T, Rosichuk S, Shapiro AM, Lakey JR. Detection of microbial contamination during human islet isolation. Cell Transplant. 2007;16(1):9- 13. 48. Bucher P, Oberholzer J, Bosco D, et al. Microbial surveillance during human pancreatic islet isolation. Transpl Int. May 2005;18(5):584-589. 49. Lakey JR, Warnock GL, Shapiro AM, et al. Intraductal collagenase delivery into the human pancreas using syringe loading or controlled perfusion. Cell Transplant. May-Jun 1999;8(3):285-292. 13 50. Kin T, Zhai X, Murdoch TB, Salam A, Shapiro AMJ, Lakey JRT. Enhancing the Success of Human Islet Isolation Through Optimization and Characterization of Pancreas Dissociation Enzyme. American Journal of Transplantation. 2007;7(5):1233-1241. 51. Goto M, Eich TM, Felldin M, et al. Refinement of the automated method for human islet isolation and presentation of a closed system for in vitro islet culture. Transplantation. Nov 15 2004;78(9):1367-1375. 52. Ricordi C, Lacy PE, Finke EH, Olack BJ, Scharp DW. Automated method for isolation of human pancreatic islets. Diabetes. Apr 1988;37(4):413-420. 53. Adewola A, Mage R, Hansen M, et al. Comparing cooling systems for the COBE 2991 cell separator used in the purification of human pancreatic islets of Langerhans. Cryo Letters. Jul-Aug 2010;31(4):310-317. 54. Barbaro B, Salehi P, Wang Y, et al. Improved human pancreatic islet purification with the refined UIC-UB density gradient. Transplantation. Nov 15 2007;84(9):1200-1203. 55. Yamamoto T, Horiguchi A, Ito M, et al. Quality control for clinical islet transplantation: organ procurement and preservation, the islet processing facility, isolation, and potency tests. J Hepatobiliary Pancreat Surg. 2009;16(2):131-136. 56. Murdoch TB, McGhee-Wilson D, Shapiro AM, Lakey JR. Methods of human islet culture for transplantation. Cell Transplant. 2004;13(6):605- 617. 57. O'Gorman D, Kin T, Murdoch T, et al. The standardization of pancreatic donors for islet isolations. Transplantation. Sep 27 2005;80(6):801-806. 58. Ponte GM, Pileggi A, Messinger S, et al. Toward maximizing the success rates of human islet isolation: influence of donor and isolation factors. Cell Transplant. 2007;16(6):595-607. 14 59. Zeng Y, Torre MA, Karrison T, Thistlethwaite JR. The correlation between donor characteristics and the success of human islet isolation. Transplantation. Mar 27 1994;57(6):954-958. 60. Balamurugan AN, Breite AG, Anazawa T, et al. Successful human islet isolation and transplantation indicating the importance of class 1 collagenase and collagen degradation activity assay. Transplantation. Apr 27 2010;89(8):954-961. 61. Goss JA, Goodpastor SE, Brunicardi FC, et al. Development of a human pancreatic islet-transplant program through a collaborative relationship with a remote islet-isolation center. Transplantation. Feb 15 2004;77(3):462-466. 62. Zoon KC. Dear colleague letter to transplant centers: allogeneic pancreatic islets for transplantation. 2000. 63. Guignard AP, Oberholzer J, Benhamou P-Y, et al. Cost Analysis of Human Islet Transplantation for the Treatment of Type 1 Diabetes in the Swiss-French Consortium GRAGIL. Diabetes Care. 04/01 2004;27(4):895- 900. 64. Frank A, Deng S, Huang X, et al. Transplantation for type I diabetes: comparison of vascularized whole-organ pancreas with isolated pancreatic islets. Ann Surg. Oct 2004;240(4):631-640; discussion 640-633. 65. Lakey JR, Ricordi C, Hering BJ. Update and New Features in cGMP Facility, Equipment and Structure of the Clinical Islet Transplant Program. Paper presented at: Human Islet Isolation and Transplantation Techniques (HIITT) 6th Training Workshop; April 9-12, 2006; Snowbird, Utah. 66. Papas KK, Suszynski TM, Colton CK. Islet assessment for transplantation. Current Opinion in Organ Transplantation. 2009;14(6):674-682 67. Caiazzo R, Gmyr V, Kremer B, et al. Quantitative In Vivo Islet Potency Assay in Normoglycemic Nude Mice Correlates With Primary Graft Function After Clinical Transplantation. Transplantation. 2008;86(2):360- 363. 15 68. Bertuzzi F, Ricordi C. Prediction of Clinical Outcome in Islet Allotransplantation. Diabetes Care. February 2007;30(2):410-417. 69. Ryan EA, Paty BW, Senior PA, Lakey JRT, Bigam D, Shapiro AMJ. Beta- Score: An assessment of β-cell function after islet transplantation Diabetes Care. February 1, 2005;28(2):343-347. 70. Baidal DA, Faradji RN, Messinger S, et al. Early metabolic markers of islet allograft dysfunction. Transplantation. Mar 15 2009;87(5):689-697. 71. Rickels MR, Naji A, Teff KL. Acute insulin responses to glucose and arginine as predictors of beta-cell secretory capacity in human islet transplantation. Transplantation. Nov 27 2007;84(10):1357-1360. 72. Papas KK, Colton CK, Qipo A, et al. Prediction of marginal mass required for successful islet transplantation. J Invest Surg. Feb 2010;23(1):28-34. 73. Ichii H, Inverardi L, Pileggi A, et al. A novel method for the assessment of cellular composition and beta-cell viability in human islet preparations. Am J Transplant. Jul 2005;5(7):1635-1645. 74. Hanson MS, Park EE, Sears ML, et al. A simplified approach to human islet quality assessment. Transplantation. May 27 2010;89(10):1178-1188. 75. Domínguez-Gil B, Delmonico FL, Shaheen FAM, et al. The critical pathway for deceased donation: reportable uniformity in the approach to deceased donation. Transplant International. 2011;24(4):373-378. 76. Ojo AO, Heinrichs D, Emond JC, et al. Organ donation and utilization in the USA. American Journal of Transplantation. 2004;4:27-37. 77. Ichii H, Sakuma Y, Pileggi A, et al. Shipment of human islets for transplantation. Am J Transplant. Apr 2007;7(4):1010-1020. 16 78. Rabkin JM, Leone JP, Sutherland DE, et al. Transcontinental shipping of pancreatic islets for autotransplantation after total pancreatectomy. Pancreas. Nov 1997;15(4):416-419. 79. Rabkin JM, Olyaei AJ, Orloff SL, et al. Distant processing of pancreas islets for autotransplantation following total pancreatectomy. Am J Surg. May 1999;177(5):423-427. 80. Goss JA, Schock AP, Brunicardi FC, et al. Achievement of insulin independence in three consecutive type-1 diabetic patients via pancreatic islet transplantation using islets isolated at a remote islet isolation center. Transplantation. Dec 27 2002;74(12):1761-1766. 81. Brunicardi FC, Atiya A, Stock P, et al. Clinical islet transplantation experience of the University of California Islet Transplant Consortium. Surgery. Dec 1995;118(6):967-971; discussion 971-962. 82. Tibell A, Bolinder J, Hagström-Toft E, et al. Experience with human islet transplantation in Sweden. Transplantation Proceedings. 2001;33(4):2535-2536. 83. Mathe Z, Langer RM, Jaray J, et al. Human Pancreatic Islet Transplantation in International Collaboration With A Distant Islet Isolation Center: Preliminary Results From the Budapest-Geneva Experience. Transplantation. 2004;78(2):353. 84. Langer RM, Mathe Z, Doros A, et al. Successful islet after kidney transplantations in a distance over 1000 kilometres: Preliminary results of the Budapest-Geneva collaboration. Transplant Proc. Dec 2004;36(10):3113-3115. 85. Benhamou PY, Oberholzer J, Toso C, et al. Human islet transplantation network for the treatment of Type I diabetes: first data from the Swiss- French GRAGIL consortium (1999-2000). Groupe de Recherche Rhin Rhjne Alpes Geneve pour la transplantation d'Ilots de Langerhans. Diabetologia. Jul 2001;44(7):859-864. 17 86. Kempf MC, Andres A, Morel P, et al. Logistics and transplant coordination activity in the GRAGIL Swiss-French multicenter network of islet transplantation. Transplantation. May 15 2005;79(9):1200-1205. 87. Kessler L, Bucher P, Milliat-Guittard L, et al. Influence of islet transportation on pancreatic islet allotransplantation in type 1 diabetic patients within the Swiss-French GRAGIL network. Transplantation. Apr 27 2004;77(8):1301-1304. 88. Lundgren T, Korsgren O, Foss A, et al. The Nordic Network for Clinical Islet Transplantation - Islet After Kidney (Iak) - Results and Experiences From the First 44 Transplantations. Transplantation. 2004;78(2):178. 89. Couri CE, Oliveira MC, Stracieri AB, et al. C-peptide levels and insulin independence following autologous nonmyeloablative hematopoietic stem cell transplantation in newly diagnosed type 1 diabetes mellitus. Jama. Apr 15 2009;301(15):1573-1579. 90. Hare JM, Traverse JH, Henry TD, et al. A randomized, double-blind, placebo-controlled, dose-escalation study of intravenous adult human mesenchymal stem cells (prochymal) after acute myocardial infarction. J Am Coll Cardiol. Dec 8 2009;54(24):2277-2286. 91. Ohira M, Ishiyama K, Tanaka Y, et al. Adoptive immunotherapy with liver allograft-derived lymphocytes induces anti-HCV activity after liver transplantation in humans and humanized mice. J Clin Invest. Nov 2009;119(11):3226-3235. 92. Macchiarini P, Jungebluth P, Go T, et al. Clinical transplantation of a tissue-engineered airway. Lancet. Dec 13 2008;372(9655):2023-2030. 93. Sueblinvong V, Weiss DJ. Stem cells and cell therapy approaches in lung biology and diseases. Transl Res. Sep 2010;156(3):188-205. 18 CHAPTER 2: HUMAN PANCREATIC ISLETS AS A TOOL IN DIABETES RESEARCH Kaddis JS, Olack BJ, Sowinski J, Cravens J, Contreras JL, Niland JC. Human pancreatic islets and diabetes research. Journal of the American Medical Association. 2009; 301(15):1580-1587 is available through the weblink, http://jama.ama-assn.org/content/301/15/1580.full.pdf+html. Authorization was granted for inclusion of this document in electronic format as a weblink only. This article is hosted on the journal web site and is available with permission of the American Medical Association. Copyright © (2009) American Medical Association. All rights reserved. 19 American Journal of Transplantation 2010; 10: 646–656 Wiley Periodicals Inc. C 2009 The Authors Journal compilation C 2009 The American Society of Transplantation and the American Society of Transplant Surgeons doi: 10.1111/j.1600-6143.2009.02962.x Multicenter Analysis of Novel and Established Variables Associated with Successful Human Islet Isolation Outcomes J. S. Kaddis a,† , J. S. Danobeitia b,† ,J.C.Niland a , T. Stiller a and L. A. Fernandez b, * a Administrative and Bioinformatics Coordinating Center, City of Hope National Medical Center and Beckman Research Institute, Division of Information Sciences, Duarte, CA b Division of Transplantation, Department of Surgery, University of Wisconsin–Madison School of Medicine and Public Health, Madison, WI * Corresponding author: Luis A. Fernandez, luisf@surgery.wisc.edu †These authors contributed equally to this work. Islet transplantation is a promising therapy used to achieve glycometabolic control in a select subgroup of individuals with type I diabetes. However, features that characterize human islet isolation success prior to transplantation are not standardized and lack valida- tion. We conducted a retrospective analysis of 806 iso- lation records from 14 pancreas-processing laborato- ries, considering variables from relevant studies in the last 15 years. The outcome was defined as postpurifi- cation islet equivalent count, dichotomized into yields ≥315000 or≤220000. Univariate analysis showed that donorcauseofdeathanduseofhormonalmedications negatively influenced outcome. Conversely, pancreata fromheavierdonorsandthosecontainingelevatedlev- els of surface fat positively influence outcome, as did heavier pancreata and donors with normal amylase levels. Multivariable logistic regression analysis iden- tified the positive impact on outcome of surgically in- tact pancreata and donors with normal liver function, and confirmed that younger donors, increased body mass index, shorter cold ischemia times, no admin- istration of fluid/electrolyte medications, absence of organ edema, use of University of Wisconsin preserva- tion solution and a fatty pancreas improves outcome. In conclusion, this multicenter analysis highlights the importance of carefully reviewing all donor, pancreas andprocessingparameterspriortoisolationandtrans- plantation. Key words: Donor selection, islet isolation, organ do- nation, pancreas preservation before isolation, pan- creas utilization, pancreatic islet transplantation Received 30 June 2009, revised 13 October 2009 and accepted for publication 04 November 2009 Introduction Pancreaticislettransplantation(PIT)hasbeenshowntobe aneffectivetherapeuticstrategytoachieveglycometabolic control in a subset of patients diagnosed with type I dia- betes mellitus (1). A recent report confirmed that in sub- jects achieving insulin independence at least once after the first islet infusion, graft function persisted in approxi- mately 75% of these individuals 40 months post follow- up (2). However, PIT is still an experimental procedure, due in part to the complexities surrounding human islet isolation and transplantation techniques, activation of in- nate and adaptive immunity and early loss of transplanted cells (3,4). Furthermore, the ultimate goal of insulin inde- pendence, alongside a diminishing incidence of secondary complications, remains elusive for many of those with la- bile disease (1,5). Overall pancreas quality has been documented as a major determinant of islet isolation yield prior to transplantation. Factors such as donor age (6–10), extended cold ischemia time(CIT)(6–9) and history ofdiabetes (6,8,9) have consis- tently been found to negatively impact isolation outcome. In contrast, organs from brain-dead donors with elevated body mass index (BMI) and increased pancreas weight (6,7,9,11–15)havebeendocumentedtobeamongthebest predictorsofsuccessfulisolationoutcome.Othervariables such as donation after cardiac death (DCD) (16,17), type of enzyme used in the digestion process (9,10,14,18), cause of death (6,7) and organ preservation methods (19–26) are still controversial with respect to their role in isolation out- come. In addition, a standard definition of isolation success re- mains a matter of debate. In 1996, Lakey et al. defined a successful isolation as a postpurification count above 100000 islet equivalents (IEQs), a standard also used in a study conducted 10 years later (6,27). Recently, others haveincreasedthecut-offpointforsuccessto250000IEQ (10,14). However, no widespread consensus has emerged as to what constitutes a successful isolation, which may help to explain why findings from previous analyses range from consistent to contradictory. Although the body of literature addressing this topic is considerable, most research has been restricted to 646 CHAPTER 3: PREDICTING HUMAN ISLET ISOLATION AND TRANSPLANTATION OUTCOMES IN TYPE 1 DIABETES Section 3.1: Multicenter Analysis of Novel and Established Variables Associated with Successful Human Islet Isolation Outcomes 20 Multicenter Analysis of Variables Associated with Islet Isolation Yield Figure 1: Distribution of Post Purifi- cation IEQs in 1023 human islet iso- lations. IEQ counts were divided into binsby100000IEQincrements.Thelast bar includes an isolation with a count of 1 132 083. The range and distribution of IEQ count was used to define the study outcome, as described in the Materials and Methods section. single-center experiences, a limited number of observa- tions, noncomparable standards of success and divergent intercentermethodologiesthattendtorestrictoverallappli- cability. Moreover, current United Network of Organ Shar- ing (UNOS) allocation policy for PIT is generally limited to organsfromdonorsabove50yearsofagewithBMIgreater than 30 kg/m 2 when suitable recipients for solid pancreas transplantarenotidentifiedatthelocal,regionalornational level (28). Therefore, we conducted this study to establish reliable markers of successful isolation yield and identify factors associated with outcome using data from the first and largest multicenter database of its kind, collected by the Islet Cell Resource (ICR) Center Consortium. The ICR Consortium has been in operation for 8 years, with data from 14 centers, resulting in a robust set of isolation and mergeddonor data, toallow fora multicenteredevaluation of factors associated with successful islet yield. Materials and Methods Human pancreatic islet isolation records A total of 1122 pancreas islet isolation records were entered into the ICR database from December 3, 2004 to October 27, 2008 for human pancre- ata obtained between July 22, 2001 through October 25, 2008 from 14 laboratories across the United States. Of those, 1094 records were linked to pancreas donor data from UNOS. IEQ counts for isolations at the post- purification stage were available for only 1023 records. Selection and definition of donor and isolation variables A literature search was performed using PubMed and combined the use of medical subject heading terms and free form text searches to identify origi- nal studies, from January 1, 1994 to January 31, 2009, that analyzed the in- fluenceofpancreasdonorand/orprocessingvariablesonisolationoutcome. A flow chart outlining our search strategy is provided as Supporting Figure S1. Thirty-one articles were identified (6,7,9,10,12–14,18,20,23,24,27,29– 47). Additional variables were obtained from review of the ICR/UNOS data sets, including donor personal and medical history, management after ad- mittance into a clinical facility, factors related to pancreas procurement, preservation, handling and data related to technical aspects of islet produc- tion. A definition for each variable used is provided as Supporting Table S1. All data for variables analyzed were solely obtained from the ICR/UNOS data sets and does not include the use of results from other previously published works. Study outcome The range and distribution of postpurification IEQs were evaluated to help identify ‘low’ and ‘high’ yielding isolations (Figure 1). Isolations were grouped into quintiles as follows: first quintile (667–145968 IEQs), sec- ondquintile(146413–221612),thirdquintile(221684–312758),fourthquin- tile (312774–432921) and fifth quintile (432929–1132083). Isolations with IEQs in the middle quintile (n = 217) were excluded from further anal- ysis because of known variability in IEQ measurements both within and between centers (48). Exclusion of the middle quintile reduces potential intra/intervariability in IEQ measurements and helps strengthen the validity of any association with a successful yield. Based on this approach, a low- yielding (unsuccessful) isolation was defined as those records with IEQ counts ≤220,000 (n = 402) and high-yielding (successful) as those with IEQ counts ≥315,000 (n = 404). A total of 806 records remained and were analyzed to help determine predictors that may increase the odds of a high yield. Postpurification purity and viability data were considered in defining the study outcome, but not used due to missing data in 124 and 617 records, respectively. Although purity was excluded as a measure of isolation suc- cess, a comparison between high- and low-yielding preparations was gen- erated using a two-sample t-test and an F-test to examine the normality of the variable (F-test p-value not reported). Univariate and multivariable analysis Percentages were reported for all categorical variables. The measure of central tendency was described using either mean (±1 SD) or median (min, American Journal of Transplantation 2010; 10: 646–656 647 21 Kaddis et al. max) for all continuous variables, depending on whether a variable was normally distributed. Coding of continuous variables to categorical group- ings was based on distribution of the variable for age (data not shown) and established laboratory reference values for kidney, liver and pancreas assays. To test for center effect, the laboratory with the largest num- ber of isolations reported in the database was selected as the baseline group. All variables were included for univariate logistic regression (LR) model testing. Chi-square p-values, the corresponding odds ratio (OR) and the 95% confidence interval (95% CI) are reported for all categorical variables. Inthecaseofcontinuousvariables,theORand95%CIarereportedinunits definedusingtheinterquartilerangebetweenthe75thand25thpercentiles, allowing for a direct comparison of OR values among all continuous and categorical predictors. The profile likelihood method was used to generate the reported statistical measures. To test for simultaneous effects of multiple predictors, all variables found to be significant from univariate LR analysis with a p-value of <0.20 were considered in defining an optimal multivariable model. If a variable was not found to be statistically significant in the analysis, but previously shown to be significant in two or more studies, it was also included in the model for testing. A matrix of significant pair-wise correlations was generated to identify possible collinear variables, defined as any two terms with a p-value <0.05 and a Pearson’s correlation coefficient of ≥0.80. Multiclass categorical variables were independently screened for colinearity using a chi-square test. All variables were added into the model and the factor with the highest p-value was removed. This process was continued until only those variables with p-values of <0.10 remained. Interaction terms were generated using a priori hypotheses for greater-than-additive effects on the dependent variable and included organ intact by procurement team, preservation solution by CIT, donor medications received by amylase test, collagenase provider by age and procurement team by CIT. Regression diagnostics were used to examine influential isolation records and outliers. The Hosmer–Lemeshow test was used to fit the model. This test may not be appropriate in all cases (49), but demonstrated to perform sufficiently in datasetswithalargesamplesize,presenceofcontinuousvariablesandlack of significant interaction terms (50), as is reflected in our data. All statistical analysiswasperformedusingSASsoftwareversion9.1.312(SASInstitute, Cary, NC). Results A total of 806 records were analyzed to help determine predictors that increase the odds of achieving isolation success. Purity A statistically significant difference in postpurification pu- rity was found between high (n = 349, 63.1% ± 21.3) and low (n = 333, 59.7% ± 22.6) yielding isolations (p = 0.0396). Univariate analysis Several donor demographic variables were found to cor- relate significantly with successful isolation outcome (Table 1). Relative to younger donors, pancreata from in- dividuals over 47.5 years of age were less likely to result in high-yielding isolations (OR = 0.61, p = 0.0004). Cause of deathwasshowntobeimportant(overallp =0.032);inpar- ticular, individuals succumbing to head trauma were less likelytoresultinhigh-yieldingisolationscomparedtothose inthecerebrovascular/strokegroup(OR =0.69,p =0.015). Although mechanism of death was only marginally signifi- cant (p = 0.053), the results were complementary to that of cause of death; relative to the blunt injury group, pan- creata from donors in the intracranial hemorrhage/stroke group were more likely to result in high-yielding isolations (OR = 1.50, p = 0.020). An increase in donor weight improved the odds of obtaining a high-yielding isolation (OR = 2.12 per an increase equal to the interquartile range valueof27kg,p < 0.0001), as did an increase in BMI (OR = 1.82 per an increase of 8.09 kg/m 2 ,p < 0.0001). A pancreas from a DCD donor was somewhat less likely to result in a high-yielding isolation (OR = 0.46, p = 0.058). Overall organ function and the number and type of med- ication given to the donor just prior to donation affected isolation outcome (Table 2). An increase in the number of medications(irrespectiveoftheirtype)administeredsimul- taneously to the donor less than 24 h prior to cross-clamp reduced the odds of obtaining a high-yielding isolation (OR = 0.87 per an increase of seven medications, p = 0.027). When medications were classified into groups according to mechanism of action, pancreata from donors receiving hormonal medications were less likely to result in high-yielding isolations (OR = 0.61, p = 0.002). More- over, pancreata from donors requiring fluid and electrolyte resuscitation showed a decrease in the likelihood of a suc- cessful outcome (OR = 0.58, p = 0.006). A list of medica- tions for each of the six groups is provided as Supporting Table S2. Finally, pancreata from donors with amylase lev- els of ≤120 l/L were more likely to result in high-yielding isolations (OR = 1.51, p = 0.016). Once the pancreas was recovered, factors related to mor- phology and surgical handling of the organ were also shown to influence isolation outcome (Table 3). The type of solution used to preserve the pancreas following organ procurement was found to be statistically significant (over- all p = 0.0023). Although this difference was not seen be- tweentheuseofaUniversityofWisconsin(UW)alonever- sustwo-layermethod(TLM;p =0.39),theuseofhistidine– tryptophan–ketoglutarate (HTK) reduced the odds of ob- taining a high-yielding isolation (OR = 0.42, p = 0.0005). In addition, pancreata classified as edematous and those with prolonged CIT greatly reduced the odds of obtaining a high-yielding isolation (OR = 0.49, p = 0.003; OR = 0.82 per increase of 4.9 h, p = 0.012, respectively). Relative to pancreata with clean or light surface fat, those with mod- erate or heavy surface fat were nearly twice as likely to result in high-yielding isolations (OR = 1.75, p = 0.0006). Likewise, pancreata with moderate or heavy fat infiltration were more likely to result in high-yielding isolations rel- ative to those organs with no or patchy fat (OR = 1.63, p = 0.002). An increase in the predistention weight of the pancreas also improved the odds of obtaining high yields (OR = 1.55 per increase of 38.5 g, p < 0.0001). 648 American Journal of Transplantation 2010; 10: 646–656 22 Multicenter Analysis of Variables Associated with Islet Isolation Yield Table 1: Organ donor demographic variables affecting islet isolation success ≤220000 IEQs ≥315000 IEQs % or Mean % or Mean Odds Name N (±1SD) N (±1 SD) ratio 1 95% CI p-Value Age (years) 00.1–47.5 182 44% 233 56% – >47.5–70.8 220 56% 171 44% 0.61 (0.46, 0.80) 0.0004 Gender Female 192 53% 170 47% – Male 210 47% 234 53% 1.26 (0.95, 1.66) 0.11 Weight 2 (kg) 402 81.4 (±19.2) 404 93.0 (±22.3) 2.12 (1.74, 2.61) <0.0001 BMI 2 (kg/m 2 ) 402 27.8 (±5.8) 404 30.8 (±7.0) 1.82 (1.51, 2.22) <0.0001 Cause of death 0.032 Cerebrovascular/stroke 209 47% 236 53% – Head trauma 156 56% 121 44% 0.69 (0.51, 0.93) 0.018 Other 2 34 45% 42 55% 1.09 (0.67, 1.79) 0.72 Mechanism of death 0.053 Blunt injury 110 58% 81 42% – Intracranial hemorrhage/stroke 212 48% 234 52% 1.50 (1.07, 2.11) 0.020 All other categories 3 68 47% 76 53% 1.52 (0.98, 2.35) 0.06 Donation after cardiac death (DCD) No 383 49% 395 51% – Yes 19 68% 9 32% 0.46 (0.20, 1.00) 0.058 History of diabetes No 390 49% 398 51% – Yes 10 63% 6 37% 0.59 (0.20, 1.60) 0.31 Insulin dependent No 5 45% 6 55% – Yes 4 100% 0 0% NA NA NA History of hypertension No 257 51% 250 49% – Yes 143 49% 148 51% 1.06 (0.80, 1.42) 0.67 Heavy alcohol use No 287 55% 233 45% – Yes 65 55% 54 45% 1.02 (0.68, 1.53) 0.91 History of cigarette use No 241 49% 248 51% – Yes 158 51% 152 49% 0.94 (0.70, 1.24) 0.64 Cocaine use No 354 50% 350 50% – Yes 42 47% 47 53% 1.13 (0.73, 1.77) 0.58 Other drug use No 298 50% 293 50% – Yes 95 48% 104 52% 1.11 (0.81, 1.54) 0.51 1 Odds ratios, confidence intervals and p-values were calculated using univariate logistic regression. Dashes indicate baseline category. 2 Odds ratio and 95% CI for continuous variables were reported between 75th and 25th percentiles. 3 Other reported causes of death include anoxia (n = 65), CNS tumor (n = 7), bacterial meningitis (n = 1), cardio-pulmonary arrest (n = 1), cerebral edema (n = 1) and spontaneous cranial bleed (n = 1). 4 Other mechanisms of death include gunshot wound (n = 72), cardiovascular (n = 36), drug intoxication (n = 12), asphyxiation (n = 8), natural causes (n = 7), drowning (n = 5) and seizure (n = 4). No technical isolation factors that were found to signifi- cantly improve the odds of obtaining a high-yielding isola- tion (Table 4). Multivariable analysis Variables selected for further analysis were assessed us- ing a multivariable logistic regression model to determine if a multifactorial influence on isolation outcome existed. Although history of diabetes had been found to be signifi- cant in two or more previous studies, it was not included in the multivariable model for testing due to small sample response size. In addition, mechanism of death was found tobecollinearwithcauseofdeathandexcludedfromtest- ing; likewise, weight was found to be collinear with BMI and also excluded from multivariable model testing. The final model yielded a Hosmer–Lemeshow p- value of 0.42, indicating a good fit of the model American Journal of Transplantation 2010; 10: 646–656 649 23 Kaddis et al. Table 2: Influence of donor medications and organ function on islet isolation outcome ≤220000 IEQs ≥315000 IEQs % or Mean % or Mean Odds Name N (±1SD) N (±1 SD) ratio 1 95% CI p-Value Number of donor man meds received 2 402 6.0 (±3.4) 404 5.5 (±3.2) 0.87 (0.76, 0.98) 0.027 Anesthetics received 3 No 362 50% 363 50% – Yes 40 49% 41 51% 1.02 (0.65, 1.62) 0.93 Antibiotics received 3 No 232 48% 248 52% – Yes 170 52% 156 48% 0.86 (0.65, 1.14) 0.29 Cardiovascular meds received 3 No 48 44% 61 56% – Yes 354 51% 343 49% 0.76 (0.51, 1.14) 0.19 Hormonal meds received 3 No 97 41% 138 59% – Yes 305 53% 266 47% 0.61 (0.45, 0.83) 0.002 Fluids/electrolytes received 3 No 326 48% 356 52% – Yes 76 61% 48 39% 0.58 (0.39, 0.85) 0.006 Other meds received 3,4 No 70 43% 91 57% – Yes 332 51% 313 49% 0.73 (0.51, 1.03) 0.07 Total number of donated organs 5 0.98 1–2 12 50% 12 50% – 3–4 272 50% 276 50% 1.02 (0.44, 2.32) 0.97 5–6 118 50% 116 50% 0.98 (0.42, 2.30) 0.97 Kidney Creatinine >1.0 mg/dL 74 51% 72 49% – Creatinine ≤1.0 mg/dL 324 50% 327 50% 1.04 (0.72, 1.49) 0.84 Liver test 1 Total bilirubin >1.0 mg/dL 64 49% 67 51% – Total bilirubin ≤1.0 mg/dL 330 50% 328 50% 0.95 (0.65, 1.38) 0.79 Liver test 2 SGOT >36 l/L 220 53% 199 47% – SGOT <36 l/L 174 47% 198 53% 1.26 (0.95, 1.67) 0.11 Liver test 3 SGPT >27 l/L 243 53% 219 47% – SGPT <27 l/L 151 46% 178 54% 1.31 (0.99, 1.74) 0.06 Pancreas test 1 Lipase >50 l/L 124 51% 117 49% – Lipase ≤50 l/L 245 49% 252 51% 1.09 (0.80, 1.48) 0.58 Pancreas test 2 Amylase >120 l/L 109 58% 80 42% – Amylase ≤120 l/L 267 48% 295 52% 1.51 (1.08, 2.10) 0.016 1 Odds ratios, confidence intervals and p-values were calculated using univariate logistic regression. Dashes indicate baseline category. 2 Represents a total for the number of different medications administered to the donor less than 24 hrs prior to cross clamp. Odds ratio and 95% CI for continuous variables were reported between 75th and 25th percentiles. 3 Categories are not mutually exclusive and represent regrouping by the authors to reduce 295 medications into manageable classes based on drug mechanism of action (see Supporting Table 2). 4 Medications with broad action placed into this category. 5 Organs that were donated included kidney (left, right, or both), lung (left, right, or both), pancreas, heart, liver and intestine. (Table 5). Several univariate factors that improved the odds of obtaining high-yielding isolations also increased the odds when examined in a multivariable model, in- cluding elevated BMI (OR = 2.26, p < 0.0001), nor- mal liver SGPT levels (OR = 1.59, p = 0.020) and pan- creas fat infiltration (OR = 1.81, p = 0.003). Likewise, extended CIT (OR = 0.77, p = 0.016), the use of flu- ids/electrolyte medications (OR = 0.55, p = 0.032), pan- creata from elderly donors (OR = 0.57, p = 0.005) and pancreas edema (OR = 0.47, p = 0.011) were all found to decrease the odds of obtaining a high- yielding isolation. Use of an organ preservation solu- tion other than UW alone and TLM also decreased the odds of obtaining a high-yielding isolation in 650 American Journal of Transplantation 2010; 10: 646–656 24 Multicenter Analysis of Variables Associated with Islet Isolation Yield Table 3: Pancreas factors tested for contribution to islet isolation outcome ≤220000 IEQs ≥315000 IEQs % or Mean % or Mean Odds Name N (±1SD) N (±1 SD) ratio 1 95% CI p-Value Pancreas preservation solution used 0.0023 UW alone 221 46% 263 54% – Two-layer 100 49% 103 51% 0.87 (0.62, 1.20) 0.39 Histidine–trypophan–ketoglutarate 56 67% 28 33% 0.42 (0.26, 0.68) 0.0005 Pancreatic surface fat Clean or light 129 58% 94 42% – Moderate or heavy 211 44% 269 56% 1.75 (1.27, 2.42) 0.0006 Fatty infiltration None or patchy 190 56% 147 44% – Moderate or heavy 140 44% 176 56% 1.63 (1.19, 2.22) 0.002 Pre distention pancreas weight 3 (g) 402 95.5 (±31.5) 404 106.7 (±31.8) 1.55 (1.30, 1.85) <0.0001 Organ intact No 44 58% 32 42% – Yes 347 48% 370 52% 1.47 (0.91, 2.38) 0.12 Organ damage No 310 48% 331 52% – Yes 78 55% 65 45% 0.78 (0.54, 1.12) 0.18 Organ edema No 313 47% 352 53% – Yes 55 65% 30 35% 0.49 (0.30, 0.77) 0.003 Pancreas procurement team Remote 267 50% 268 50% – Local 110 48% 121 52% 1.10 (0.81, 1.50) 0.53 Cold ischemia time 2,3 (h) 374 8.7 (±4.8) 376 7.7 (±4.8) 0.82 (0.70, 0.95) 0.012 1 Odds ratios, confidence intervals and p-values were calculated using univariate logistic regression. Dashes indicate baseline category. 2 Odds ratio and 95% CI for continuous variables were reported between 75th and 25th percentiles. 3 Cold ischemia time was defined as the time from cross clamp to the start of dissection by the pancreas processing laboratory. the univariate and multivariable analyses (OR = 0.31, p = 0.0007). However, there were three variables that were found tobe significant in only one form of the analysis. First, an intact pancreas was not shown to be significant in the univariate analysis, but nearly doubled the odds of obtaining a high- yielding isolation in the multivariable model (OR = 1.95, p = 0.031). Second, amylase levels below 120 U/L was found to be an important factor in the univariate analysis, but failed to show any statistically significant influence in obtaininghighyieldsinthemultivariablemodel(OR =1.48, p = 0.082). Finally, center effect (see Supporting Table S3) was found to be significant in the univariate analysis (over- all p < 0.0001), for only four of the 13 isolation laborato- ries compared to the baseline facility, but failed to show anystatisticalsignificanceinthemultivariatemodel(overall p = 0.09). Discussion Inthisstudy,weanalyzeddatasetsmadeavailabletous throughtheICRandUNOS,includingasubsetoffactorsre- portedintheliteratureoverthepast15years.Ourobjective was to identify variables associated with a dichotomized outcome of islet isolation by using a unique and exten- sive collection of heterogeneous data generated by 14 laboratories. Severalfactorswereshowntoconsistentlyinfluencepost- purification isolation outcome in both the univariate and multivariable analysis. Our study is the first to show that when grouping donor medications by mechanism of ac- tion, fluid/electrolyte administration less than 24 h prior to cross-clamp negatively impacts isolation outcome. We did not find any collinear variables that might help to explain this observation, nor were there any statistically signifi- cant interaction terms. The use of hormonal medications likewise negatively influenced isolation outcome, although only univariately. In a related, but earlier study, Fiedor et al. found that the hemodynamic and metabolic status of the patient were variables significantly associated with islet yield, but not fluid intake (31). Nonetheless, it is possi- ble that certain combinations of drugs, beyond our broad groupings, have synergistic or antagonistic effects yet to be identified. Pancreata from older donors were found to negatively im- pactisolationoutcome.Thesefindingscomplementanear- lier report, where islet recovery and purity both improved when using pancreata from younger donors (8), but con- tradict later studies in which pancreata from older donors resulted in high-yielding isolations versus those from American Journal of Transplantation 2010; 10: 646–656 651 25 Kaddis et al. Table 4: Influence of pancreas processing factors on isolation outcome ≤220000 IEQs ≥315000 IEQs Variable name N %, Mean (±1SD), or median (min, max) 2 N %, Mean (±1SD), or median (min, max) 2 Odds ratio 1 95% CI p-Value Collagenase provider 0.10 Roche 272 48% 293 52% – Serva 88 57% 66 43% 0.70 (0.49, 1.00) 0.048 Sigma 21 57% 16 43% 0.71 (0.25, 1.38) 0.31 Predistention pancreas temperature 3 ( ◦ C) 41 2.1 (0.2, 15.2) 26 2.1 (0.4, 6.7) 0.71 (0.36, 1.25) 0.28 Postdistention pancreas temperature 3 ( ◦ C) 14 5.1 (3.0, 32.0) 12 4.0 (4.0, 14.4) 0.79 (0.32, 1.53) 0.51 Digestion time 3,4 (min) 397 19.3 (±10.4) 401 18.7 (±10.4) 0.95 (0.84, 1.07) 0.40 Dilution times 3,5 (min) 303 37.5 (±18.6) 307 39.9 (±19.6) 1.14 (0.97, 1.36) 0.12 Total 2 (digestion and dilution) (mins) 303 56.9 (±17.1) 307 59.1 (±18.9) 1.14 (0.96, 1.35) 0.14 Purification method 0.21 Continuous 340 49% 351 51% – Discontinuous 49 58% 36 42% 0.71 (0.45, 1.12) 0.14 Both 10 40% 15 60% 1.45 (0.65, 3.38) 0.37 Type of gradient used 0.22 Bicoll (Ficoll) 273 49% 288 51% – Euroficoll 33 56% 26 44% 0.75 (0.43 1.28) 0.29 Optiprep 69 57% 53 43% 0.73 (0.49, 1.08) 0.11 More than one of the above 11 37% 19 63% 1.64 (0.78, 3.62) 0.20 Other 13 45% 16 55% 1.17 (0.55, 2.51) 0.69 1 Odds ratios, confidence intervals and p-values were calculated using univariate logistic regression. Dashes indicate baseline category. 2 Number and % are reported for all categorical variables. Use of mean or median for continuous factors based on if variable was normally distributed. 3 Odds ratio and 95% CI for continuous variables were reported between 75th and 25th percentiles. 4 Also known as phase I or digestion phase. Begins when the collagenase is first circulated through the pancreas digestion chamber. 5 Also known as phase II or dilution phase. Begins with influx of albumin into pancreas digestion chamber to stop enzymatic action of collagenase. younger donors or at the extremes of life (6,7,9,10). It is hypothesized that the collagen composition of the ex- tracellular matrix of the pancreas changes with age, thus impacting collagenase affinity during the digestion phase of the isolation process (6,27). This may partially help to explain why no clear consensus has emerged as to ideal age boundaries for optimal donors. We also observed a detrimental impact of CIT on isola- tion yield, along with a progressive decline in the odds of success when pancreata were subject to prolonged cold storageperiods.Thisobservationisinagreementwithpre- vious studies in which extended CIT inversely correlated withorganquality,postisolationyieldandgraftfunction(6– 9). Furthermore, the type of solution used to preserve the pancreas duringtransportationwas also identifiedas a sig- nificant variable associated with isolation outcome. There is substantial controversy regarding pancreas preservation due to the lack of agreement in the literature and the costs associated with each method. In 2002, a series of sep- arate studies indicated that the use of TLM for pancreas preservationwassuperiortoUWaloneandimprovedposti- solation islet recovery (21,22,26). However, recent reports challenge these findings and indicate that there is no real benefit in using TLM preservation (20,24,51). In our study, we found that the TLM does not significantly change the odds of obtaining a high-yielding isolation when compared to UW alone. However, we did find a significant nega- tive correlation between isolation success and the use of HTK when compared to UW alone. Data available on pan- creaspreservationforisletisolationwithHTKisscarce,but Brandhorst et al. reported in 1995 that preservation with UW and HTK was equivalent for the purpose of islet isola- tion(19),afindingthatwaslaterreproducedbySalehietal. in a study analyzing data from 96 human isolations (52). Conversely, Stewart et al. recently reported a large mul- ticenter experience and concluded that pancreas preser- vation with HTK for solid organ pancreas transplant was associatedwithdecreasedgraftsurvivalandearlygraftloss (53),furthersupportingourobservationthatHTKpreserva- tion was inferior to both UW and TLM for islet isolation purposes. Pancreata from high BMI donors were found to positively influence isolation outcome. This observation is in univer- sal agreement with previous reports showing that isola- tions using pancreata from overweight and obese donors areconsistentlysuperiorinyield(7–11,13,18,35,54).More- over,wefoundthatanincreaseinbothdonorandpancreas weightpositivelyinfluencedisolationoutcomeunivariately. Interestingly, although pancreas weight is not known prior to recovery, others have established a direct relationship 652 American Journal of Transplantation 2010; 10: 646–656 26 Multicenter Analysis of Variables Associated with Islet Isolation Yield Table 5: Multivariable logistic regression model of factors influencing islet isolation success Variable N Odds ratio 1 95% CI p-Value BMI 2 526 2.26 (1.72, 3.01) <0.0001 Cold ischemia time 2 526 0.77 (0.62, 0.95) 0.016 Fluids/electrolytes received No 448 – Yes 78 0.55 (0.31, 0.94) 0.032 Pancreas function Amylase >120 l/L 128 – Amylase <120 l/L 398 1.48 (0.95, 2.31) 0.08 Liver function SGPT >27 l/L 319 – SGPT ≤27 l/L 207 1.59 (1.08, 2.35) 0.020 Age 0.1–47.5 270 – >47.5–70.8 256 0.57 (0.39, 0.84) 0.005 Organ edema No 456 – Yes 70 0.47 (0.26, 0.83) 0.011 Fatty infiltration None or patchy 271 – Moderate or heavy 255 1.81 (1.22, 2.69) 0.003 Organ intact No 58 – Yes 468 1.95 (1.07, 3.62) 0.031 Pancreas preservation solution used 0.003 UW alone 328 – Two-layer 143 0.94 (0.60, 1.46) 0.77 All other solutions 55 0.31 (0.15, 0.59) 0.0007 1 Oddsratios,confidenceintervalsandp-valueswerecalculatedusingmultivariablelogisticregression.Dashesindicatebaselinecategory. 2 Odds ratio and 95% CI for continuous variables were reported between 75th and 25th percentiles. between donor and pancreas weight, using it to estimate organ weight from donor information for pancreas selec- tion purposes (12). A number of related factors influenced postisolation out- come in the multivariable analysis, but not consistently shown to do so univariately. Of note, we found the use of intact pancreata and those containing moderate to heavy surface fat or with moderate to heavy fat infiltration in- creases the likelihood of isolation success, but the pres- ence of organ edema to negatively impact outcome. With the exception of organ edema, these observations are in agreement with previous studies (7,9,14,54). Pancreas edema has been shown to be associated with increased postpurification yield and purity, affecting outcome by en- hancing enzymatic action during the digestion phase (55); however, our data show that the presence of edema di- minished by half the probability of a favorable outcome, an observation supported by Mahler et al. who found that fibrotic and edematous pancreata result in inferior postisolation yields (56). This might be explained by the observed relationship between pancreas edema and sub- optimal preservation, as well as inadequate collagenase diffusion after intraductal distention in the edematous pan- creas (23). In addition, normal liver and pancreas func- tion tests were also found to positively influence isola- tion outcome. Although these results are in general agree- ment with available data (7,57), our study is the first to show that both normal liver and pancreas function tests are important in isolation outcome within a multivariable model. Despite being cited as relevant elsewhere, we were not abletoconfirmtherelationshiptooutcomeforseveralvari- ables of interest. Contrary to previous reports (6,9,18,58), we failed to see an effect on outcome for pancreata re- covered using a local versus remote procurement team. Furthermore, we did not find that variables relating to the donors medical history to be significant, although noted as important elsewhere (7,18) Moreover, it was striking that noneofthepancreas-processingvariableshadaneffecton outcome. In our study, the type of enzyme used, digestion period and purification methods did not modify the odds of obtaining a high-yielding isolation, despite being exten- sively described as being significant in other publications (9,10,14,18,27). Although manufacturer lot number was available for nearly all isolation records (n = 789 of 806), data on enzymatic activity (n = 0), collagenase (n = 77) and neutral-protease (n = 43) concentrations were lim- ited. In addition, beyond the standard certificate of analy- sis,thereisnoconsensusonbiochemicalassaysproposed to characterize the collagenase and neutral-protease prod- uct (59,60), thus limiting the collection and analysis of this kind of data in a multicenter setting. American Journal of Transplantation 2010; 10: 646–656 653 27 Kaddis et al. OurresultsindicatedthattheuseofDCDdonorsnegatively impacts isolation outcome, a noteworthy, but marginally significant, finding given that such pancreata have been successfully used for PIT (42). Closely related to DCD sta- tus, cause of death was found to be a determinant to iso- lation success, as donors who were declared brain-dead following traumatic injury were less likely to result in a successfulisolation.Braindeathischaracterizedbysevere hemodynamic, metabolic and immunological alterations that compromise adequate perfusion and oxygenation of abdominal organs. O’Gorman et al. described a similar as- sociationbetweencauseofdeath,especiallywithabdomi- nalinjury,andadecreasedscoreintheirscreeningprocess leading to a reduced possibility of a successful outcome (7).Wealsoobservedthatstatisticalsignificanceformech- anism of death closely mirrored cause of death. The main differenceintheanalysiswasthereferencegroupusedfor baseline comparison. Although the findings were comple- mentary, the definitions for each category within a variable were slightly different (see Supporting Table S1) and may help explain the difference in statistical significance. Center effect was not a significant variable in the multi- variable model. Although we considered analyzing other factors that might account for center effect (such as the number of years in operation, years of staff/technician ex- perience and turnover rate), such data was not available to us for examination and precluded a more extensive analysis. Postpurification IEQ count was selected as the outcome because it was documented in nearly all records analyzed and has been regarded as the most important criterion for transplant suitability, along with preparation purity and via- bility. However, islet quantity does not necessarily reflect preparation quality, limiting the significance of our data in the transplant scenario. Additional research is needed to linkdonorandorgancharacteristicstoinvivoisletfunction, andavoidconfoundingvariablesrelatedtotheisolationpro- cedure itself. Therefore, until better quality indicators are tested, validated, and commonly reported, islet yield re- mains the single best and most widely available metric to determine isolation outcome. In summary, this multicenter analysis reveals that nonede- matousintactpancreatawithmoderatetoheavyfatinfiltra- tion from younger, obese, hemodynamically stable donors with normal liver function increase the odds of achiev- ing postpurification islet isolation success. Minimization of CIT and the use of UW solution or the TLM, but not HTK, for pancreas preservation also improve the odds of isolation success. The data derived from this study serves as a reference, along with established methodolo- gies (61) and regulatory guidelines (62–64), to help screen candidate donors, potentially lowering overall pancreas processing costs, maximizing efficient use of limited re- sources, and improving islet extraction and transplantation rates which may ultimately lead to superior posttransplant outcomes. Acknowledgments Funding source: This work was funded by the cooperative efforts of the NCRR and the NIDDK, a component of the US NIH, in conjunction with the generous contributions of the JDRF. ICR centers, past and present, include: (1) Washington University, St. Louis, MO (2001–2006; U42 RR 016597 to T. Mohanakumar), (2) University of Colorado, Denver CO (2001– 2006; U42 RR 016599 to R.G. Gill), (3) University of Tennessee, Nashville, TN (2001–2006; U42 RR 016602 to A.O. Gaber), (4) Puget Sound Blood Center, Seattle, WA (2001–2006; U42 RR 016604 to J. Reems), (5) Joslin Diabetes Center, Boston, MA (2001–2006; U42 RR 016606 to G.C. Weir), (6) Columbia University, New York, NY (2001–2006; U42 RR 016629 to M.A. Hardy), (7) University of Minnesota, Minneapolis, MN (2001–2009; U42 RR 016598 to B.J. Hering), (8) University of Pennsylvania, Philadelphia, PA (2001–2009; U42 RR 016600 to A. Naji), (9) University of Miami, Miami, FL (2001–2009; U42 RR 016603 to C. Ricordi), (10) City of Hope National Medical Center, Duarte, CA (2001–2009; U42 RR 016607 to F. Kandeel), (11) University of Wisconsin, Madison, WI (2006–2009; U42 RR 023240 to L.A. Fernandez), (12) Chicago Consortium (University of Illinois at Chicago andNorthwesternUniversity),Chicago,IL(2006–2009;U42RR023245toJ. Oberholzer,includingsubcontracttoNorthwestern,sub-PI,D.Kaufman)and (13)UniversityofAlabama,Birmingham,AL(2006–2009;U42RR023246to J.Contreras).TheICR-ABCCislocatedattheCityofHopeNationalMedical Center (2001–2009; U42 RR 017673 to J.C. Niland). Data: Organ donor information was supplied by the United Network of Organ Sharing as the contractor for the Organ Procurements and Trans- plantation Network (OPTN). The interpretation and reporting of such data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by the OPTN or the U.S. Govern- ment. Pancreas and islet isolation data was acquired though the ICR-ABCC database and entered by staff members from 14 contributing ICR facilities (see funding sources mentioned earlier for the facility name, years in ser- vice, grant funding and principal investigator name of each participating ICR laboratory). Other contributions: The authors are equally grateful to the scientists and techniciansineachisolationlaboratorywhohavespentmanylonghoursen- teringdataintotheICR-ABCCdatabase.AspecialthankstoJamesCravens in providing assistance and input with some of the technical aspects of the data analysis and to both Dr. Alejandro Munoz and the journal reviewers for careful evaluation of the manuscript. A special acknowledgment to the families of organ donors and Organ Procurement Organizations across the country without whom this work would not have been possible. References 1. Shapiro AM, Ricordi C, Hering BJ et al. International trial of the Edmonton protocol for islet transplantation. N Engl J Med 2006; 355: 1318–1330. 2. Alejandro R, Barton FB, Hering BJ, Wease S. 2008 Update from the Collaborative Islet Transplant Registry. Transplantation 2008; 86: 1783–1788. 3. Ricordi C, Inverardi L, Kenyon NS, Goss J, Bertuzzi F, Alejandro R.Requirementsforsuccessinclinicalislettransplantation.Trans- plantation 2005; 79: 1298–1300. 4. Robertson RP. Islet transplantation as a treatment for diabetes—a work in progress. N Engl J Med 2004; 350: 694–705. 654 American Journal of Transplantation 2010; 10: 646–656 28 Multicenter Analysis of Variables Associated with Islet Isolation Yield 5. Fiorina P, Shapiro AM, Ricordi C, Secchi A. The clinical impact of islet transplantation. Am J Transplant 2008; 8: 1990–1997. 6. Lakey JR, Warnock GL, Rajotte RV et al. Variables in organ donors that affect the recovery of human islets of Langerhans. Transplan- tation 1996; 61: 1047–1053. 7. O’GormanD,KinT,MurdochTetal.Thestandardizationofpancre- atic donors for islet isolations. Transplantation 2005; 80: 801–806. 8. Zeng Y, Torre MA, Karrison T, Thistlethwaite JR. The correlation between donor characteristics and the success of human islet isolation. Transplantation 1994; 57: 954–958. 9. Ponte GM, Pileggi A, Messinger S et al. Toward maximizing the success rates of human islet isolation: Influence of donor and isolation factors. Cell Transplantation 2007; 16: 595–607. 10. Hanley SC, Paraskevas S, Rosenberg L. Donor and isolation vari- ables predicting human islet isolation success. Transplantation 2008; 85: 950–955. 11. Brandhorst H, Brandhorst D, Hering BJ, Federlin K, Bretzel RG. Bodymassindexofpancreaticdonors:Adecisivefactorforhuman islet isolation. Exp Clin Endocrinol Diabetes 1995; 103(Suppl 2): 23–26. 12. Kin T, MurdochTB, Shapiro AM,Lakey JR.Estimation ofpancreas weight from donor variables. Cell Transplant 2006; 15: 181–185. 13. MatsumotoI,SawadaT,NakanoMetal.Improvementinisletyield from obese donors for human islet transplants. Transplantation 2004; 78: 880–885. 14. NanoR,ClissiB,MelziRetal.Isletisolationforallotransplantation: Variables associated with successful islet yield and graft function. Diabetologia 2005; 48: 906–912. 15. Neidlinger NA, Odorico JS, Sollinger HW, Fernandez LA. Can ‘ex- treme’ pancreas donors expand the donor pool? Curr Opin Organ Transplant 2008; 13: 67–71. 16. Markmann JF, Deng S, Desai NM et al. The use of non-heart- beating donors for isolated pancreatic islet transplantation. Trans- plantation 2003; 75: 1423–1429. 17. Matsumoto S, Okitsu T, Iwanaga Y et al. Successful islet trans- plantation from nonheartbeating donor pancreata using modified Ricordi islet isolation method. Transplantation 2006; 82: 460–465. 18. Benhamou PY, Watt PC, Mullen Y et al. Human islet isolation in 104 consecutive cases. Factors affecting isolation success. Trans- plantation 1994; 57: 1804–1810. 19. Brandhorst H, Hering BJ, Brandhorst D et al. Comparison of histidine-tryptophane-ketoglutarate (HTK) and University of Wis- consin (UW) solution for pancreas perfusion prior to islet isola- tion, culture and transplantation. Transplant Proc 1995; 27: 3175– 3176. 20. Caballero-Corbalan J, Eich T, Lundgren T et al. No beneficial effect of two-layer storage compared with UW-storage on human islet isolation and transplantation. Transplantation 2007; 84: 864–869. 21. Fraker CA, Alejandro R, Ricordi C. Use of oxygenated perfluorocar- bon toward making every pancreas count. Transplantation 2002; 74: 1811–1812. 22. Hering BJ, Matsumoto I, Sawada T et al. Impact of two-layer pan- creaspreservationonisletisolationandtransplantation.Transplan- tation 2002; 74: 1813–1816. 23. Hubert T, Gmyr V, Arnalsteen L et al. Influence of preservation solution on human islet isolation outcome. Transplantation 2007; 83: 270–276. 24. Kin T, Mirbolooki M, Salehi P et al. Islet isolation and transplanta- tionoutcomesofpancreaspreservedwithUniversityofWisconsin solution versus two-layer method using preoxygenated perfluoro- carbon. Transplantation 2006; 82: 1286–1290. 25. Lakey JR, Tsujimura T, Shapiro AM, Kuroda Y. Preservation of the human pancreas before islet isolation using a two-layer (UW solution-perfluorochemical) cold storage method. Transplantation 2002; 74: 1809–1811. 26. Tsujimura T, Kuroda Y, Kin T et al. Human islet transplantation from pancreases with prolonged cold ischemia using additional preservation by the two-layer (UW solution/perfluorochemical) cold-storage method. Transplantation 2002; 74: 1687–1691. 27. Sabek OM, Cowan P, Fraga DW, Gaber AO. The effect of donor factors on human islet yield and their in vivo function. Prog Trans- plant 2006; 16: 350–354. 28. United Network for Organ Sharing (UNOS) Website. Policy 3.8 Organ Distribution: Pancreas Allocation. Available at http:// www.unos.org/PoliciesandBylaws2/policies/pdfs/policy_10.pdf. Accessed February 28, 2009. 29. Balamurugan AN, Chang Y, Bertera S et al. Suitability of human juvenile pancreatic islets for clinical use. Diabetologia 2006; 49: 1845–1854. 30. Brandhorst H, Brandhorst D, Hesse F et al. Successful human islet isolation utilizing recombinant collagenase. Diabetes 2003; 52: 1143–1146. 31. FiedorP,GoodmanER,SungRS,CzerwinskiJ,RowinskiW,Hardy MA. The effect of clinical and biochemical donor parameters on pancreatic islet isolation yield from cadaveric organ donors. Ann Transplant 1996; 1: 59–62. 32. Hubert T, Strecker G, Gmyr V et al. Acute insulin response to arginine in deceased donors predicts the outcome of human islet isolation. Am J Transplant 2008; 8: 872–876. 33. Ichii H, Pileggi A, Molano RD et al. Rescue purification maximizes theuseofhumanisletpreparationsfortransplantation.AmJTrans- plant 2005; 5: 21–30. 34. Ichii H, Wang X, Messinger S et al. Improved human islet isolation using nicotinamide. Am J Transplant 2006; 6: 2060–2068. 35. Ihm SH, Matsumoto I, Sawada T et al. Effect of donor age on function of isolated human islets. Diabetes 2006; 55: 1361–1368. 36. KinT,SeniorP,O’Gorman D,RicherB,Salam A,ShapiroAM.Risk factors for islet loss during culture prior to transplantation. Transpl Int 2008; 21: 1029–1035. 37. Kin T, Zhai X, Murdoch TB, Salam A, Shapiro AMJ, Lakey JRT. Enhancing the success of human islet isolation through optimiza- tion and characterization of pancreas dissociation enzyme. Am J Transplant 2007; 7: 1233–1241. 38. Kinasiewicz A, Juszczak M, Pachecka J, Fiedor P. Pancreatic islets isolation using different protocols with in situ flushing and intra- ductal collagenase injection. Physiol Res 2004; 53: 327–333. 39. Lakey JR, Rajotte RV, Warnock GL, Kneteman NM. Human pan- creas preservation prior to islet isolation. Cold ischemic tolerance. Transplantation 1995; 59: 689–694. 40. LinetskyE,BottinoR,LehmannR,AlejandroR,InverardiL,Ricordi C. Improved human islet isolation using a new enzyme blend, liberase. Diabetes 1997; 46: 1120–1123. 41. LiuX,MatsumotoS,OkitsuTetal.Analysisofdonor-andisolation- relatedvariablesfromnon-heart-beatingdonors(NHBDs)usingthe Kyoto islet isolation method. Cell Transplant 2008; 17: 649–656. 42. Markmann JF, Deng SP, Desai NM et al. The use of non-heart- beating donors for isolated pancreatic islet transplantation. Trans- plantation 2003; 75: 1423–1429. 43. Ricordi C, Fraker C, Szust J et al. Improved human islet isolation outcome from marginal donors following addition of oxygenated perfluorocarbontothecold-storagesolution.Transplantation2003; 75: 1524–1527. 44. Rose NL, Palcic MM, Shapiro AMJ, Lakey JRT. Endogenous pancreatic enzyme activity levels show no significant effect on human islet isolation yield. Cell Transplant 2004; 13: 153– 160. American Journal of Transplantation 2010; 10: 646–656 655 29 Kaddis et al. 45. TsujimuraT,KurodaY,AvilaJGetal.Influenceofpancreaspreser- vation on human islet isolation outcomes: Impact of the two-layer method. Transplantation 2004; 78: 96–100. 46. TsujimuraT,KurodaY,ChurchillTAetal.Short-termstorageofthe ischemically damaged human pancreas by the two-layer method prior to islet isolation. Cell Transplant 2004; 13: 67–73. 47. Yamamoto T, Ricordi C, Messinger S et al. Deterioration and vari- ability of highly purified collagenase blends used in clinical islet isolation. Transplantation 2007; 84: 997–1002. 48. Kissler HJ, Niland JC, Olack B et al. Validation of methodologies forquantifyingisolatedhumanislets:Anisletcellresourcesstudy. Clin Transplant 2009. Epub ahead of print. 49. Hosmer DW, Hosmer T, Cessie SL, Lemeshow S. A comparison ofgoodness-of-fittestsforthelogisticregressionmodel.StatMed 1997; 16: 965–980. 50. Hosmer DW, Lemeshow S. Applied logistic regression. 2nd Ed. New York: Wiley, 2000. 51. AgrawalA,GurusamyK,PowisS,GrayDW,FullerB,DavidsonBR. A meta-analysis of the impact of the two-layer method of preser- vation on human pancreatic islet transplantation. Cell Transplant 2008; 17: 1315–1322. 52. Salehi P, Hansen MA, Avila JG et al. Human islet isolation out- comes from pancreata preserved with Histidine-Tryptophan Ke- toglutarate versus University of Wisconsin solution. Transplanta- tion 2006; 82: 983–985. 53. Stewart ZA, Cameron AM, Singer AL, Dagher NN, Montgomery RA, Segev DL. Histidine-tryptophan-ketoglutarate (HTK) is associ- ated with reduced graft survival in pancreas transplantation. Am J Transplant 2009; 9: 217–221. 54. Hughes SJ, McShane P, Contractor HH, Gray DW, Clark A, John- son PR. Comparison of the collagen VI content within the islet- exocrineinterfaceofthehead,body,andtailregionsofthehuman pancreas. Transplant Proc 2005; 37: 3444–3445. 55. Taylor MJ, Baicu S, Leman B, Greene E, Vazquez A, Brassil J. Twenty-four hour hypothermic machine perfusion preservation of porcine pancreas facilitates processing for islet isolation. Trans- plant Proc 2008; 40: 480–482. 56. Mahler R, Franke FE, Hering BJ et al. Evidence for a significant correlationofdonorpancreasmorphologyandtheyieldofisolated purified human islets. J Mol Med 1999; 77: 87–89. 57. LakeyJR,RicordiC,HeringBJ.UpdateandnewfeaturesincGMP facility, equipment and structure of the Clinical Islet Transplant Program. In:Human Islet Isolation and Transplantation Techniques (HIITT) 6th Training Workshop, April 9-12, 2006; Snowbird, Utah. 58. Lee TC, Barshes NR, Brunicardi FC et al. Procurement of the hu- man pancreas for pancreatic islet transplantation. Transplantation 2004; 78: 481–483. 59. Barnett MJ, Zhai X, LeGatt DF, Cheng SB, Shapiro AM, Lakey JR. Quantitative assessment of collagenase blends for human islet isolation. Transplantation 2005; 80: 723–728. 60. Bertuzzi F, Cainarca S, Marzorati S et al. Collagenase isoforms for pancreas digestion. Cell Transplant 2009; 18: 203–206. 61. Pileggi A, Ricordi C, Kenyon NS et al. Twenty years of clinical islet transplantation at the Diabetes Research Institute—University of Miami. Clin Transpl 2004: 177–204. 62. Zoon KC. Dear colleague letter to transplant centers: Allogeneic pancreatic islets for transplantation. Washington, DC: US Depart- ment of Health and Human Services, 2000. 63. Weber DJ. FDA regulation of allogeneic islets as a biological prod- uct. Cell Biochem Biophys 2004; 40(3 Suppl): 19–22. 64. Weber DJ, McFarland RD, Irony I. Selected Food and Drug Admin- istration review issues for regulation of allogeneic islets of langer- hansassomaticcelltherapy.Transplantation2002;74:1816–1820. Supporting Information The following supporting information is available for this article online: Figure S1: Literature search strategy to identify rele- vant articles containing analysis of human islet isola- tion outcomes. Table S1: Definitions for variables used in the analysis Table S2: Drugs given to organ donors grouped by class Table S3: Univariate analysis of center effect across 14 human islet isolation laboratories Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials sup- plied by the authors. Any queries (other than missing ma- terial) should be directed to the corresponding author for the article. 656 American Journal of Transplantation 2010; 10: 646–656 30 CHAPTER 3: PREDICTING HUMAN ISLET ISOLATION AND TRANSPLANTATION OUTCOMES IN TYPE 1 DIABETES Section 2.1: Optimization of Microarray Analysis Techniques: Influence of RNA Labeling on High-Throughput Expression Profiling of MicroRNAs CHAPTER 3.2.1 ABSTRACT Background: Although a number of technical parameters are now being examined to optimize microRNA profiling experiments, it is unknown if changes made to the ligation mix used during the labeling step affect the starting RNA requirements or microarray performance. Methodology: Human brain/lung samples were each labeled in duplicate, at 1.0ug, 0.5ug, 0.2ug, and 0.1ug of total RNA, using both a standard and ligation- mix modified RNA labeling kit. Statistical measures of reliability and validity were used to evaluate microarray data. Cross-platform confirmation was accomplished using TaqMan microRNA assays. Synthetic microRNA spike-in experiments were also performed to establish the signal dynamic range of the microarray using the ligation-modified kit. 31 Results: Technical replicate correlations of signal intensity values were high using both kits, but improved with the ligation-modified assay. The drop in detection call sensitivity and gene list correlations observed using reduced amounts of standard-labeled RNA was considerably improved with the ligation- modified kit. Microarray signal dynamic range was found to be linear across 3 orders of magnitude from 4.88 to 5000 attomoles. Significance: Our results show that optimization of the microRNA labeling step can result in at least a 10-fold decrease in microarray total RNA requirements with little compromise to data quality. Clinical investigations bottlenecked by the amount of starting material may employ a ligation mix modification strategy to reduce total RNA requirements. CHAPTER 3.2.1 INTRODUCTION MicroRNA (miRNA) expression profiling platforms have burgeoned in the last decade 1-4 , with a promising potential to aid in the diagnosis and prognosis of human diseases such as cancer 5 . In response, intra-, inter- and cross-platform comparisons have recently emerged that examine the use of microarray technology alone 6,7 or in contrast to other profiling methods 8-12 . Such studies have reported on the reliability and validity of widely used miRNA discovery tools and to highlight the advantages and limitations of these platforms. 32 However, there also remains a need to evaluate pre-experimental factors, such as sample handling, processing, storage, and nucleic acid quality, that may influence miRNA expression profiling experiments 13 . While 2 studies have demonstrated that miRNA detection was possible in formalin-fixed paraffin- embedded (FFPE) samples of up to 10 14 and 12 15 years old, stability and comparability of miRNA expression signal using paired fresh frozen (FF) versus FFPE samples was reported in human and mouse specimens using different tissues ranging in age from newly acquired 16 up to 3 years old 15,17 . In contrast, when detecting the expression of longer RNA transcripts, Abdueva et al found differences in paired FF versus FFPE samples of up to 7.5 years of age, but concluded that useable profiles were obtainable 18 . A number of studies examined RNA extraction methods for miRNA, with 1 that compared 3 isolation kits and found general comparability between them 19 , 2 that used 5 kits and identified a preferred approach 20,21 , and 1 that evaluated 8 kits and found R 2 correlations exceeding 0.857 in all comparisons made 22 . Conflicting reports have surfaced on the length of time RNA can be stored after extraction for miRNA profiling, with one study describing degradation after 3 days 23 and another demonstrating stability of up to 10 months 24 , although cells from different species were used in each study. Likewise, the quality of RNA 25 has also been called into question, with some showing that medium to high integrity samples are required for reproducible miRNA profiling using bovine 26 and mouse 27 samples, and another reporting negligible effects on highly degraded human samples 28 . 33 Consideration of these experimental factors are crucial to ensuring valid and reliable miRNA expression profiling results within and across laboratories. With the exception of next generation sequencing (NGS), all miRNA expression profiling approaches use a labeling method to tag the target molecules of interest. In the case of transcripts > 200 nucleotides, the method used to amplify and/or label the molecules of interest prior to detection has been shown to be important. Studies examining linear 29-33 , exponential 34 , and linear versus exponential 35,36 amplification methods revealed that while total RNA requirements can be reduced significantly by using amplified RNA, quality metrics such as comparability, sensitivity, and accuracy were in some cases compromised when compared to un-amplified RNA. However, these measures can be improved when robust data methods were used 37 . Next, evaluation of direct, indirect, and other cDNA labeling approaches have shown that the method chosen can reduce total RNA requirements, but that gains in signal intensity values were also achieved alongside, in some cases, a loss in labeling performance 38-43 . Similar observations have been made when comparing cRNA labeling methods 44-47 . Even when performing replicate experiments within 48 or across 49 laboratories, RNA labeling has been shown to contribute to the variability seen in gene expression outcome. 34 While the body of work examining target labeling is extensive, it is unclear if and how these findings can be applied to low abundance short transcripts < 200 nucleotides. This is especially true in the case of miRNAs, when considering that they represent only 0.5-9.2% of total RNA in human samples and 0.1-1.3% in mouse and rat samples 50 . Since many studies are limited in the amount of total RNA available, such as those utilizing tissues derived from needle biopsies or laser capture microdissection, more sensitive target labeling procedures are critical. Accordingly, a growing number of amplification and labeling strategies have been developed to aid in the detection of these low abundance RNAs 51 . While there are at least 2 studies that have examined the effects of amplification on miRNA expression 8,52 , there are currently no studies that compare amplification or labeling effects when direct labeling of mature miRNA is employed. We therefore conducted a study to directly compare the Genisphere® FlashTag™ biotin-HSR labeling kit to the previously recommended FlashTag™ biotin-only assay for the Affymetrix® GeneChip® miRNA Arrays. The biotin-HSR kit is a ligation-mix modified version of the biotin-only assay. A total of 32 chips were used for this evaluation, followed by cross-platform validation using TaqMan miRNA assays. In addition, the linear dynamic range of the GeneChip was assessed using synthetic miRNA spiked-in at different amounts, labeled with the biotin-HSR kit, on 12 chips. A total of 44 microarray chips were used in this study. 35 CHAPTER 3.2.1 MATERIALS AND METHODS Human Normal Tissue RNA. FirstChoice® Total Brain (Lot#0906005) and Lung (Lot#0904002) RNA samples were obtained from Ambion, Inc., certified to contain small RNAs, and quantified at 1mg/mL (Applied Biosystems/Ambion, Foster City, CA). Labeling of RNA and Microarray Processing. Brain and lung total RNA samples were each labeled using both the FlashTag biotin-only and FlashTag biotin-HSR RNA labeling kits (Genisphere, Hatfield, PA) for the Affymetrix® GeneChip® miRNA array. Each microarray contains sequences for 6703 miRNA probes, 499 small nucleolar RNA (snoRNA), including 274 C/D and 127 H/ACA box RNA, 22 small cajal body-specific RNA (scaRNA), 10 5.8s ribosomal RNA , 162 Affymetrix controls, and 22 oligonucleotide spike-in controls. Probes were developed by Affymetrix using Sanger miRNA database v11 and contained additional snoRNAs and scaRNAs. RNA labeling was performed by the kit manufacturer and efficiency assessed using an Enzyme Linked Oligosorbent Assay (ELOSA). Hybridization, washing, and scanning of slides were performed according to the Affymetrix and Genisphere protocols. Scanning was performed using the miRNA-1_0 library file from Affymetrix for the Flashtag biotin-only and miRNA- 1_0_2Xgain for the FlashTag biotin-HSR labeled samples. 36 RNA Titration Experiments. Total RNA was prepared at recommended and reduced input amounts from two different tissues to examine the performance of the Affymetrix GeneChip miRNA array and evaluate the significance of using a new RNA labeling kit on the quality of data generated. Brain samples were prepared containing 0.1ug, 0.2ug, 0.5ug, and 1.0ug of total RNA. Two aliquots from each sample were taken and labeled with either the biotin-only or biotin- HSR kit. This process was carried out in duplicate for each tissue. A total of 32 microarray chips were used for this study. Synthetic miRNA Spike-In Experiments. The mirVana miRNA Reference Panel v9.1 (Lot#072307) was obtained from Ambion, Inc. and contains an equimolar pool of 470 human, 224 mouse, and 42 rat synthetic miRNA oligonucleotides according to the Sanger miRBase sequence database, release 9.1 (Applied Biosystems/Ambion, Foster City, CA). Two-fold serial dilution samples were then prepared that contained from 5000 to 4.88 attomoles of miRNA (12 total, including a negative control). Synthetic miRNAs were labeled using the FlashTag biotin-HSR RNA labeling kits. Hybridization, washing, and scanning of slides were performed by Novartis according to Affymetrix and Genisphere protocols. Scanning was performed using the miRNA-1_0_2Xgain library file from Affymetrix. 37 Array Quality, Data Processing, and Detection Calls. Raw intensity signal values of Affymetrix spike-in controls indicated that array hybridization was successful (i.e. bioB<bioC<bioD<Cre). Likewise, raw intensity values from an additional 5 Genisphere oligonucleotide spike-ins control probes were all found to be >1000, used to demonstrate poly(A) tailing and ligation (3 RNA oligos), ligation (1 poly(A) RNA), and ligation and lack of RNases in the RNA sample (1 poly(dA) DNA; data not shown). Because labeling was an experimental factor being tested, raw signal intensity values were Robust Multichip Average (RMA) background corrected, quantile normalized, median polish summarized, and log 2 transformed separately for FlashTag biotin-only and FlashTag biotin-HSR chip data 53-57 . Processing was performed using Partek Genomics Suite, version 6.5, build 6.10.0412 copyright 2010 (Partek Inc, St. Louis, MO USA). Present/absent detection calls for each probe, on every array, were generated using the Affymetrix detection algorithm and statistical significance testing implemented through the miRNA QCTool, version 1.0.33.0. Details on how the algorithm determines these calls are described in Appendix A of the Affymetrix miRNA QC Tool User’s Guide 6 that, along with the software utility, are freely available on the manufacturer’s website. Reverse Transcription and Real Time PCR. Total RNA was converted into cDNA and quantified using individual Taqman assay kits for each miRNA of interest (Applied Biosystems, Foster City, CA). Real time PCR followed the reverse 38 transcription (RT) reaction, which was modified to allow for the simultaneous synthesis of cDNA by combining RT primers into pools A and B. The multiplex RT protocol is a modification of the original kit instructions and made available through the manufacturer by request. RNU6B, SNORD44, and SNORD48 were selected as candidate artificial normalization controls and previously shown to be stable in a panel of 38 normal human tissues, including Brain and Lung 58 . Identification of hsa-let-7d and hsa- miR-151-5p as candidate endogenous human miRNA normalization controls, specific to the present experiment, was performed, using microarray data, in the following way: 1) A signal detection call of ‘present’ in both Brain and Lung tissue across all input RNA titrations and replicates using both the old and new labeling methods (79 of 847 miRNA probesets remained), 2) The absence of a statistically significant fold change difference between Brain and Lung tissue (20 of 79 remained), 3) Removal of half the probesets with the lowest mean signal intensity values and those previously shown 59 as inadequate (8 of 20 remained), and 4) Utilization of NormFinder 60 to rank (data not shown) and select the most stable miRNAs (2 of 8 remained). Subsequent analysis of all 5 candidate control Ct values by geNorm 61 identified the combination of no greater than 2 miRNAs, i.e. hsa-miR-151-5p/SNORD44 and hsa-let-7d/SNORD44, as the most stably expressed housekeeping miRNAs 39 in pools A and B, respectively (data not shown). The expression of miRNAs, relative to selected controls, was determined using the 2 (– Ct) method 62,63 . Data Analysis. The mean or median was used as a measure of central tendency for all continuous numerical variables, depending on whether the values followed a normal Gaussian probability distribution. A student’s t-test was used when performing 2-group comparisons using parametric data. An F-test for equality of variance was performed prior to significance testing. A Wilcoxon rank sum test was used when performing 2-group comparisons using non-parametric data. For descriptive statistics, significance was indicated if p<0.05. Coefficient of variation (CV) for replicate chips was calculated as the standard deviation of the processed signal intensity value over the mean for miRNAs called as present. The Pearson Product Moment Correlation Coefficient (r) was used to measure the linear relationship of signal intensity values between replicate chips. Spearman’s rank correlation coefficient ( ρ) was used to evaluate intra- and cross- platform gene list comparisons. Unlike r, ρ measures the relative rank of a gene in list A against its relative rank in list B, thereby minimizing the effects of outlier data. For both r and ρ, the null hypothesis of no relationship was indicated if p>0.05. Statistical significance for differential gene expression was determined using a false discover rate (FDR) p-value of <0.05, unless otherwise noted. Data analysis was performed using either Partek Genomics Suite or SAS software, version 9.1.3 SP 4 (SAS Institute, Cary, NC). 40 CHAPTER 3.2.1 RESULTS Reliability (RNA Titrations). Qualitative and quantitative measures of random error within and between labeling kits were examined at all input amounts of total RNA from both brain and lung tissue. The number of present and absent detection calls was first calculated for all replicate chips (Supplemental Figure 1). Percent concordant (present and absent) calls for all probes on the array using total RNA labeled with the biotin-only kit was found to range from a grand mean of 95.0% for brain to 95.8% for lung. Those values increased to 96.6% and 97.4%, respectively, when the RNA was labeled using the biotin-HSR kit, although this was not a statistically significant finding (p>0.05 for both comparisons made). Upon restricted examination of only the 847 human miRNA gene probes, a statistically significant increase in the number of detection calls was seen when concurrently examining the differences in labeling kits alongside the amount of input RNA used for microarray hybridization (Figure 1). For brain samples, as the amount of biotin-only labeled RNA is reduced from 1ug to 0.1ug, the mean percentage of detected overlapping probes was decreased by 31%. However, this was statistically significantly improved to a mean value of only 7% when using biotin-HSR labeled RNA (p=0.0490). Likewise, for lung samples, the loss was statistically significantly improved from a mean value of 25% to 5% (p=0.04). 41 Irrespective of detection call, the reproducibility of processed signal intensity values between duplicate chips was next examined (Table 1). A statistically significant improvement in the mean Pearson’s correlation coefficient value was observed for duplicate chips labeled using the biotin-HSR kit. This finding was true for RNA isolated from either brain or lung tissue (p<0.0001 for both comparisons made). Individual correlations are provided in a 16 x 16 correlation matrix table for both brain and lung samples (Supplemental Tables 1-2). Finally, the CVs of the processed signal intensity values were calculated for all study samples (Figure 2). There was a statistically significant reduction in the grand median CV value for brain samples labeled with the biotin-HSR vs. biotin- only kit (2.59% to 1.25%; Wilcoxon rank sum p<0.05). However, a statistically significant reduction in the grand median CV value for lung samples was not observed (1.96% to 1.16%; Wilcoxon rank sum p>0.05). Validity (RNA Titrations and RT Real Time PCR). Systematic error was evaluated by examining the inter-platform agreement and cross-platform concordance of brain to lung expression ratios for selected human miRNAs. First, using differential gene expression analysis, the numerical rank of all probes from the 1.0ug biotin-only labeled RNA list was compared to the order from lists using reduced amounts of input RNA (Table 2). This process was repeated following sequential application of filters, defined in Table 2, to the 1.0ug biotin- 42 only gene list to selectively reduce the number of candidate probes evaluated. Regardless of gene filters used, as the amount of biotin-only labeled RNA is reduced, correlations with the 1.0ug biotin-only gene list also falls. However, when using reduced amounts of biotin-HSR labeled RNA, gene-list correlations remain steady. Depending on the gene list filter used, there is a 60% to 93% reduction in the variation of correlation values when using reduced amounts of biotin-HSR vs. biotin-only labeled RNA. Second, differentially-expressed gene lists were also generated for biotin-HSR labeled samples at each input RNA amount and evaluated for overlap in statistically significant transcripts (Figure 3). Of the 1779 human probes, a total of 474 differentially-expressed genes were identified. As the amount of input RNA increases, the percentage of genes unique to each list also increases from 4.3% for 0.1ug and 4.1% for 0.2ug to 8.8% for 0.5ug and 16.9% for 1.0ug. The percentage of genes that were detected in more than 1 list, but not by all, is 22.8% for 0.1ug, 34.4% for 0.2ug, 55.6% for 0.5ug, and 68.0% for 1.0ug. There were 134 genes robustly detected by all lists with a minimum absolute fold change of at least 10 or greater for 86 genes, and 2 to <10 for the remaining 48. Next, we took the 272 genes, from table 2, and compared it with the 474 identified genes from figure 3, to determine the gene-list effects of using a new labeling kit (Supplemental Table 3). Of f the 272 genes identified using 1.0ug of 43 biotin-only labeled RNA, 136 of them were also found to be statistically significant in one or more, but not all, biotin-HSR gene lists, 122 were common to all biotin- HSR gene lists, and 14 that were unique to the biotin-only gene list. To further investigate the differences in these gene lists, a cross platform analysis was undertaken comparing microarray results to TaqMan MiRNA assays for 11 selected human miRNAs (Figure 4). Brain to lung expression ratios derived from microarrays using the biotin-HSR kit better correlated with TaqMan assay results than did data from chips using biotin-only labeled RNA. Pearson’s correlation coefficient improved from 0.915 to 0.986 for 0.1ug, 0.964 to 0.995 for 0.2ug, 0.975 to 0.994 for 0.5ug, and 0.994 to 0.996 for 1.0ug samples. The use of expression ratios to correlate microarray with TaqMan data has been previously demonstrated 64 . Linear Dynamic Range (Synthetic miRNA Spike-In Experiments). The concentration range in which valid brain and lung sample measurements can be made was examined by using spike-in synthetic miRNAs at increasing equimolar amounts (Figure 5). The change in amount of spike-in used on the array varied by 3 orders of magnitude from a low of 4.88 to a high of 5000 attomoles (0 attomole sample used as control). The change in signal intensity values across this range was found to be linear for human (r=0.991), mouse (r=0.989), and rat (r=0.987) synthetic spike-in miRNAs. Signal intensity values from brain and lung 44 samples of the same probes used to detect the synthetic human miRNAs were also plotted on the same graph and shown to fall within the linear dynamic range tested. CHAPTER 3.2.1 DISCUSSION Previous studies in the mRNA literature have shown that when laboratories are left to use their own labeling protocols, the comparability of microarray data within and across centers is fair to low 65 , but can be improved when standardizing technical, experimental, and analytical factors 64 . This is the first study to examine the issue in small non-coding RNAs. Here, we analyzed miRNA expression in human brain and lung samples and evaluated the linear dynamic range of the Affymetrix miRNA GeneChip. Our objective was to determine the differences, using 2 RNA labeling kits, in the reliability and validity of miRNA expression profiling data using recommended and reduced amount of input RNA. With the exception of the ligation mix, each of the 2 kits used contained identical reagents. Several improvements to the quality of data were seen when different RNA labeling kits were used. First, a statistically significant increase in the mean correlation coefficient for technical replicates in the brain and lung were seen when comparing samples labeled using the biotin-HSR vs. biotin-only kits (Table 1; Supplemental Tables 1-2). These values were consistent with other studies 45 that examined miRNA replicate chip signal intensity correlations 7,8,11 . Second, improvements in the grand median CV values between labeling kits for each tissue was also observed (Figure 2). Although the reduction in CVs was only statistically significant for brain samples, median CV values never exceeded 4%, regardless of the amount of input RNA or labeling kit used. This is in stark contrast to data from Sato and company, who found median CVs across 5 different miRNA microarrays, using two different tissues, to range from 15- 100% 7 . Likewise, detection calls were also improved by labeling of RNA with the biotin-HSR kit (Supplemental Figure 1). The fact that these improvements were not statistically significant suggest that both kits performed equally well, with values ranging from 93.3%-98.4% percent concordant. This data was generally superior to the numbers reported by Sato 7 , as well as the percentages in the MicroArray Quality Control project 64 , although that study examined protein coding gene expression profiling. Taken together, this data shows that both the biotin- only and biotin-HSR labeling approaches are reliable, and that changes to the labeling reagent in the biotin-HSR improves reproducibility of the data. Moreover, this is the first study to show that changes to the miRNA labeling reagent allows for a reduction in the amount of starting material used. First, it was shown that the number of present calls for human miRNA genes was reduced by only a mean value of 7% and 5% for brain and lung samples, respectively, when the amount of biotin-HSR labeled RNA used was decreased 46 by 10-fold from 1.0 ug to 0.1 ug (Figure 1). However, the decline was increased to a mean value of 31% and 25% respectively, when using the biotin-only labeled RNA. This large detection loss seen using biotin-only labeling was similar to that observed by Lynch and company, who reduced input RNA requirements by 2.5, 5 and 25 fold and found a 14%, 14%, and 33% decrease, respectively, in expression detection sensitivity 57 . These data show that detection calls using as little as 0.1ug of biotin-HSR labeled RNA were comparable to those generated when using 1.0ug of biotin-only labeled RNA. Second, differentially expressed gene lists, from reduced input RNA amounts, were correlated to the one derived using 1 ug total RNA labeled with the biotin-only kit, which was the standard recommended amount by Affmetrix/Genisphere. We found that, in contrast to the gene list correlations using reduced amounts of biotin-only labeled RNA, variation in gene rankings were minimized when using reduced amounts of biotin-HSR labeled RNA, suggesting that the use of less starting material does not affect the order of identified genes on a list (Table 2). Third, when directly comparing filtered gene lists, generated using different amounts of biotin-HSR labeled RNA (Figure 3), at least 80% of the genes from each list were differentially co-detected in one or more samples of varying input RNA amounts, indicating a high degree of overlap regardless of the amount of starting material used. 47 A cross-platform validation was undertaken to ensure the integrity of the microarray data. We choose 11 human miRNAs based on statistically significant high/ low differential expression values in both brain and lung samples, as well as a small subset with no difference (Figure 4). Much like the reports from some 7,19,64 , but not all 9 , we found good correlation between TaqMan and microarray data. Furthermore, the correlation values remained >0.98 for all comparisons made using microarray values from any of the RNA dilutions labeled using the biotin-HSR. In contrast, although still high, the correlations fell to a minimal value of 0.915 when evaluating biotin-only labeled RNA samples. Finally, our spike-in experiments showed that the dynamic range of the microarray remained linear for equimolar increases in synthetic miRNAs that varied by 3 orders of magnitude. Similar results were reported by Wang and company, who demonstrated linearity over 2 orders of magnitude using the Agilent microarray platform 66 . These experiments were performed to ensure that the signal intensity values in our experiments did not exceed the linearity limits of the microarray when using the new RNA labeling kit. CHAPTER 3.2.1 CONCLUSION Results from this study highlight RNA labeling as an important factor that can be used as a means of reducing input RNA requirements and improving data quality for high-throughput miRNA expression profiling studies. Ligation mix 48 modifications can result in time and cost savings, by eliminating or minimizing the need for pre-amplification strategies, and enables the profiling of clinical samples with limited quantities of total RNA. CHAPTER 3.2.1 ACKNOWLEDGEMENTS Precision Biomarker Resources performed all the hybridization, washing, and scanning steps for the RNA titration experiments. The authors thank James Cravens and Stacy Berger for their input in data review and presentation. 49 CHAPTER 3.2.1 REFERENCES 1. Benes V, Castoldi M. Expression profiling of microRNA using real-time quantitative PCR, how to use it and what is available. Methods. Apr 2010;50(4):244-249. 2. Creighton CJ, Reid JG, Gunaratne PH. Expression profiling of microRNAs by deep sequencing. Briefings in Bioinformatics. Sep 2009;10(5):490-497. 3. Liu CG, Spizzo R, Calin GA, Croce CM. Expression profiling of microRNA using oligo DNA arrays. Methods. 2008;44(1):22-30. 4. William K, Jian-Jun Z, Lili H, Jin QC. Strategies for profiling MicroRNA expression. Journal of Cellular Physiology. 2009;218(1):22-25. 5. Calin GA, Croce CM. MicroRNA signatures in human cancers. Nat Rev Cancer. 2006;6(11):857-866. 6. Sah S, McCall MN, Eveleigh D, Wilson M, Irizarry RA. Performance evaluation of commercial miRNA expression array platforms. BMC Res Notes. Mar 18 2010;3(1):80. 7. Sato F, Tsuchiya S, Terasawa K, Tsujimoto G. Intra-platform repeatability and inter-platform comparability of microRNA microarray technology. PLoS One. 2009;4(5):e5540. 8. Chen Y, Gelfond J, McManus L, Shireman P. Reproducibility of quantitative RT-PCR array in miRNA expression profiling and comparison with microarray analysis. BMC Genomics. 2009;10(1):407. 9. Git A, Dvinge H, Salmon-Divon M, et al. Systematic comparison of microarray profiling, real-time PCR, and next-generation sequencing technologies for measuring differential microRNA expression. RNA. May 2010;16(5):991-1006. 50 10. Pradervand S, Weber J, Lemoine F, et al. Concordance among digital gene expression, microarrays, and qPCR when measuring differential expression of microRNAs. Biotechniques. Mar 2010;48(3):219-222. 11. Willenbrock H, Salomon J, Sokilde R, et al. Quantitative miRNA expression analysis: comparing microarrays with next-generation sequencing. RNA. Nov 2009;15(11):2028-2034. 12. Yauk CL, Rowan-Carroll A, Stead JD, Williams A. Cross-platform analysis of global microRNA expression technologies. BMC Genomics. May 26 2010;11(1):330. 13. Nelson PT, Wang WX, Wilfred BR, Tang G. Technical variables in high- throughput miRNA expression profiling: much work remains to be done. Biochimica et biophysica acta. Nov 2008;1779(11):758-765. 14. Xi Y, Nakajima G, Gavin E, et al. Systematic analysis of microRNA expression of RNA extracted from fresh frozen and formalin-fixed paraffin- embedded samples. RNA. Oct 2007;13(10):1668-1674. 15. Szafranska AE, Davison TS, Shingara J, et al. Accurate Molecular Characterization of Formalin-Fixed, Paraffin-Embedded Tissues by microRNA Expression Profiling. J Mol Diagn. September 1, 2008;10(5):415-423. 16. Li J, Smyth P, Flavin R, et al. Comparison of miRNA expression patterns using total RNA extracted from matched samples of formalin-fixed paraffin-embedded (FFPE) cells and snap frozen cells. BMC Biotechnol. 2007;7:36. 17. Zhang X, Chen J, Radcliffe T, Lebrun DP, Tron VA, Feilotter H. An array- based analysis of microRNA expression comparing matched frozen and formalin-fixed paraffin-embedded human tissue samples. The Journal of molecular diagnostics : JMD. Nov 2008;10(6):513-519. 18. Abdueva D, Wing M, Schaub B, Triche T, Davicioni E. Quantitative Expression Profiling in Formalin-Fixed Paraffin-Embedded Samples by Affymetrix Microarrays. J Mol Diagn. July 1, 2010;12(4):409-417. 51 19. Ach RA, Wang H, Curry B. Measuring microRNAs: comparisons of microarray and quantitative PCR measurements, and of different total RNA prep methods. BMC Biotechnol. 2008;8:69. 20. Doleshal M, Magotra AA, Choudhury B, Cannon BD, Labourier E, Szafranska AE. Evaluation and validation of total RNA extraction methods for microRNA expression analyses in formalin-fixed, paraffin-embedded tissues. The Journal of molecular diagnostics : JMD. May 2008;10(3):203- 211. 21. Hammerle-Fickinger A, Riedmaier I, Becker C, Meyer HH, Pfaffl MW, Ulbrich SE. Validation of extraction methods for total RNA and miRNA from bovine blood prior to quantitative gene expression analyses. Biotechnology letters. Jan 2010;32(1):35-44. 22. Wang WX, Wilfred BR, Baldwin DA, et al. Focus on RNA isolation: obtaining RNA for microRNA (miRNA) expression profiling analyses of neural tissue. Biochimica et biophysica acta. Nov 2008;1779(11):749-757. 23. Bravo V, Rosero S, Ricordi C, Pastori RL. Instability of miRNA and cDNAs derivatives in RNA preparations. Biochemical and biophysical research communications. Feb 23 2007;353(4):1052-1055. 24. Mraz M, Malinova K, Mayer J, Pospisilova S. MicroRNA isolation and stability in stored RNA samples. Biochemical and biophysical research communications. Dec 4 2009;390(1):1-4. 25. Schroeder A, Mueller O, Stocker S, et al. The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Mol Biol. 2006;7:3. 26. Becker C, Hammerle-Fickinger A, Riedmaier I, Pfaffl MW. mRNA and microRNA quality control for RT-qPCR analysis. Methods. Jan 15 2010; 50(4):237-43. 27. Ibberson D, Benes V, Muckenthaler MU, Castoldi M. RNA degradation compromises the reliability of microRNA expression profiling. BMC Biotechnol. 2009;9:102. 52 28. Jung M, Schaefer A, Steiner I, et al. Robust microRNA stability in degraded RNA preparations from human tissue and cell samples. Clinical chemistry. Jun 2010;56(6):998-1006. 29. Zhu B, Xu F, Baba Y. An evaluation of linear RNA amplification in cDNA microarray gene expression analysis. Molecular genetics and metabolism. Jan 2006;87(1):71-79. 30. Viale A, Li J, Tiesman J, et al. Big results from small samples: evaluation of amplification protocols for gene expression profiling. Journal of biomolecular techniques : JBT. Jul 2007;18(3):150-161. 31. Singh R, Maganti RJ, Jabba SV, et al. Microarray-based comparison of three amplification methods for nanogram amounts of total RNA. American journal of physiology. Cell physiology. May 2005;288(5):C1179- 1189. 32. Patel OV, Suchyta SP, Sipkovsky SS, et al. Validation and application of a high fidelity mRNA linear amplification procedure for profiling gene expression. Veterinary immunology and immunopathology. May 15 2005;105(3-4):331-342. 33. Francois P, Garzoni C, Bento M, Schrenzel J. Comparison of amplification methods for transcriptomic analyses of low abundance prokaryotic RNA sources. Journal of microbiological methods. Feb 2007;68(2):385-391. 34. Nagy ZB, Kelemen JZ, Feher LZ, Zvara A, Juhasz K, Puskas LG. Real- time polymerase chain reaction-based exponential sample amplification for microarray gene expression profiling. Analytical biochemistry. Feb 1 2005;337(1):76-83. 35. Puskas LG, Zvara A, Hackler L, Jr., Van Hummelen P. RNA amplification results in reproducible microarray data with slight ratio bias. Biotechniques. Jun 2002;32(6):1330-1334, 1336, 1338, 1340. 36. Bak M, Conley L, Hedegaard J, et al. Evaluation of two methods for generating cRNA for microarray experiments from nanogram amounts of total RNA. Analytical biochemistry. Nov 1 2006;358(1):111-119. 53 37. Cope L, Hartman SM, Gohlmann HW, Tiesman JP, Irizarry RA. Analysis of Affymetrix GeneChip data using amplified RNA. Biotechniques. Feb 2006;40(2):165-166, 168, 170. 38. Yu J, Othman MI, Farjo R, et al. Evaluation and optimization of procedures for target labeling and hybridization of cDNA microarrays. Molecular vision. Apr 26 2002;8:130-137. 39. Richter A, Schwager C, Hentze S, Ansorge W, Hentze MW, Muckenthaler M. Comparison of fluorescent tag DNA labeling methods used for expression analysis by DNA microarrays. Biotechniques. Sep 2002;33(3):620-628, 630. 40. Ouellet M, Adams PD, Keasling JD, Mukhopadhyay A. A rapid and inexpensive labeling method for microarray gene expression analysis. BMC Biotechnol. 2009;9:97. 41. Manduchi E, Scearce LM, Brestelli JE, Grant GR, Kaestner KH, Stoeckert CJ, Jr. Comparison of different labeling methods for two-channel high- density microarray experiments. Physiological genomics. Sep 3 2002;10(3):169-179. 42. Kimura N, Tamura TA, Murakami M. Evaluation of the performance of two carbodiimide-based cyanine dyes for detecting changes in mRNA expression with DNA microarrays. Biotechniques. May 2005;38(5):797- 806. 43. Badiee A, Eiken HG, Steen VM, Lovlie R. Evaluation of five different cDNA labeling methods for microarrays using spike controls. BMC Biotechnol. Dec 11 2003;3:23. 44. 't Hoen PAC, de Kort F, van Ommen GJB, den Dunnen JT. Fluorescent labelling of cRNA for microarray applications. Nucl. Acids Res. March 1, 2003;31(5):e20. 45. Sawada A, Mizufune S, Kaji N, Tokeshi M, Baba Y. Evaluation of amplified cRNA targets for oligonucleotide microarrays. Analytical and bioanalytical chemistry. Apr 2007;387(8):2645-2654. 54 46. Ma C, Lyons-Weiler M, Liang W, et al. In vitro transcription amplification and labeling methods contribute to the variability of gene expression profiling with DNA microarrays. The Journal of molecular diagnostics : JMD. May 2006;8(2):183-192. 47. Lonergan W, Whistler T, Vernon SD. Comparison of target labeling methods for use with Affymetrix GeneChips. BMC Biotechnol. 2007;7:24. 48. Zakharkin SO, Kim K, Mehta T, et al. Sources of variation in Affymetrix microarray experiments. BMC Bioinformatics. 2005;6:214. 49. Ach RA, Floore A, Curry B, et al. Robust interlaboratory reproducibility of a gene expression signature measurement consistent with the needs of a new generation of diagnostic tools. BMC Genomics. 2007;8:148. 50. Tissot C. Application Note: Analysis of miRNA content in total RNA preparations using the Agilent 2100 bioanalyzer. Publication Number 5989-787OEN. February 2008:1-8. 51. Mora JR, Getts RC. High-sensitivity detection methods for low-abundance RNA species: applications for functional genomics research. Expert Review of Molecular Diagnostics. 2007;7(6):775-785. 52. Saba R, Booth S. Target labelling for the detection and profiling of microRNAs expressed in CNS tissue using microarrays. BMC Biotechnol. 2006;6(1):47. 53. Wu Z, Irizarry RA, Gentleman R, Martinez-Murillo F, Spencer F. A Model- Based Background Adjustment for Oligonucleotide Expression Arrays. Journal of the American Statistical Association. 2004;99(468):909. 54. Irizarry RA, Hobbs B, Collin F, et al. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. Apr 2003;4(2):249-264. 55 55. Rao Y, Lee Y, Jarjoura D, et al. A comparison of normalization techniques for microRNA microarray data. Statistical Applications in Genetics and Molecular Biology. 2008;7(1):Article22. 56. Pradervand S, Weber J, Thomas J, et al. Impact of normalization on miRNA microarray expression profiling. RNA. Mar 2009;15(3):493-501. 57. Lynch A, Hadfield J, Dunning M, Osborne M, Thorne N, Tavare S. The cost of reducing starting RNA quantity for Illumina BeadArrays: A bead- level dilution experiment. BMC Genomics. 2010;11(1):540. 58. Application Note TaqMan MicroRNA Assays: Endogenous Controls for Real-Time Quantitation of miRNA Using TaqMan MicroRNA Assays. Applied Biosystems Publication 127AP11-01. August 2007:1-8. 59. Peltier HJ, Latham GJ. Normalization of microRNA expression levels in quantitative RT-PCR assays: identification of suitable reference RNA targets in normal and cancerous human solid tissues. RNA. May 2008;14(5):844-852. 60. Andersen CL, Jensen JL, Orntoft TF. Normalization of Real-Time Quantitative Reverse Transcription-PCR Data: A Model-Based Variance Estimation Approach to Identify Genes Suited for Normalization, Applied to Bladder and Colon Cancer Data Sets. Cancer Res. August 1, 2004 2004;64(15):5245-5250. 61. Vandesompele J, De Preter K, Pattyn F, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. Jun 18 2002;3(7):RESEARCH0034. 62. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. Dec 2001;25(4):402-408. 63. Livak KJ. ABI Prism 7700 Sequence Detection System. User Bulletin no.2 1997; http://docs.appliedbiosystems.com/pebiodocs/04303859.pdf. 56 64. Shi L, Reid LH, Jones WD, et al. The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. Nature Biotechnology. Sep 2006;24(9):1151-1161. 65. Bammler T, Beyer RP, Bhattacharya S, et al. Standardizing global gene expression analysis between laboratories and across platforms. Nature Methods. May 2005;2(5):351-356. 66. Wang H, Ach RA, Curry B. Direct and sensitive miRNA profiling from low- input total RNA. RNA. January 2007;13(1):151-159. 57 Supplemental Figure 1. Concordance of detection calls within and between RNA labeling kits. For each input RNA amount (0.1-1.0ug), the number of absent probes are represented by grey bars, present by black and white checker bars, and discrepant as a black line. Discrepant probes are those with an absent call on one chip and a present call on a corresponding replicate array. Detection calls were generated for 7788 of 7819 probes on the array. 58 Brain 0 50 100 150 200 250 300 1 vs 1 1 vs. 0.5 1 vs. 0.2 1 vs. 0.1 Input RNA (in ug) Detected Human miRNA Genes Lung 0 50 100 150 200 250 300 1 vs 1 1 vs. 0.5 1 vs. 0.2 1 vs. 0.1 Input RNA (in ug) Detected Human miRNA Genes Biotin-HSR vs. Biotin-HSR Biotin-Only vs. Biotin-HSR Biotin-Only vs. Biotin-Only Figure 1. Comparison of Detected miRNA Genes. Probes for all 847 human miRNAs were examined in Brain and Lung samples (top and bottom panel, respectively). The number of present calls is indicated by the y-axis. Comparisons on the x-axis was done to determine the number of overlapping present calls on chips containing 1ug of input RNA vs. 0.5, 0.2, and 0.1ug of starting material. One vs. 1ug indicates the maximum possible overlap. Evaluations using the biotin-only kit are represented by black circles, biotin-HSR by black squares, and biotin-only vs. biotin-HSR by white triangles. Statistical significance represented by a star and found for comparisons made in both brain and lung samples (p<0.05). 59 Table 1. Correlation of Replicate Chip Signal Intensity Values. The processed signal intensity value for each probe on a chip was correlated with the corresponding probe on a replicate array (7819 probes total). p-values for individual correlations between replicate arrays were all found to be <0.0001 (supplemental tables 1-2). Global mean calculations did not include correlations of a chip with itself, i.e. where r=1. Pearson Product Moment Correlation (r) Labeling Kit Tissue Biotin-Only Global Mean (±1SD) Biotin-HSR Global Mean (±1SD) p-value Brain 0.955 (0.023) 0.978 (0.010) <0.0001 Lung 0.976 (0.011) 0.985 (0.006) <0.0001 60 Biotin-Only Biotin-HSR Input RNA Amount (Brain Samples) 1.0ug rep1 1.0ug rep2 0.5ug rep1 0.5ug rep2 0.2ug rep1 0.2ug rep2 0.1ug rep1 0.1ug rep2 1.0ug rep1 1.0ug rep2 0.5ug rep1 0.5ug rep2 0.2ug rep1 0.2ug rep2 0.1ug rep1 0.1ug rep2 1.0ug rep1 1.000 0.988 0.984 0.984 0.959 0.964 0.921 0.916 0.966 0.967 0.969 0.969 0.971 0.970 0.965 0.966 1.0ug rep2 0.988 1.000 0.983 0.984 0.957 0.963 0.919 0.912 0.966 0.966 0.969 0.967 0.970 0.969 0.965 0.964 0.5ug rep1 0.984 0.983 1.000 0.984 0.969 0.972 0.936 0.931 0.950 0.950 0.955 0.954 0.963 0.963 0.963 0.963 0.5ug rep2 0.984 0.984 0.984 1.000 0.967 0.971 0.933 0.927 0.952 0.952 0.957 0.956 0.964 0.963 0.964 0.962 0.2ug rep1 0.959 0.957 0.969 0.967 1.000 0.971 0.953 0.949 0.908 0.908 0.916 0.916 0.933 0.934 0.940 0.941 0.2ug rep2 0.964 0.963 0.972 0.971 0.971 1.000 0.953 0.947 0.917 0.917 0.925 0.924 0.940 0.941 0.946 0.947 0.1ug rep1 0.921 0.919 0.936 0.933 0.953 0.953 1.000 0.951 0.858 0.859 0.868 0.867 0.890 0.891 0.902 0.904 Biotin Only 0.1ug rep2 0.916 0.912 0.931 0.927 0.949 0.947 0.951 1.000 0.849 0.850 0.859 0.859 0.881 0.884 0.894 0.898 1.0ug rep1 0.966 0.966 0.950 0.952 0.908 0.917 0.858 0.849 1.000 0.994 0.991 0.992 0.977 0.980 0.961 0.963 1.0ug rep2 0.967 0.966 0.950 0.952 0.908 0.917 0.859 0.850 0.994 1.000 0.991 0.992 0.977 0.980 0.961 0.963 0.5ug rep1 0.969 0.969 0.955 0.957 0.916 0.925 0.868 0.859 0.991 0.991 1.000 0.991 0.981 0.983 0.968 0.969 0.5ug rep2 0.969 0.967 0.954 0.956 0.916 0.924 0.867 0.859 0.992 0.992 0.991 1.000 0.980 0.983 0.967 0.969 0.2ug rep1 0.971 0.970 0.963 0.964 0.933 0.940 0.890 0.881 0.977 0.977 0.981 0.980 1.000 0.983 0.976 0.976 0.2ug rep2 0.970 0.969 0.963 0.963 0.934 0.941 0.891 0.884 0.980 0.980 0.983 0.983 0.983 1.000 0.976 0.979 0.1ug rep1 0.965 0.965 0.963 0.964 0.940 0.946 0.902 0.894 0.961 0.961 0.968 0.967 0.976 0.976 1.000 0.975 Biotin HSR 0.1ug rep2 0.966 0.964 0.963 0.962 0.941 0.947 0.904 0.898 0.963 0.963 0.969 0.969 0.976 0.979 0.975 1.000 Supplemental Table 1. Brain Sample Correlations of Processed Signal Intensity Values. Correlations of processed signal intensity values for all 16 chips with brain RNA labeled using either the Biotin-Only (8 samples) or Biotin-HSR (8 samples) kits. All correlation p-values<0.0001. 61 Supplemental Table 2. Lung Sample Correlations of Processed Signal Intensity Values. Correlations of processed signal intensity values for all 16 chips with lung RNA labeled using either the Biotin-Only (8 samples) or Biotin-HSR (8 samples) kits. All correlation p-values<0.0001. Biotin-Only Biotin-HSR Input RNA Amount (Lung Samples) 1.0ug rep1 1.0ug rep2 0.5ug rep1 0.5ug rep2 0.2ug rep1 0.2ug rep2 0.1ug rep1 0.1ug rep2 1.0ug rep1 1.0ug rep2 0.5ug rep1 0.5ug rep2 0.2ug rep1 0.2ug rep2 0.1ug rep1 0.1ug rep2 1.0ug rep1 1.000 0.993 0.990 0.989 0.982 0.975 0.958 0.958 0.971 0.968 0.972 0.971 0.976 0.973 0.973 0.973 1.0ug rep2 0.993 1.000 0.989 0.989 0.980 0.973 0.954 0.955 0.976 0.973 0.977 0.975 0.979 0.976 0.976 0.975 0.5ug rep1 0.990 0.989 1.000 0.990 0.984 0.979 0.964 0.963 0.963 0.961 0.966 0.964 0.971 0.970 0.971 0.970 0.5ug rep2 0.989 0.989 0.990 1.000 0.985 0.980 0.967 0.964 0.962 0.960 0.965 0.963 0.971 0.970 0.971 0.971 0.2ug rep1 0.982 0.980 0.984 0.985 1.000 0.983 0.975 0.975 0.945 0.941 0.949 0.947 0.961 0.959 0.964 0.964 0.2ug rep2 0.975 0.973 0.979 0.980 0.983 1.000 0.976 0.975 0.933 0.930 0.937 0.933 0.949 0.949 0.954 0.954 0.1ug rep1 0.958 0.954 0.964 0.967 0.975 0.976 1.000 0.976 0.907 0.904 0.912 0.909 0.928 0.927 0.936 0.936 Biotin Only 0.1ug rep2 0.958 0.955 0.963 0.964 0.975 0.975 0.976 1.000 0.906 0.901 0.911 0.908 0.928 0.925 0.937 0.935 1.0ug rep1 0.971 0.976 0.963 0.962 0.945 0.933 0.907 0.906 1.000 0.995 0.994 0.994 0.987 0.982 0.976 0.976 1.0ug rep2 0.968 0.973 0.961 0.960 0.941 0.930 0.904 0.901 0.995 1.000 0.994 0.994 0.985 0.982 0.974 0.974 0.5ug rep1 0.972 0.977 0.966 0.965 0.949 0.937 0.912 0.911 0.994 0.994 1.000 0.995 0.989 0.985 0.980 0.979 0.5ug rep2 0.971 0.975 0.964 0.963 0.947 0.933 0.909 0.908 0.994 0.994 0.995 1.000 0.989 0.984 0.979 0.978 0.2ug rep1 0.976 0.979 0.971 0.971 0.961 0.949 0.928 0.928 0.987 0.985 0.989 0.989 1.000 0.987 0.987 0.986 0.2ug rep2 0.973 0.976 0.970 0.970 0.959 0.949 0.927 0.925 0.982 0.982 0.985 0.984 0.987 1.000 0.983 0.983 0.1ug rep1 0.973 0.976 0.971 0.971 0.964 0.954 0.936 0.937 0.976 0.974 0.980 0.979 0.987 0.983 1.000 0.985 Biotin HSR 0.1ug rep2 0.973 0.975 0.970 0.971 0.964 0.954 0.936 0.935 0.976 0.974 0.979 0.978 0.986 0.983 0.985 1.000 62 Figure 2. Coefficient of Variation within and between RNA Labeling Kits. The percent CV was averaged for replicate chips at each input RNA amount (0.1-1.0ug) within and between labeling kits for both brain and lung samples. The interquartile range of the processed signal intensity value is represented by grey boxes, median value by a black line within a box, the 5 th and 95 th percentiles by the whiskers outside of each box and the number of detected miRNA genes by black boxes. CVs were only calculated for miRNA genes called as present. 63 Gene List Filters Applied A (number of probes) 0 (n=7819) 1 (n=1779) 2 (n=601) 3 (n=422) 4 (n=272) Spearmans Rank Correlation Coefficient 0.5ug 0.765 0.734 0.961 0.972 0.978 0.2ug 0.693 0.662 0.915 0.938 0.949 0.1ug 0.588 0.558 0.830 0.879 0.903 Biotin-Only Kit Range (Max-Min) 0.177 0.176 0.131 0.093 0.075 0.5ug 0.791 0.762 0.962 0.968 0.973 0.2ug 0.766 0.733 0.957 0.970 0.977 0.1ug 0.721 0.692 0.937 0.960 0.972 Biotin-HSR Kit Range (Max-Min) 0.070 0.070 0.025 0.010 0.005 % Reduction in Range Using Biotin-HSR Kit 60 60 81 89 93 Table 2. Gene List Agreement within and between RNA Labeling Kits. The brain vs. lung differentially expressed gene list, derived from1.0ug of biotin-only labeled RNA, was used in all pairwise correlations made with lists generated using reduced amounts of starting material. The use of at least 1.0ug of biotin-only labeled total RNA is required by the manufacturer and thus defined as the standard. A) Filters to the 1.0ug biotin-only gene list included 0) none, i.e. all probes on the array, 1) human probes only, 2) filter 1 + probe must be present in at least 1 tissue, 3) filter 2 + absolute fold change >1.5, and 4) filter 3 + false discovery rate p<0.05. All correlations were statistically significant (p<0.0001; data not shown). 64 Figure 3. Evaluation of Statistically Significant Differentially-Expressed Genes Using Biotin-HSR Labeled RNA. Genes were included in this analysis using only human probes detected as present in at least brain or lung tissue, with absolute fold change > 1.5, and an FDR p-value of <0.05. A total of 419, 376, 218, and 184 differentially expressed genes were identified when using 1.0ug, 0.5ug, 0.2ug, and 0.1ug of input RNA, respectively. 0.2ug (218) 0.5ug (376) 0.1ug (184) 1.0ug (419) 65 0.1ug r=0.915 -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 1 Log2 (TaqMan Ratio) Log2 (Array Ratio) 0.1ug 8 10 r =0.986 -10 -8 -6 -4 -2 0 2 4 6 -10-8 -6 -4 -2 024 68 10 Log2 (Array Ratio) 0 Log2 (TaqMan Ratio) 0.2ug r =0.964 -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 1 Log2 (TaqMan Ratio) Log2 (Array Ratio) 0 0.5ug r =0.975 -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 1 Log2 (TaqMan Ratio) Log2 (Array Ratio) 0 1.0ug r =0.994 -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 1 Log2 (TaqMan Ratio) Log2 (Array Ratio) 0 0.2ug r =0.995 -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 10 Log2 (Array Ratio) Log2 (TaqMan Ratio) 0.5ug r -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 Log2 (TaqMan Ratio) Log2 (Array Ratio) =0.994 10 1.0ug r -10 -8 -6 -4 -2 0 2 4 6 8 10 -10-8 -6 -4 -2 024 68 Log2 (TaqMan Ratio) Log2 (Array Ratio) =0.996 10 Biotin-HSR Labeled RNA Biotin-Only Labeled RNA Figure 4. Correlation of Microarray Data Using TaqMan MicroRNA Assays. Selection of human miRNAs based on statistically significant differential expression a) in the brain (1 high, 2 lows) and lung (1 high, 1 low), and b) found exclusively using either Biotin-HSR labeled RNA (1 high, 1 low) or Biotin-Only labeled RNA (2 lows). Two additional miRNAs that were not statistically significant by either labeling method (1 high, 1 low), and 1 low detected using Biotin-only labeled RNA, were found to be undetectable by TaqMan Assay and therefore excluded. A total of 8 miRNAs were used to generate these graphs. Fold change values and names of each miRNA can be found in supplemental table 3. Linear regression line determined using least squares method. 66 5 6 7 8 9 10 11 12 13 14 15 2468 10 12 Log2 (Attomoles) Log2 (Array Signal) Synthetic Human miRNAs (n=465); r=0.991 Human Brain miRNAs (n=129) Human Lung miRNAs (n=141) Figure 5. Dynamic Range of miRNA Gene Chip Using Biotin-HSR Labeling Kit. An equimolar pool of synthetic human, mouse, and rat miRNA was labeled using the Biotin-HSR kit and hybridized in amounts ranging from 4.88 to 5000 attomoles (0 datapoint not shown; 12 arrays total). Probes for 465 of 470 human, 223 of 224 mouse, and 42 of 42 rat synthetic miRNAs were detected in all hybridizations (rat and mouse data not shown). Global mean ( ± SEM) represented by X and error bars, respectively. Linear regression line in black. Intensity values from the same human probes detected in all brain or lung hybridizations (0.1-1.0ug) were plotted as black squares or white triangles, respectively. Brain and lung data plotted as grand mean ( ± grand SEM). 67 CHAPTER 3: PREDICTING HUMAN ISLET ISOLATION AND TRANSPLANTATION OUTCOMES IN TYPE 1 DIABETES Section 2.2: Optimization of Microarray Analysis Application: Identification of Predictive Genetic Markers Associated with Favorable Human Islet Transplantation Outcomes CHAPTER 3.2.2 ABSTRACT Human islet transplantation relies on the use of allogeneic organ donors and subsequent pancreas processing to isolate the cell clusters of interest. In chapter 3, section 1 we identified several important donor and pancreas processing factors associated with successful human islet isolation outcomes. However, very little work has been done to understand the influence of donor islet genomics on transplantation success. This pilot study seeks to determine the genomic basis of human islet graft variability, after adjusting for organ donor, pancreas processing, and islet characterization factors. A total of 26 human islet preparations from different organ donors were used for this study. Preliminary results show that there exists genomic differences between transplant quality preparations that do not correlate with accepted indicators of successful human islet isolation outcomes, suggesting that islet-intrinsic biological variability may play a under appreciated role in transplant success. Future work is currently underway to explore these findings. 68 CHAPTER 3.2.2 INTRODUCTION Islet transplantation is one form of cellular therapy currently being used to treat a select subgroup of type 1 diabetic individuals and requires the removal of islets from the pancreas of an organ donor for transplantation into a qualified recipient 1 . Long-term insulin independence rates now hover around 50%, and while some of these transplant recipients will resume some form of exogenous insulin therapy, the benefits of retaining graft function include measurable c-peptide production, a reduction or elimination of severe hypoglycemic unawareness, stabilization of blood glucose, and improvements in patient quality of life 2 . Nonetheless, there is a need to identify and understand the factors predictive of clinical outcomes. A number of studies appearing in the literature have sought to investigate parameters that could help to dissect the variables involved in discriminating between highly functional islet grafts post-transplant and those that loose their ability to respond to physiological changes in vivo. In Chapter 3, section 1, a thorough review of all published original studies in the last 15 years that analyzed the influence of organ donor characteristics and islet processing variables on human islet isolation outcome was conducted 3 . Of 45 variables examined, a total of 17 were univariately found to influence the odds of isolating high quality pancreatic islets. A multivariable logistic regression model showed that minimizing cold ischemia time of a carefully preserved intact pancreas free from edema with moderate to heavy fat infiltration from a hemodynamically 69 stable, high BMI, younger donor will improve the odds of obtaining a high yielding human islet isolation. However, these factors alone have not yet been shown to be associated with successful transplantation outcome. Moreover, the contributions of donor islet genomic and functional factors have not yet been evaluated. We therefore conducted a pilot study to identify the modifying effects of islet genomics data on functional, donor, and pancreas processing characteristics associated with successful human islet isolation outcome. CHAPTER 3.2.2 MATERIALS AND METHODS Organ Procurement and Human Islet Isolation. Human pancreata were obtained from heart-beating brain-dead cadaveric multi-organ donors and preserved in either University of Wisconsin (UW) or histidine-tryptophan-ketoglutarate (HTK) solution. Islets were isolated in the cell isolation laboratory at the University of Illinois at Chicago (UIC), Illinois following a method previously described 4,5 . All procedures regarding the acquisition of human islets were approved by the Human Research Ethics Committee at UIC. Preparation Quality and RNA Extraction. Purity of the islets was assessed using Dithizone staining, as previously described 6 . Viability was determined using fluorescent staining with Syto-Green/Ethidium Bromide, as previously 70 described 7,8 . Islet function was evaluated using the glucose stimulated insulin release (GSIR) assay, measured as a stimulation index (SI) after static incubation with low and high glucose conditions, as previously described 9 . Islet yield was expressed as islet equivalent (IEQ) count, based on a universally- adopted standardized scale 10 . Endotoxin was reported as Endotoxin Units (EU) per mL tissue volume and determined using a standard assay. A total of 27 samples were included in this study that met the following established clinical transplantation lot release criteria 11 : a) purity >30%, b) viability >70%, c) SI > 1, d) yield > 105,000 IEQs, calculated using a 35kg transplant recipient and minimum of 3,000IEQ/kg bodyweight, and 5) contamination of < 5 EU/kg recipient bodyweight. We used 35kg as our theoretical number because this was the smallest known weight of a transplant recipient 12 , although the mean value was 66kg. We used 16mL to calculate total EU per preparation because this was the largest cell volume ever reported 12 , although the mean value was 3.5mL. RNA from 27 total isolations was extracted using human islets sampled from the highest purity fraction of the preparation. Covariate Data. Pancreas islet isolation records for 27 samples were obtained from the Islet Cell Resource Center (ICR) database, as previously described 3 . Of those, records from 26 samples were linked to organ donor data from the United Network for Organ Sharing (UNOS). Complete covariate data was available for 26 of 27 human islet isolations. 71 RNA Labeling and Microarray Processing. RNA from 26 human islet preparations was labeled, hybridized, washed, and scanned by the Cleveland Clinic microarray core laboratory using the Illumina Sentrix Human HT-12 Expression BeadChip. Twelve additional samples were profiled, but not used for this study because they did not meet the above preparation quality criteria. Each chip contains 12 arrays and each array contains >47,000 probes for sequences derived from RefSeq and UniGene content, including coding and non-coding transcripts from well-established and provisional annotations (n=43,770 probes) and experimentally confirmed mRNA sequences that align to EST clusters (n=3461 probes). Array Quality, Data Processing, and Detection Calls. Chip quality was assessed by the Cleveland Clinic microarray core laboratory. Raw intensity signal values were background subtracted and average normalized using Illumina GenomeStudio data analysis software. Present probe calls were based on detection p-values of <0.05. Detection p-values are generated in the software by testing for nonspecific cross-hybridization by looking at differences in the signals of the reporting probe vs. the thermodynamically equivalent negative control compliment that lacks specific targets in the transcriptome. Probes were used if the detection p-values<0.05 in at least 1 sample; therefore, 30,689 remained. Of 72 those, only 22,070 were used for this study based on RefSeq content of well- annotated coding and non-coding transcripts. Composite Donor Score. The multivariate logistic regression analysis performed by Kaddis and colleagues 3 was used to develop a composite donor score for each sample to numerically segregate high quality samples using statistically significant factors shown to influence human islet isolation outcomes (Figure 1). Exponentiated beta-parameters, i.e. odds ratio (OR), from the multivariate equation were converted into points using a scale adopted from Selby et al 13 as follows: a value of 1 was assigned to an OR between 1.1 and 1.49, 2 for an OR between 1.50 and 1.99, and 3 for OR’s >2. Corresponding negative ORs were assigned -1, -2, and -3 points, respectively. Values for the 2 continuous variables in the model were assigned as follows: a) BMI – a BMI specific OR distribution was generated using the model coefficient for 8.09 kg/m 2 increases in BMI, from the lowest to highest reported BMI, and then used to categorize BMI into values ranging from 0 to 9, b) Cold ischemia (CI) – a CI specific OR distribution was generated using the model coefficient for 4.90 hrs increases in CI, from the lowest to highest reported CI, and then used to categorize CI into values ranging from 0 to -3. A composite donor score was generated by adding up the values from each factor. Isolations were then grouped into tertiles as follows: first tertile (-8 to 0 points), second tertile (1 to 4 points), and third tertile (5 to 11 points). Samples in the 2 nd and 3 rd tertile were then combined into a single 73 group and represent intermediate to poor quality preparations. Samples with the highest scores represent those human islet isolations of best quality based on significant donor, pancreas, and islet isolation factors from the multivariate model (Figure 2). A scoring system has been successfully used in the past to categorize human islet preparations 14 , but points were based only on perceived importance of factors. Data Analysis. The mean or median was used a measure of central tendency for all continuous variables, depending on whether the values were normally distributed. Fisher’s exact test was used to measure the agreement between the composite donor score and successful human islet isolation outcome. Principal Component Analysis (PCA) plots were used to visualize processed microarray signal data. Two-group comparisons, from clusters identified in an unsupervised fashion, were performed using one-way analysis of variance (ANOVA). All resultant p-values were false discovery rate (FDR) corrected using the method of Benjamini and Hochberg and defined as statistically significant if <0.05. . All data analysis was performed using either Partek Genomics Suite or SAS software, version 9.1.3 SP4 (SAS Institute, Cary, NC). CHAPTER 3.2.2 PRELIMINARY RESULTS Sample Assessment. Donor, pancreas, and post-isolation characteristics of all 26 human islet preparations used in this study were examined (Table 1). 74 Performance of Composite Donor Score. Independent predictors of human islet isolation success were used to generate a composite point score for each preparation. The point scale ranged from a maximum possible of 15 to a minimum of -12; however, the actual scores varied from 11 to -8. There was statistically significant agreement between the total number of points and the number of successful human islet isolations (p=0.0093; Figure 2). Exploratory Gene Expression Analysis of Human Islet Preparations. Principal Component Analysis (PCA) mapping was performed to examine clusters of potential biological importance (Figure 3). Using all 22,070 probes on the array, we identified 2 clusters that natively aggregated bilaterally and explained 45.9% of the total variation seen by principal components (PCs) 1-3 (Figure 3A). A one- way ANOVA by cluster group found that only 396 of these probes were statistically significantly different between the 2 groups (FDR p<0.05, absolute fold change >1.5). The total variation explained by PCs 1-3 was improved to 68.8% when restricting the PCA map to only those probes (Figure 3B). Cluster group was not statistically significantly associated with post purification IEQ count (p=1; IEQ count as categorized in figure 2), islet morphology score (p=1), or composite donor score (p=0.0256). Fishers exact test was used to measure these associations and statistical significance was indicated only if bonferroni adjusted p<0.017. 75 CHAPTER 3.2.2 DISCUSSION AND FUTURE WORK High throughput expression profiling has been used to examine the genomic contributions of human pancreatic islets under a number of different conditions. Early studies focused on the baseline transcript expression of human islets after isolation 15,16 and culture 17-19 . Later work clarified the genes enriched in human islets 20 vs. those from pancreatic endocrine neoplasms 21,22 , human insulinomas 23 , pancreatic exocrine, liver, and kidney tissues 24 . A desire to understand islet dysfunction 25 has also resulted in studies using type 2 diabetic islets 26-28 , free fatty acid 29 and cytokine 30-36 stimulated human islets. The inflammatory profile of human islets has also been examined in order to understand the autoimmune response in type 1 diabetes 37-41 . Recent work has now focused on the effects of drug treatment in human islets 42 , cellular ER stress 43 , in-vitro expansion 44 , and identification of beta-cell specific biomarkers 45,46 . The accumulation of this data has resulted in islet specific database repositories, such as GeneSpeed Beta Cell 47 and EuroDia 48 . Nearly all of these studies were restricted to less than 10 human islet preparations and none examined differences between samples that might account for the variability in performance seen post-transplant. Our pilot study is the first to identify genomic variability between preparations, even after accounting for established factors associated with successful human islet isolation outcomes, using one of the largest cohort of samples to date (n=26). 76 Although additional samples were available for this study, we chose to limit our analysis to only those islet preparations that could have been, and in some cases were (n=5), used for human islet transplantation. However, this sample size was not suitable for correlations with clinical outcome. Additional samples are currently being processed for expansion of this dataset. Other limitations are currently being addressed. First, although established clinical lot release criteria were used to screen samples, these measures are seen as crude when compared to new islet characterization techniques 11 . This study utilized basic islet functional data, but detailed analysis 49,50 of all islet preparation are currently underway. Second, the composite donor score performed well when we tested it using our training dataset, but we did not generate sensitivity and specificity receiver operator characteristic (ROC) based curves to optimize cut points. We are currently acquiring an independent dataset to formally evaluate these measures. Finally, future work is focused on identifying a potential biological role for the identified cluster groups. The Database for Annotation, Visualization, and Integrated Discovery (DAVID; available at http://david.abcc.ncifcrf.gov) is being used to examine genes for enrichment of functional categories 51 . Primary focus is on the annotation clusters with the Gene Ontology Terms for Biological Process (GOTERM-BP), Cellular Function (GOTERM-CC) and Molecular Function 77 (GOTERM-MF). Each annotation cluster is assigned an enrichment score to quantify the global importance of gene groups within a cluster. KEGG pathway analysis will also be performed through DAVID. CHAPTER 3.2.2 CONCLUSION This pilot study provides some preliminary evidence that despite optimization of organ donor selection, pancreas and isolation procedures, intrinsic biological variability exists between islet preparations that may be used to help predict transplantation success. CHAPTER 3.2.2 ACKNOWLEDGEMENTS Human pancreatic islets, RNA, and islet function data were provided by Dr. Jose Oberholzer from the UIC. Microarray samples were processed at the core laboratory facility of the Cleveland Clinic under the direction of Dr. Jan Jensen, who also provided the processed data files for this study. Organ donor information was supplied by UNOS as the contractor for the Organ Procurement and Transplantation Network (OPTN). The interpretation and reporting of such data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by the OPTN or the U.S. Government. 78 Pancreas and islet processing data was acquired through the ICR Administrative and Bioinformatics Coordinating Center database, made possible through the cooperative funding efforts of the National Center for Research Resources and the National Institute of Diabetes and Digestive and Kidney Diseases, a component of the US National Institutes of Health, in conjunction with the generous contributions of the Juvenile Diabetes Research Foundation (U42 RR 017673 to J.C. Niland and U42 RR 023245 to J. Oberholzer). 79 CHAPTER 3.2.2 REFERENCES 1. Kaddis JS, Olack BJ, Sowinski J, Cravens J, Contreras JL, Niland JC. Human Pancreatic Islets and Diabetes Research. JAMA. April 15, 2009 2009;301(15):1580-1587. 2. Robertson RP. Islet Transplantation a Decade Later and Strategies for Filling a Half-Full Glass. Diabetes. 06/01 2010;59(6):1285-1291. 3. Kaddis JS, Danobeitia JS, Niland JC, Stiller T, Fernandez LA. Multicenter Analysis of Novel and Established Variables Associated with Successful Human Islet Isolation Outcomes. American Journal of Transplantation. 2010;10(3):646-656. 4. Qi M, Strand BL, Morch Y, et al. Encapsulation of human islets in novel inhomogeneous alginate-ca2+/ba2+ microbeads: in vitro and in vivo function. Artificial cells, blood substitutes, and immobilization biotechnology. 2008;36(5):403-420. 5. Ricordi C, Lacy PE, Scharp DW. Automated islet isolation from human pancreas. Diabetes. Jan 1989;38 Suppl 1:140-142. 6. Ricordi C, Gray DW, Hering BJ, et al. Islet isolation assessment in man and large animals. Acta diabetologica latina. Jul-Sep 1990;27(3):185-195. 7. Yang H, Acker J, Chen A, McGann L. In situ assessment of cell viability. Cell Transplant. Sep-Oct 1998;7(5):443-451. 8. Avila JG, Tsujimura T, Oberholzer J, et al. Improvement of pancreatic islet isolation outcomes using glutamine perfusion during isolation procedure. Cell Transplant. 2003;12(8):877-881. 9. Tsujimura T, Kuroda Y, Kin T, et al. Human islet transplantation from pancreases with prolonged cold ischemia using additional preservation by the two-layer (UW solution/perfluorochemical) cold-storage method. Transplantation. Dec 27 2002;74(12):1687-1691. 80 10. Ricordi C. Quantitative and qualitative standards for islet isolation assessment in humans and large mammals. Pancreas. Mar 1991;6(2):242-244. 11. Papas KK, Suszynski TM, Colton CK. Islet assessment for transplantation. Current Opinion in Organ Transplantation. 2009;14(6):674-682 12. Collaborative Islet Transplant Registry (CITR) 6th Annual Report, 2009. Last accessed 03/11/2011: https://web.emmes.com/ study/isl/reports/CITR %206th%20Annual%20Data%20Report%20120109.pdf. 13. Selby JV, Karter AJ, Ackerson LM, Ferrara A, Liu J. Developing a prediction rule from automated clinical databases to identify high-risk patients in a large population with diabetes. Diabetes Care. Sep 2001;24(9):1547-1555. 14. O'Gorman D, Kin T, Murdoch T, et al. The Standardization of Pancreatic Donors for Islet Isolations. Transplantation. 2005;80(6):801-806. 15. Cras-Meneur C, Inoue H, Zhou Y, et al. An expression profile of human pancreatic islet mRNAs by Serial Analysis of Gene Expression (SAGE). Diabetologia. Feb 2004;47(2):284-299. 16. Shalev A, Pise-Masison CA, Radonovich M, et al. Oligonucleotide microarray analysis of intact human pancreatic islets: identification of glucose-responsive genes and a highly regulated TGFbeta signaling pathway. Endocrinology. Sep 2002;143(9):3695-3698. 17. Barry R, Tadayyon M, Green IC. Reproducibility of targeted gene expression measurements in human islets of Langerhans. Biochem Biophys Res Commun. Nov 1 2002;298(3):350-356. 18. Hui H, Wang C, Li H, et al. Gene expression profiling of cultured human islet preparations. Diabetes Technol Ther. Aug 2004;6(4):481-492. 81 19. Sabek OM, Marshall DR, Penmetsa R, Scarborough O, Gaber AO. Examination of gene expression profile of functional human pancreatic islets after 2-week culture. Transplant Proc. Dec 2006;38(10):3678-3679. 20. Kaestner KH, Lee CS, Scearce LM, et al. Transcriptional program of the endocrine pancreas in mice and humans. Diabetes. Jul 2003;52(7):1604- 1610. 21. Maitra A, Hansel DE, Argani P, et al. Global expression analysis of well- differentiated pancreatic endocrine neoplasms using oligonucleotide microarrays. Clin Cancer Res. Dec 1 2003;9(16 Pt 1):5988-5995. 22. Capurso G, Crnogorac-Jurcevic T, Milione M, et al. Peanut-like 1 (septin 5) gene expression in normal and neoplastic human endocrine pancreas. Neuroendocrinology. 2005;81(5):311-321. 23. Nabokikh A, Ilhan A, Bilban M, et al. Reduced TGF-beta1 expression and its target genes in human insulinomas. Exp Clin Endocrinol Diabetes. Nov 2007;115(10):674-682. 24. Maffei A, Liu Z, Witkowski P, et al. Identification of tissue-restricted transcripts in human islets. Endocrinology. Oct 2004;145(10):4513-4521. 25. Marshall D, Sabek O, Fraga D, Kotb M, Gaber AO. Examination of the molecular signature associated with islet dysfunction. Transplant Proc. Mar 2005;37(2):1311-1312. 26. Gunton JE, Kulkarni RN, Yim S, et al. Loss of ARNT/HIF1beta mediates altered gene expression and pancreatic-islet dysfunction in human type 2 diabetes. Cell. Aug 12 2005;122(3):337-349. 27. Ostenson CG, Gaisano H, Sheu L, Tibell A, Bartfai T. Impaired gene and protein expression of exocytotic soluble N-ethylmaleimide attachment protein receptor complex proteins in pancreatic islets of type 2 diabetic patients. Diabetes. Feb 2006;55(2):435-440. 82 28. Igoillo-Esteve M, Marselli L, Cunha DA, et al. Palmitate induces a pro- inflammatory response in human pancreatic islets that mimics CCL2 expression by beta cells in type 2 diabetes. Diabetologia. Jul 2010;53(7):1395-1405. 29. Bikopoulos G, da Silva Pimenta A, Lee SC, et al. Ex vivo transcriptional profiling of human pancreatic islets following chronic exposure to monounsaturated fatty acids. J Endocrinol. Mar 2008;196(3):455-464. 30. Liuwantara D, Elliot M, Smith MW, et al. Nuclear factor-kappaB regulates beta-cell death: a critical role for A20 in beta-cell protection. Diabetes. Sep 2006;55(9):2491-2501. 31. Riachy R, Vandewalle B, Moerman E, et al. 1,25-Dihydroxyvitamin D3 protects human pancreatic islets against cytokine-induced apoptosis via down-regulation of the Fas receptor. Apoptosis. Feb 2006;11(2):151-159. 32. Bergholdt R, Karlsen AE, Hagedorn PH, et al. Transcriptional profiling of type 1 diabetes genes on chromosome 21 in a rat beta-cell line and human pancreatic islets. Genes Immun. Apr 2007;8(3):232-238. 33. Sarkar SA, Wong R, Hackl SI, et al. Induction of indoleamine 2,3- dioxygenase by interferon-gamma in human islets. Diabetes. Jan 2007;56(1):72-79. 34. Chambers KT, Unverferth JA, Weber SM, Wek RC, Urano F, Corbett JA. The role of nitric oxide and the unfolded protein response in cytokine- induced beta-cell death. Diabetes. Jan 2008;57(1):124-132. 35. Klein D, Timoneri F, Ichii H, Ricordi C, Pastori RL. CD40 activation in human pancreatic islets and ductal cells. Diabetologia. Oct 2008;51(10):1853-1861. 36. Sarkar SA, Kutlu B, Velmurugan K, et al. Cytokine-mediated induction of anti-apoptotic genes that are linked to nuclear factor kappa-B (NF- kappaB) signalling in human islets and in a mouse beta cell line. Diabetologia. Jun 2009;52(6):1092-1101. 83 37. Johansson U, Olsson A, Gabrielsson S, Nilsson B, Korsgren O. Inflammatory mediators expressed in human islets of Langerhans: implications for islet transplantation. Biochem Biophys Res Commun. Aug 29 2003;308(3):474-479. 38. Olsson A, Johansson U, Korsgren O, Frisk G. Inflammatory gene expression in Coxsackievirus B-4-infected human islets of Langerhans. Biochem Biophys Res Commun. May 6 2005;330(2):571-576. 39. Ylipaasto P, Kutlu B, Rasilainen S, et al. Global profiling of coxsackievirus- and cytokine-induced gene expression in human pancreatic islets. Diabetologia. Aug 2005;48(8):1510-1522. 40. Hultcrantz M, Huhn MH, Wolf M, et al. Interferons induce an antiviral state in human pancreatic islet cells. Virology. Oct 10 2007;367(1):92-101. 41. Jackson A, McWilliams C, Kaizer E, et al. Gene expression profiling of human pancreatic islets undergoing a simulated process of instant blood- mediated inflammatory reaction. Transplant Proc. Mar 2008;40(2):430- 432. 42. Couto FM, Minn AH, Pise-Masison CA, et al. Exenatide blocks JAK1- STAT1 in pancreatic beta cells. Metabolism. Jul 2007;56(7):915-918. 43. Kennedy J, Katsuta H, Jung MH, et al. Protective unfolded protein response in human pancreatic beta cells transplanted into mice. PLoS One. 2010;5(6):e11211. 44. Kutlu B, Kayali AG, Jung S, et al. Meta-analysis of gene expression in human pancreatic islets after in vitro expansion. Physiol Genomics. Sep 9 2009;39(1):72-81. 45. Marselli L, Thorne J, Ahn YB, et al. Gene expression of purified beta-cell tissue obtained from human pancreas with laser capture microdissection. J Clin Endocrinol Metab. Mar 2008;93(3):1046-1053. 84 46. Flamez D, Roland I, Berton A, et al. A genomic-based approach identifies FXYD domain containing ion transport regulator 2 (FXYD2)gammaa as a pancreatic beta cell-specific biomarker. Diabetologia. Jul 2010;53(7):1372- 1383. 47. Quayum N, Kutchma A, Sarkar SA, et al. GeneSpeed Beta Cell: an online genomics data repository and analysis resource tailored for the islet cell biologist. Exp Diabetes Res. 2008;2008:312060. 48. Liechti R, Csardi G, Bergmann S, et al. EuroDia: a beta-cell gene expression resource. Database (Oxford). 2010;2010:baq024. 49. Adewola AF, Lee D, Harvat T, et al. Microfluidic perifusion and imaging device for multi-parametric islet function assessment. Biomed Microdevices. Jun 2010;12(3):409-417. 50. Mohammed JS, Wang Y, Harvat TA, Oberholzer J, Eddington DT. Microfluidic device for multimodal characterization of pancreatic islets. Lab Chip. Jan 7 2009;9(1):97-106. 51. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protocols. 2008;4(1):44-57. 85 Most Pure Fraction A Entire Islet Preparation Factors of Interest %, Mean (±SD), or Median (min, max) Donor C Age (years) 0.1-47.5 >47.5-70.8 42% 58% ND B BMI (kg/m 2 ) 30.6 ± 7.4 ND B Fluids/electrolytes received No Yes 65% 35% ND B Liver SGPT Levels (u/L) >27 <27 46% 54% ND B Pancreas C ND B Fatty infiltration 1=None or patchy 2=Moderate or Heavy 15% 85% ND B Organ edema No Yes 88% 12% ND B Organ Intact No Yes 0% 100% ND B Pancreas Preservation Solution UW Alone HTK 46% 54% ND B Cold ischemia time (hrs) 7.1 ± 2.7 ND B Post-Purification Islet Assessment IEQ count 265,604 ± 113,797 334,237 ± 156,521 Purity (%) 90 (75,95) 79 (43,90) Viability (%) 95 ± 3 Not reported SI 2.3 ± 1.2 Not reported Endotoxin (EU/mL tissue volume) 0.04 (0.01, 0.16) ND B Islet Morphology Score (1-10 scale) <6 7-10 42% 58% A Islets for microarray analysis were sampled from this fraction B No Difference (ND) between first fraction and entire isolation C Variables selected based on statistically significant multivariate factors identified by Kaddis et al. to influence human islet isolation outcome Table 1. Characteristics of 26 Human Islet Preparations Used in This Study. 86 Figure 1. Assessing Human Islet Preparations Using Multivariable Logistic Regression Analysis. Model was derived using statistically significant multivariate factors, reported by Kaddis et al in 2010. Factors in grey increase the odds of a successful human islet isolation, while those in red negatively impact outcome. Panel A) continuous and B) categorical factors. Body Mass Index (BMI); Cold Ischemia (CI) time; Administration of Fluids/Electrolytes (Fluids/ELYT) during terminal hospital stay. 87 0 1 2 3 4 5 6 7 8 9 Group B: -8 to 4 Group A: 5 to 11 Total Points # Successful Isolations (>315,000 IEQs) Observed Expected overall p =0.0093 Figure 2. Agreement Between Composite Donor Score and Human Islet Isolation Outcomes. The observed (white) and expected (black) number of successful human islet isolations stratified by number of total points. The total number of isolations in group B was 18, and 8 in group A. Fisher’s exact test was used to examine the association between # of successful isolations and total points stratified by group. 88 A B Figure 3. PCA Mapping of 26 High Quality Human Islet Samples. Highlighted in blue and red are 2 bilateral clusters obtained from unsupervised Principal Component (PC) mapping A) using all 22,070 probes on the array and explained 45.9% of the total variation by PC#1 (26.0%), PC#2(10.8%), and PC#3 (9.1%), or B) using only 396 probes that were statistically significantly differentially expressed between cluster groups 1 and 2 (FDR p<0.05 and absolute fold change >1.5) that explained 68.8% of the total variation by PC#1 (54.5%), PC#2 (8.6%), and PC#3 (5.7%). 89 CHAPTER 4: PHYSIOLOGICAL GENETIC MODELING OF HUMAN BETA CELLS IN DIABETES: A NEW FRAMEWORK FOR SIMULATING COMPLEX DISEASE GENETICS CHAPTER 4 INTRODUCTION Mathematical modeling of pancreatic beta cells has been used for at least 40 years in the field of diabetes to unravel the underlying biology of the normal and disease state 1 . For example, Toffolo and Mari used C-peptide data to model beta-cell behavior 2,3 . Bertuzzi et al represented glucose-stimulated insulin release by modeling the dynamics of exocytosis 4 . Magombedze et al modeled the effects of cytotoxic and regulatory t-cells on the onset of type 1 diabetes 5 . These, and other recent examples 6 , illustrate the potential of mathematical modeling to assess beta cell function in-vivo. At the same time, major technological advances in the last 20 years have allowed for unprecedented access to expanded coverage of the entire genome 7,8 . The benefits of this data have been demonstrated using genome- wide association (GWA) studies that identified common genetic variation in polygenic diseases such as prostate 9 and breast 10 cancer, type 1 11 and 2 12-14 diabetes mellitus, coronary heart disease 15 , and others 16 . The drawbacks include the lack of data analysis methodologies and difficulty of interpretation 17-20 . 90 To address this problem, pathway-based analysis methods aimed at incorporating prior biological knowledge may be a promising approach to understanding the functional importance of and dependency between genes 21,22 . However, in order to test and investigate the performance of statistical genetic approaches 23-25 , simulation models that reflect the underlying complex genetic and biologic interactions are requisite. The traditional approach for simulating genetic data to test statistical methodology assumes an underlying genetic model using a penetrance function to relate genetic variation to disease status or variation in a quantitative trait (QT). To simulate a binary outcome, the penetrance function takes a probabilistic form that usually incorporates population-level-based information, such as disease prevalence 26 . For QTs, the penetrance function might take the form of a variance components model 27,28 . Both these approaches rely solely on a stochastic process to characterize the link between genetic variation and the outcome. Expansion of these simulations to gene-gene or gene-environment joint action typically oversimplifies to the inclusion of statistical interaction terms specifying effects. These approaches have been used, in part, because it has been difficult to create appropriate simulations in the absence of exact knowledge of the relations among genes and environment in vivo. The lack of a method to simulate representative data may, in part, explain why many statistical genetic methods tested under traditional statistical genetic modeling conditions have performed well in simulation, but less optimally on “real data”. Therefore, the underlying biology needs to be 91 appropriately modeled to evaluate novel statistical genetic methods designed to better capture or incorporate complex biological relations. We approach this problem from a different perspective and introduce a new simulation framework we call “Physiological Genetic Modeling” (PGM). PGM attempts to bridge the gap between genetic variation and complex phenotypes through the use of biologically based mathematical modeling. PGM consists of the following steps: 1) identification and selection of a phenotype of interest, 2) construction of a mathematical model representing an underlying physiological state, 3) acquisition of candidate genes attributable to the defined phenotype, 4) assignment of genes to model parameters, and 5) alteration of model parameters to test for physiological relationships. This approach allows us to stochastically simulate the effect of single or multiple genes without directly assuming an overly simplified statistical representation by using a biologically based model that should manifest the integrated effect of the entire regulatory and feedback processes that may influence the trait of interest and not just genetic variation. The study of glucoregulation in the pathogenesis of type 2 diabetes mellitus (T2DM) provides unique opportunities to understand these relationships. 92 CHAPTER 4 MATERIALS AND METHODS 1.1. Model Development 1.1.1. Compartmentalization of Glucose Regulation Glucose regulation is a dynamic process known to involve the action of several hormones on different organs and tissues 29 . We therefore began by adopting a single kinetic model of the glucose regulatory system, described by Huckberg et al 30 . In brief, a single representation of glucose regulation was idealized as compartments controlled by inputs and outputs, described by 5 differential equations. 1.1.2. Addition of Genetic Component Our knowledge of candidate genes is used as a guide to specify which corresponding parameters within the compartmental model are conceivably altered by candidate gene variants. Genotype-specific model parameters for simulated individuals are derived by utilizing a population distribution of the model parameter of interest, defining biologically plausible values for that parameter, and incorporating reported allele frequencies for candidate genes of interest. 1.2. Simulation Scenarios Simulation studies were carried out to assess the performance and sensitivity of the proposed compartmental model. For all simulations, model parameter values were derived from average values reported in the human literature 31-33 . Where 93 human data were not available, parameter values from the animal literature were used 34 . Unless noted elsewhere, model parameter values were fixed and held constant as indicated in supplemental Table I. Specific simulation scenarios are outlined below. 1.2.1. Scenario 1: Model Consistency with Published Data The first series of simulations were designed to assess whether our proposed compartmental model would behave in a manner consistent with in vivo human observations. We therefore generated Oral Glucose Tolerance Test (OGTT) profiles and compared the simulated glucose time courses with published data. The OGTT is one of several well characterized clinical protocols 35 used to evaluate the body’s ability to remove glucose from the bloodstream. For the OGTT, we compared our model predictions with those obtained by Shapiro et al. 36 who performed 240 minute OGTTs in 9 human subjects with normal glucose tolerance. The average pre-hepatic insulin secretory profile, determined by C-peptide deconvolution, was used as a known input to the model (S(t)). We empirically derived a glucose absorption pattern for a 70 kg individual based on the assumption of a 75 g oral glucose load absorbed over the course of the 240- minute experiment. 1.2.2. Scenario 2: Effects of Population Parameter Variation The second series of simulations were designed to assess if and which population parameters influence a) the OGTT profile, and b) insulin sensitivity. 94 This was done by increasing and decreasing the mean population value of each parameter by 25%, 50%, 75%, and 100% for 1 individual. After each simulation, we categorized the individual as normal glucose tolerant (NGT), impaired glucose tolerant (IGT), or diabetic, and calculated both the Matsudo and Stumvoll indices of insulin sensitivity. 1.3. Data Analysis For simulations that included genetic effects, we randomly selected genotype- specific parameter values to generate one replicate of 1000 individuals. Each individual was then classified as being NGT, IGT, or diabetic using the standard clinical diagnostic criteria of a 2-hour OGTT glucose concentration >140 mg/dl (7.8 mM) 37 . 1.4. Software Packages Simulations and subsequent data analyses were performed using the R software environment for statistical computing freely available at http://www.r-project.org/) with the odesolve, lattice, hmisc, scatterplot3d, reshape, and mass libraries. CHAPTER 4 PRELIMINARY RESULTS Simulation Scenario Assessments Scenario 1 was constructed to determine if model behavior was consistent with in vivo physiology. Model-predicted glucose and insulin time courses were superimposed upon average literature values for experimentally determined 95 OGTT profiles. Figure 1 demonstrates that the model was able to mimic published average glucose and insulin concentration profiles with remarkable fidelity, but that deviations were also observed. These differences can be explained by the fact that model parameters were fixed to average literature values and not optimized for this particular data. Scenario 2 was designed to identify the population parameters with the largest influence on OGTT profiles. Figure 2 demonstrated that changes to K 01 had the largest influence on the OGTT profile up to 100 minutes. Beyond that, variation in p 3 , K g , and K 21 created the greatest changes in the OGTT profile. A 75% increase in the mean population value of p 3 resulted in IGT, as did a 75% increase in kg, a 50% decrease in K 21 , and a 100% increase in K 01 . Diabetes was seen when the mean population value of p 3 and K g were increased by 100%, K 21 was increased by 75% and 100%. CHAPTER 4 DISCUSSION AND FUTURE WORK Characterization of a disease often begins with the phenotypic appearance of undesirable traits followed by intense genotypic investigation for the causative gene(s). With the rare exception of monogenic diseases, such analyses can result in incomplete and complex phenotype-genotype relationships that are frequently confounded by environmental factors. In this preliminary report, our 96 aim was to demonstrate the potential utility of PGM using T2DM as our phenotype of interest. Initial work focused on reproducing the mathematical model of Huckberg et al 30 to obtain OGTT profiles that mimicked real-world experimental data. The OGTT was used because it has been shown to be a reliable measure of beta cell function, if performed appropriately 38 ; however, recent works have called into the use of OGTT-derived indices of insulin sensitivity 30,39 . We found that our simulations performed remarkably well when compared to experimental data reported in the literature 36,40 . The model used accounts for many, but not all, of the factors that affect glucose and insulin. For example, the model currently does not account for the known effects of free fatty acids on glucose turnover and β-cell function 41,42 . Thus, improvements can be made to the model. In fact, as greater biologic knowledge is gained, new findings can be incorporated into our model system to provide a more accurate representation of the system. We are now working to better understand how different population parameters from the compartmental model affect different areas of the OGTT curve. Observed were dramatic early effects when K 01 was altered and late effects when p 3 , K g , and K 21 were altered. The interaction of different parameters on the OGTT profile is currently being examined. 97 Our next step will be to demonstrate how genetic variation can be incorporated into the compartmental modeling framework to simulate the effect of single or multiple genes on beta-cell function. This is accomplished by modeling the distribution of parameter values for the simulated individuals as a linear function of an overall mean (γ 0 ) and a genotype-specific effect (γ X ). We will use a single parameter, single gene model to simulate individual-specific values, p i μ , for a specific parameter, p, determined by a single underlying gene X. i Xi p X p p i G ε γ γ μ + + = 0 (1) ) , 0 ( ~ where , 2 p i N σ ε Here, G Xi is a variable indicating the number of variant alleles at a suspected gene X for individual i that influences parameter p. The genotypes for X are assumed to be in Hardy-Weinberg equilibrium. The allele frequencies are pre- specified to reflect knowledge of a particular gene known to influence the biologic function of a parameter in the compartmental model, or may be determined by the experimental conditions, such as an examination of the range in allele frequency on the level of impact of genetic variation. Here, γ 0 and γ X are pre- specified. However, we constrain the frequency-weighted average of the genotype-specific values (for a single, additive gene model, p p G X 0 0 γ μ = = , p X p p G X γ γ μ + = = 0 0 , p X p p G X γ γ μ 2 0 0 + = = ) to equal the population average, p Pop μ . That is, p G X p G X p G X p Pop X X X G G G 2 1 0 ) 2 Pr( ) 1 Pr( ) 0 Pr( = = = = + = + = = μ μ μ μ . (2) 98 Values for the mean population, p Pop μ , and standard deviation, p σ , are obtained from previously reported values in the literature of the particular parameter of interest or by empirical distributions of the model parameters from clinical data. As a further constraint, we limited the genotype-specific mean effects to be within a “biologically” plausible range defined by the 25 th - and 75 th -percentile of the parameter distribution observed from clinical data. That is, p Min p G X μ μ ≥ =0 and p Max p G X μ μ ≤ =2 , where p Min μ and p Max μ correspond to the 25 th - and 75 th -percentile values, respectively. We define the limit of this range in which p Min p G X μ μ = =0 and p Max p G X μ μ = =2 as having a single gene effect explaining 100% of the plausible effect. In essence, the ultimate impact of these multiple constraints based upon a priori knowledge of gene frequencies and plausible parameter distributions puts a biological constraint on the impact that a single gene may have on a single parameter in our compartmental model. CHAPTER 4 CONCLUSION In summary, we introduce the concept of PGM, a biologically based framework to simulate the effect of genetic variation on phenotypes. This framework replaces the traditional penetrance function with a biologically based mathematical model to directly incorporate biologic knowledge into genetic simulations. We illustrated the potential of PGM by introducing a compartmental model of glucoregulation and an overall framework for incorporating the effect of genetic variation into the 99 model. Thus, the PGM approach represents a unique blend of statistical genetic and compartment modeling methodology, and provides a novel simulation tool for statistical genetic methods designed to untangle the genetic architecture of complex diseases. 100 CHAPTER 4 REFERENCES 1. Boutayeb A, Chetouani A. A critical review of mathematical models and data used in diabetology. Biomed Eng Online. 2006;5:43. 2. Toffolo G, Breda E, Cavaghan MK, Ehrmann DA, Polonsky KS, Cobelli C. Quantitative indexes of Î²-cell function during graded up&down glucose infusion from C-peptide minimal models. American Journal of Physiology - Endocrinology And Metabolism. January 1, 2001;280(1):E2-E10. 3. Mari A, Camastra S, Toschi E, et al. A model for glucose control of insulin secretion during 24 h of free living. Diabetes. February 1, 2001;50(suppl 1):S164. 4. Bertuzzi A, Salinari S, Mingrone G. Insulin granule trafficking in Î²-cells: mathematical model of glucose-induced insulin secretion. American Journal of Physiology - Endocrinology And Metabolism. July 1, 2007; 293(1):E396-E409. 5. Magombedze G, Nduru P, Bhunu CP, Mushayabasa S. Mathematical modelling of immune regulation of type 1 diabetes. Biosystems. Nov-Dec 2010;102(2-3):88-98. 6. Soria B, Tuduri E, Gonzalez A, et al. Pancreatic islet cells: a model for calcium-dependent peptide release. HFSP journal. Apr 2010;4(2):52-60. 7. Kahvejian A, Quackenbush J, Thompson JF. What would you do if you could sequence everything? Nat Biotech. 2008;26(10):1125-1133. 8. Mardis ER. The impact of next-generation sequencing technology on genetics. Trends Genet. Mar 2008;24(3):133-141. 9. Gudmundsson J, Sulem P, Manolescu A, et al. Genome-wide association study identifies a second prostate cancer susceptibility variant at 8q24. Nature Genetics. May 2007;39(5):631-637. 101 10. Easton DF, Pooley KA, Dunning AM, et al. Genome-wide association study identifies novel breast cancer susceptibility loci. Nature. Jun 2007;447(7148):1087-U1087. 11. Todd JA, Walker NM, Cooper JD, et al. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet. 2007;39(7):857-864. 12. Scott LJ, Mohlke KL, Bonnycastle LL, et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science. Jun 2007;316(5829):1341-1345. 13. Sladek R, Rocheleau G, Rung J, et al. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature. Feb 2007;445(7130):881-885. 14. Saxena R, Voight BF, Lyssenko V, et al. Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science. Jun 2007;316(5829):1331-1336. 15. McPherson R, Pertsemlidis A, Kavaslar N, et al. A common allele on chromosome 9 associated with coronary heart disease. Science. Jun 2007;316(5830):1488-1491. 16. Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature. Jun 7 2007;447(7145):661-678. 17. Hunter DJ, Kraft P. Drinking from the Fire Hose -- Statistical Issues in Genomewide Association Studies. N Engl J Med. July 18, 2007; 357(5):436-9. 18. Christensen K, Murray JC. What Genome-wide Association Studies Can Do for Medicine. N Engl J Med. March 15, 2007;356(11):1094-1097. 19. Thomas DC. Are We Ready for Genome-wide Association Studies? Cancer Epidemiol Biomarkers Prev. April 1, 2006;15(4):595-598. 102 20. Pop M, Salzberg SL. Bioinformatics challenges of new sequencing technology. Trends in Genetics. 2008;24(3):142-149. 21. Curtis RK, Oresic M, Vidal-Puig A. Pathways to the analysis of microarray data. Trends in Biotechnology. 2005;23(8):429-435. 22. Wang K, Li M, Bucan M. Pathwayâ€Based Approaches for Analysis of Genomewide Association Studies. The American Journal of Human Genetics. 2007;81(6):1278-1283. 23. Hung RJ, Brennan P, Malaveille C, et al. Using hierarchical modeling in genetic association studies with multiple markers: application to a case- control study of bladder cancer. Cancer Epidemiol Biomarkers Prev. Jun 2004;13(6):1013-1021. 24. Conti DV, Cortessis V, Molitor J, Thomas DC. Bayesian modeling of complex metabolic pathways. Hum Hered. 2003;56(1-3):83-93. 25. Cortessis V, Thomas DC. Toxicokinetic genetics: an approach to gene- environment and gene-gene interactions in complex metabolic pathways. IARC Sci Publ. 2004(157):127-150. 26. Risch N. Linkage strategies for genetically complex traits. I. Multilocus models. Am J Hum Genet. Feb 1990;46(2):222-228. 27. Fisher RA. The correlation between relatives on the supposition of Mendelian inheritance. Trans Royal Soc Edinburgh. 1918;52(399-433). 28. Hopper JL. Variance components for statistical genetics: applications in medical research to characteristics related to human diseases and health. Stat Methods Med Res. 1993;2(3):199-223. 29. Kruger DF, Martin CL, Sadler CE. New Insights Into Glucose Regulation. The Diabetes Educator. March 1, 2006;32(2):221-228. 103 30. Hucking K, Watanabe RM, Stefanovski D, Bergman RN. OGTT-derived Measures of Insulin Sensitivity Are Confounded by Factors Other Than Insulin Sensitivity Itself. Obesity. 2008;16(8):1938-1945. 31. Cobelli C, Toffolo G, Ferrannini E. A model of glucose kinetics and their control by insulin, compartmental and noncompartmental approaches. Math Biosci. 1984;72:291-315. 32. Ferrannini E, Smith JD, Cobelli C, Toffolo G, Pilo A, DeFronzo RA. Effect of insulin on the distribution and disposition of glucose in man. J Clin Invest. Jul 1985;76(1):357-364. 33. Watanabe RM, Steil GM, Bergman RN. Critical evaluation of the combined model approach for estimation of prehepatic insulin secretion. Am J Physiol. Jan 1998;274(1 Pt 1):E172-183. 34. Ader M, Ni TC, Bergman RN. Glucose effectiveness assessed under dynamic and steady state conditions. Comparability of uptake versus production components. J Clin Invest. Mar 15 1997;99(6):1187-1199. 35. Trout KK, Homko C, Tkacs NC. Methods of Measuring Insulin Sensitivity. Biol Res Nurs. April 1, 2007;8(4):305-318. 36. Shapiro ET, Tillil H, Rubenstein AH, Polonsky KS. Peripheral insulin parallels changes in insulin secretion more closely than C-peptide after bolus intravenous glucose administration. J Clin Endocrinol Metab. Nov 1988;67(5):1094-1099. 37. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. Jul 1997;20(7):1183-1197. 38. Mari A, Ferrannini E. β-cell function assessment from modelling of oral tests: an effective approach. Diabetes, Obesity and Metabolism. 2008;10:77-87. 104 39. Buchanan TA, Watanabe RM, Xiang AH. Limitations in Surrogate Measures of Insulin Resistance. J Clin Endocrinol Metab. November 1, 2010;95(11):4874-4876. 40. Watanabe RM, Volund A, Roy S, Bergman RN. Prehepatic beta-cell secretion during the intravenous glucose tolerance test in humans: application of a combined model of insulin and C-peptide kinetics. J Clin Endocrinol Metab. Oct 1989;69(4):790-797. 41. Liu YQ, Tornheim K, Leahy JL. Fatty acid-induced beta cell hypersensitivity to glucose. Increased phosphofructokinase activity and lowered glucose-6-phosphate content. J Clin Invest. May 1 1998;101(9):1870-1875. 42. Rebrin K, Steil GM, Getty L, Bergman RN. Free fatty acid as a link in the regulation of hepatic glucose output by peripheral insulin. Diabetes. Sep 1995;44(9):1038-1045. 105 Model Parameter Definition Units Value K 21 Fractional glucose disappearance from G 2 to G 1 min -1 0.043 K 12 Fractional glucose disappearance from G 1 to G 2 min -1 0.059 K 01 Fractional glucose disappearance from G 1 min -1 0.012 K 02 Fractional glucose disappearance from G 2 min -1 0.0055 V G Glucose distribution volume L 114 HGO 0 Hepatic glucose output at zero glucose mg/min 274.4 K L Fractional disappearance from G 2 to G 1min -1 0.00917 p2 Fractional disappearance from G 2 to G 1min -1 0.02 p3 Fractional disappearance from G 2 to G 1 min -2 per U/ml 0.00001 F X Hepatic fraction of the remote insulin effect fraction 0.01 F Fractional hepatic insulin extraction fraction 0.5 K I Fractional insulin disappearance min -1 0.175 V I Insulin distribution volume L 4.55 K G -cell sensitivity to glucose fraction per mg/dl 0.009 Supplemental Table 1. Assumed Model Parameters. Values taken from the literature, as detailed in materials and methods. 106 Figure 1. OGTT Simulation Results. Average glucose (bottom) and insulin (top) profiles from the OGTT. Solid red line represents the model-predicted glucose and insulin profiles from the compartmental model when fixing model parameters as described in the methods, while blue dots represent published data (see main text). 107 Figure 2. Variation of Selected Population Parameters from the Glucoregulatory Compartmental Model. 108 Computational Statistics & DataAnalysis 51 (2007) 5494–5506 www.elsevier.com/locate/csda A matching algorithm for the distribution of human pancreatic islets Dajun Qian ∗ , John Kaddis, Joyce C. Niland Division of Information Sciences and Administrative and Bioinformatics Coordinating Center for the Islet Cell Resource Center Consortium, City of Hope National Medical Center, Duarte, CA, USA Received 16 December 2006; accepted 28 February 2007 Available online 7 March 2007 Abstract Thesuccessofhumanpancreaticislettransplantationinasubsetoftype1diabeticpatientshasledtoanincreaseddemandforthis tissue in both clinical and basic research, yet the availability of such preparations is limited and the quality highly variable. Under the current process of islet distribution for basic science experimentation nationwide, specialized laboratories attempt to distribute islets to one or more scientists based on a list of known investigators. This local decision making (LDM) process has been found to be ineffective and suboptimal. To alleviate these problems, a computerized Matching Algorithm for Islet Distribution (MAID) was developed to better match the functional, morphological, and quality characteristics of islet preparations to the criteria desired by basic research laboratories, i.e., requesters. The algorithm searches for an optimal combination of requesters using detailed screening, sorting, and search procedures. When applied to a data set of 68 human islet preparations distributed by the Islet Cell Resource (ICR) Center Consortium, MAID reduced the number of requesters that (a) did not receive any islets, and (b) received mis-matched shipments. These results suggest that MAID is an improved more efficient approach to the centralized distribution of human islets within a consortium setting. © 2007 Elsevier B.V.All rights reserved. Keywords: Islet distribution; Matching algorithm; Exhaustive search; Space reduction; Importance sampling 1. Introduction The inability to respond to or produce insulin results in diabetes and leads to potentially severe secondary compli- cations in individuals with this disease (The Diabetes Control and ComplicationsTrial Research Group, 1993; Nathan, 1993; Pinto et al., 2004; Taylor et al., 2004; Weir and Bonner-Weir, 2004). The malfunction and/or destruction of the pancreatic beta cell play a key role in the progression of diabetes. Beta cells are responsible for the production and storage of insulin, the hormone that controls blood glucose levels, and are one of four main cell types in the pancreas clustered together to form what are known as “islets of Langerhans” (Hardikar, 2004;Ahren and Taborsky, 2003). Emerging strategies for the treatment of diabetes include islet replacement and/or renewal therapies (Ramiya and Schatz, 2004; Trucco, 2005). Islet replacement strategies for a subset of type 1 diabetic patients are predominately ∗ Corresponding author. Tel.: +16262564673x62685; fax: +16264717106. E-mail addresses: dqian@coh.org (D. Qian), jkaddis@coh.org (J. Kaddis), jniland@coh.org (J.C. Niland). 0167-9473/$-see front matter © 2007 Elsevier B.V.All rights reserved. doi:10.1016/j.csda.2007.02.030 CHAPTER 5: IMPROVING AVAILABILITY OF ISLET TRANSPLANTATION IN TYPE 1 DIABETES Section 5.1: A Matching Algorithm for the Distribution of Human Pancreatic Islets 109 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 5495 based on the use of pancreata from a limited supply of human cadaveric donors (Nanji and Shapiro, 2006). The use of specializedpancreasprocessingfacilitiesisrequiredinordertogeneratehighqualityhumanisletsderivedfromdonated cadaveric pancreata. Organ procurement, islet isolation, final batch preparations, and product release characterization are highly complex procedures involving a chain of intricate processing steps. Relatively few centers around the world arecapableofprovidinghumanisletsforclinicaltransplantationpurposes,yettheneedforsuchisletsisgrowingrapidly. It is estimated that between 1999 through the middle 2004, 43 institutions worldwide processed human pancreata for islet transplantation in over 470 recipients (Gaglia et al., 2005). As a result of the improved outcomes demonstrated in patients receiving islet transplants (Shapiro et al., 2000), the need for human islets has increased in both clinical and basic research settings (Berney et al., 2005; Ridgway etal.,2005).Withthehopethatimprovementsinhumanisletisolationandtransplantationwillleadtobettertreatments and durable insulin independence for type 1 diabetes, the National Islet Cell Resource Center (ICR) consortium was established in 2001 (Knazek, 2002). The ICRs are a group of academic laboratories across the United States charged with providing human pancreatic islets for clinical and basic research purposes. One of the key responsibilities of the ICR Administrative and Bioinformatics Coordinating Center (ABCC) is to coordinate the islet distribution activities of the ICR basic science Human Islet Distribution program.The existing islet distributionapproachallowseachproducer,i.e.,ICRlaboratory,tomakelocaldecisionsaboutwhoreceivesisletsbased on a centrally provided ABCC list of requesters, i.e., ICR approved investigators. Under this local decision making (LDM) approach, producers are often limited in their ability to closely match the characteristics of available islets to the desired criteria described in requesters’ applications. Thus, requesters with identical needs may not receive a comparable number of islets, if any at all. The LDM approach, however, is used because it is often the quickest way to distribute islets in the least amount of time, ideally within 24h of production, thus minimizing the degradation of functional viability. The problem of equitably allocating a limited supply of islets to multiple requesters in a limited amount of time can be viewed as a problem in combinatorial optimization (Nemhauser and Wolsey, 1988). The classic binary knapsack problemseekstofindanoptimalsetofobjectsthatwillfitintoaknapsackoffixedsizeandvolume,whilesimultaneously considering the unique attributes of each object (Martello and Toth, 1990). However, the problem of matching islet offers differs somewhat in that the parameters of a requesting investigator vary, rather than representing a fixed set of known attributes. Moreover,thisvariationisoneofseveralfactorstoconsiderwhenattemptingtomatchisletsgeneratedbyaproducer toapprovedrequesters.Notonlydothecharacteristicsofdesiredisletsdifferbasedontheintendedresearchpurposeof eachrequester,theshortageoforgandonationandtheunder-utilizationofpancreataalsocreateanunpredictablesupply of human islets that often falls short of demand at any point in time. Furthermore, the variable quality of available organs and the differing pancreas processing procedures used by a given producer of islets cause the resulting islets to vary in amount, purity and viability. To preserve maximum islet quality, islets must be shipped immediately following thecompletionofpancreaticprocessing,usuallywithin24h,tominimizedegradationoftheisletsovertime.Therefore the proximity of requesters to producers must be taken into account, along with considerations such as study start and end dates, and frequency, quality, and quantity of desired islets for experimentation. A major challenge is that such an islet distribution algorithm should not only be computationally efficient, but also aidintheregulatoryandfunctionaloversightactivitiesoftheICRconsortium.Theseincludetheabilitytodecreasethe priority of a requester who has previously rejected a valid islet offer, and increase the priority of certain investigators conducting research deemed critical by the ICR funding agencies.The preferences of a requester to receive islets from a selected producer can be based on the proximity, ongoing collaborative relationship, and/or the perceived quality of islets by the producer. In addition to these fairly subjective criteria, the algorithm must take into account the objective criteriaofisletpurity(minimumandideal),viability(minimumandideal),quantity(minimumandideal)andfrequency (for the duration of the scientific project) of shipments for all approved studies. To accommodate all of these objective and subjective criteria, a matching algorithm for islet distribution (MAID) was developed and tested as a tool to optimize the islet distribution activities in a practical setting. In this paper we describe the development and testing of this islet matching algorithm.We first apply MAID using actual data to model ICRDistributionProgrambehavior,includingICRproductionactivityoverafixedtimeperiod,andshipmentsofislets to approved investigators during this timeframe. We then apply the algorithm to simulated data, varying certain key conditions anticipated to impact the ability to successfully match islet cell offers from producers to the pool of waiting islet requesters. 110 5496 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 2. Methods 2.1. MAID development We consider an islet distribution program to consist of I producers and J requesters. Each islet isolation, labeled u,is generated from a donated cadaveric pancreas organ and characterized by various parameters, including (a) producer i generatingtheislets,(b)date tonwhichisletsareisolated,(c)numberofavailableislets Q,fordistributionmeasuredin unitofisletequivalent(IEQ),and(d)isletpurityPandviabilityV,bothmeasuredaspercentagesfrom0%to100%.(We omit the index u in each of the above isolation parameters for simplicity.) Each requester j must submit an application for islets prior to approval, and the following information within the application is used to classify the requester: (a) date of application approval t j,apv , (b) supplier preference vector to all producers {y j (i)|1iI}, (c) minimum daysbetweenisletshipmentsx j ,(d)minimumandidealnumbersofIEQspershipmentq j,min andq j,ideal ,(e)minimum and ideal purity values p j,min and p j,ideal , and (f) minimum and ideal viability values v j,min and v j,ideal . TheMAIDalgorithmusesthesevariablestogenerateanoptimalofferlist(OOL)ofrequestersviathreecomputational components:(a)ascreeninganalysistoidentifyallqualifiedrequesters,(b)asortinganalysistoscoreandrankqualified requesters, and (c) a search procedure for selection of OOL under required conditions. Requesters on the OOL are contacted to confirm acceptance of the islet offer. If a requester rejects the offered islets, those islets will be used to generateanalternativeOOL.Fig.1providesanalgorithmschemaofMAIDunderaconsortiumsetting,andthesections below describe each computational component in more details. 2.1.1. Screening analysis for qualified requesters When a batch of islets, characterized by parameters u, i, t, Q, P andV, are generated and available for distribution, a screeninganalysisisfirstperformedtoidentifyalistofallqualifiedrequesters.Specifically,arequester j ∈{1,...,J} is qualified if all of the following criteria are met: 1. For a requester j who has received at least 1 islet shipment, the time since last shipment should be at least the minimum days between shipments desired by the requester, i.e., t − t j,LS x j , where t j,LS is the date of last shipment to requester j. 2. Producer i is an acceptable supplier to requester j, i.e., y j (i) = 1. 3. The number of available IEQs is at least the minimum number of IEQs per shipment desired by requester j, i.e., Qq j,min . Screening: Apply screening criteria to identify qualified requesters Sorting: Score & rank qualified requesters in descending order of priority scores Selection: Select optimal islet offers using a semi-exhaustive search procedure Send offers Ship islets accepted Rerun MAID for rejected IEQs rejected Input Data: Available islets, criteria for all requesters, and history of previous offers Fig. 1.Algorithm schema of MAID. 111 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 5497 4. The purity and viability of available islets are not lower than the corresponding minimal acceptable values, respec- tively, i.e., Pp j,min and Vv j,min . For ease of notation, all qualified requesters {j 1 ,...,j K }⊆{1,...,J} identified in the screening analysis will be indexed as k = 1,...,K in subsequent analyses. 2.1.2. Sorting requesters by priority scores We next compute a priority score s k for each qualified requester k using a form of s k = w k (a k b k c k d k ) + e k , (1) and sort the K requesters in descending order of their scores to generate an ordered list of candidates to be included in the OOL. In order to appropriately incorporate the priorities mandated by the ICR Steering Committee and funding agencies, the coefficient terms in Eq. (1) are determined as follows: 1. w k =waiting days of requester k, in addition to the minimum days between shipments as stated in requester’s application. Specifically, w k equals t −t k,apv if no offer has ever been made to requester k,or t −(t k,LS +x k ) if the last offer on date t k,LS was accepted and shipped, or t − t k,LO if the last offer on date t k,LO was rejected. 2. a k =1.1ifthedistancebetweenproducerandrequesterallowssamedaydelivery,or1ifnextdaydeliveryisrequired. 3. b k = 1.1 if the requester has peer reviewed funding support, or 1 if no such funding is present. 4. c k =1.5 if the purity of available islets is within 5% of the ideal value desired by the requester, or 1 if the difference is larger. 5. d k =1.5iftheviabilityofavailableisletsiswithin5%oftheidealvaluedesiredbytherequester,or1ifthedifference is larger. 6. e k =max(w 1 ,...,w K )(1.1 ×1.1 ×1.5 ×1.5) if the requester has preferred priority status to receive islets, or 0 if the requester has standard priority status. The rationale behind Eq. (1) reflects several considerations. First, waiting time is considered as a baseline value in quantifying the priority scores among all qualified requesters. The choice of w k = t − t k,LO indicates that the waiting time will be reset to 0 at the time of rejecting a matched offer.As a side note, the dates of last shipment and last offer are equal (i.e., t k,LS =t k,LO for requester k) if the last offer has been accepted and shipped, or the date of last shipment is earlier (i.e., t k,LS <t k,LO ) if the last offer has been rejected. Second, coefficients a k and b k for delivery distance and fundingstatusrespectivelyarechosentoincreasethescoresby10%underfavorableconditions.Thesetwocoefficients are found to have little or no influence on the resulting proportion of unmatched islets.Third, coefficients c k =1.5 and d k =1.5 for purity and viability matches respectively, are chosen to minimize the proportion of unmatched islets.And fourth, the preferred priority status is considered as a dominant factor that allows a small portion of requesters to have the highest priority scores when so designated by the funding agencies. 2.1.3. Search for OOL ThesearchfortheOOLevaluatesallpossiblesubsetsamongtheKqualifiedrequestersobtainedinscreeninganalysis, toidentifytheoptimaloneundergivenconditions.Thescreeninganalysisgenerallyyieldsmultiplepotentialrequesters, and the final OOL consists of up to 10 requesters who will be offered IEQs from a given isolation in order to utilize all islets. We search the OOL under the following three conditions: (A) the number of requesters in the optimal solution is less than or equal to N max , where N max = 10 is the default value and 1N max 9 are the alternative choices when needed, (B) the number of unmatched islets is minimized, and (C) the average priority score is maximized. When K is small, we identify the optimal solution based on an exhaustive search.Alternatively, when K is large, we search the optimal solution using a semi-exhaustive procedure, coupled with importance sampling and extended local search for improved search performance. Several notations are needed before describing the proposed search procedure. Let={1,...,K} denote the full list of K qualified requesters sorted in descending order of their priority scores, N ={1,...,N|NK} denote a subset of with the N top-score requesters in descending order of scores, and (N) ={k 1 ,...,k N |NK} denote a 112 5498 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 subset of with any N requesters in descending order of scores. Of note, the number of matched IEQs in any offer list (N) must be equal to or less than the number of available IEQs, Q. We begin by describing the search procedure under a simple scenario when the number of qualified requesters K is small. Let M = max(10,N max ) denote a small number for which an exhaustive search is feasible. When KM,an exhaustive search in the full space of 2 K − 1 subsets is performed to retain any subsets (N) such that NN max and (N) results in 0 or minimal unmatched islets.The OOL is chosen to be the retained (N) with maximal average score, or a random one when multiple retained (N) have the same maximal average score. We now describe the search procedure under an alternative and general scenario when the number of qualified requestersislarge(i.e.,K>M),andanexhaustivesearchamonganexponentialnumberofsubsetsiscomputationally infeasible. In this case, a semi-exhaustive search procedure, coupled with importance sampling and extended local search, is used to identify the OOL under the required conditions (A)–(C). The search procedure can be described in three steps as follows. Step 1: Reduce search space to a manageable size. We obtain a reduced search space M containing M requesters forthesearchofOOLintwomutuallyexclusivesituations.Let Q( N )= k∈N q k,ideal and Q( (N) )= k∈ (N) q k,ideal denote the sums of ideal IEQs per shipment desired by requesters in N and (N) , respectively. In a common situation that K>M and Q( M )Q, the reduced search space is chosen to be all subsets in M top-score requesters, i.e., M ={ (N) | (N) ⊆ M ,NM},thengotoStep2.Inpractice,theconditionQ( M )Qisexpectedtoholdinmore than 90% of occasions because the number of IEQs Q generated from a cadaveric pancreas organ is rarely sufficient for offering islets to 10 or more requesters. In a less common situation whereK>M and Q( M )<Q (e.g., the M top-score requesters all desire very small numbersofIEQspershipment),animportancesamplingroutineisemployedtofindareducedspace M ={ (N) | (N) ⊆ ∗ (M) ,NM}, where ∗ (M) is a set of M selected requesters with the sum of their ideal IEQs per shipment equal to or greater than the number of available IEQs (i.e., Q( ∗ (M) )Q). To do this, we choose an initial (M) to be the M top-score requesters M ={1,...,M}, and then update (M) iteratively by substituting one requester k ∈ (M) with another one l ∈ ( − (M) ) based on assigned sampling probabilities {f k ,k ∈ (M) } and {g l ,l ∈ ( − (M) )}, respectively.Numerically,wechoosef k inproportionto2 −(M−k+1) foreachk ∈ (M) andg l inproportiontos l ×q l,ideal for each l ∈ ( − (M) ), where k in quantity 2 −(M−k+1) represents that requester k has the kth largest score in (M) . Of note, these sampling probabilities are chosen to force (M) to favor requesters with both higher scores and larger requested islet amounts per shipment. If Q( (M) )Q is reached within 1000 iterations, we choose ∗ (M) = (M) to form a reduced search space M , and go to Step 2. Otherwise, we choose (M) at iteration 1000 as a temporary offer list, and go to Step 3. Step2: Exhaustive search in reduced search space M .Ifareducedsearchspace M isfoundin Step1,weperform an exhaustive search by evaluating a maximum of 2 M − 1 subsets in M and choose an optimal one ∗ under the required conditions (A)–(C). If the identified ∗ results in zero unmatched islets, we choose the OOL equal to ∗ , and stop the search procedure. Otherwise, if ∗ results in non-zero unmatched islets, we choose ∗ as a temporary offer list, and go to Step 3. Step 3: Extended local search when needed. Occasionally, we may either fail in Step 1 in finding a reduced search space,orfailinStep2inidentifyinganoptimalsubsetwithzerounmatchedislets.Insuchcases,anextendedlocalsearch in the full space of all K qualified requesters is performed to identify an OOL under the required conditions (A)–(C). To do this, we start with a temporary offer list (N) containing NM requesters as obtained in Step 1, or Step 2, when appropriate.Wethensubstituteonerequesteratatimein (N) withpossiblecombinationsof1toM −N +1requesters in ( − (N) ), retain all subsets with minimal unmatched islets, and update the offer list (N) when substitutions are done for all requesters in (N) and conditional on minimal unmatched islets. If at least one subset has zero unmatched islets, we choose the OOL to be the one with maximum average score, and stop the search procedure. Otherwise, if no subset has zero unmatched islets, we choose the OOL to be the one with minimum unmatched islets, and stop the search procedure. Of note, one loop by substituting each requester once in a corresponding offer list (N) is generally sufficient in the local search analysis above, because experimental evaluations indicated no change for the resulting OOL with additional loops. Infrequently, the available islets are not acceptable to any requesters (i.e., K = 0), or the quantity of available islets is more than the sum of ideal amounts per shipment desired by all qualified requesters or the N max ones requesting the largest amounts per shipment. In such cases, all or a portion of islets cannot be matched, and those unmatched islets are offered to any approved requesters willing to accept them, with no influence on future offers. 113 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 5499 2.2. Actual and simulated data sets 2.2.1. Actual ICR consortium data We evaluated the proposed matching algorithm using a retrospective pilot data set of 68 human pancreatic islet isolationsdistributedunderaLDMmodelandcollectedbytheICRconsortiumfromJanuary–September2005.Within the 68 islet isolations, a total of 6,653,944 IEQs were shipped in 217 shipments from 8 producers to 62 requesters located in 18 states across the United States (Fig. 2).As can be seen in Fig. 3, certain requesters received more than the desired quantity of islets, while others received no islets over a given time period, even though their requesting criteria and other conditions may be similar. (Among the 62 requesters, 23 (37%) were on the requesters list for the entire 9 months,andtheremaining7(11%),26(42%)and6(10%)ofrequesterswereaddedintothelistduringthefirst,middle and last 3 months, respectively.) The islet supply versus demand ratio, defined as the ratio of total produced islets by all producers to total minimum requested islets by all requesters in the 9 months period was 0.68. Producer Requester(s) Fig. 2. Geographical locations of 8 producers and 62 requesters in an ICR consortium study from January–September 2005. Each triangle represents an ICR producer, and each circle represents one or multiple requesters, with its area proportional to the number of requesters at a same or very close location. 10 20 30 40 50 60 Requester 1 10 10 2 10 3 10 4 10 5 10 6 10 7 Total IEQs Requested range (real data, n=44) Requested range (imputed,n=18) Received in actual distributions Received via MAID (30% rejection) Fig.3.Distributionof68isletisolationsfortheICRconsortiumdatafromJanuary–September2005usingtheLDMmodelandafterapplyingMAID. Each vertical black line represents the range of total islet amount desired by a requester during the study period, estimated by the corresponding minimum and ideal amounts per shipment, minimum days between shipments and length of time in study. Each vertical gray line has the same meaning as a black line, except that the range was imputed by combining known data and population mode values for missing values. Each circle “◦” represents the total islets received by a requester in actual distribution activities, and each cross “×” represents the total matched islets received by a requester in MAID-derived results, assuming a 30% rejection rate across all offers. 114 5500 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 To apply the MAID algorithm to these 68 isolations, we estimated the rejection rate and imputed missing val- ues as follows. The rate at which requesters rejected islet offers was estimated as 30%, based on an ongoing ICR consortium study from January–June 2006 (details not shown); hypothetical rejection rate of 15% was also tested to determine the impact of this factor. Missing values in 17 (27%) of requesters, each with at least one missing data point, were imputed as follows to ensure complete data on all requesting criteria: (i) missing values for de- sired minimum days between shipments were imputed based on the mode value of 30 days, (ii) no missing values existed for supplier preference data, (iii) for minimum and ideal numbers of IEQs per shipment, mode values of 5000 and 10,000 IEQs were used, respectively, if both values were missing, and ±50% of the stated value was used if one value was missing, (iv) for paired minimum and ideal purity values, mode values of 0.50 and 0.90 were used, respectively, if both values were missing, and ±50% of the stated value was used if one value was miss- ing, and (v) the imputation for minimum and ideal viability values was done the same way as the paired purity values. We also assume that islets from rejected offers can be redistributed to other qualified requesters by run- ning the algorithm repeatedly for a maximum of 5 times; unmatched islets also include rejected islets from the last algorithm run. 2.2.2. Data simulation Datasimulationmimickedaconsortiumforisletdistributionconsistingof8producersand80requestersovera1year period. Each simulation corresponds to two source data sets, one for 80 requesters and the other for a varied number of isolations during the study period. Both requester and isolation variables were simulated based on the corresponding data distributions in the above mentioned consortium pilot data set. Specifically, the variables for each requester j were simulated as follows: 1. Date of application approval t j,apv follows a density distribution such that the length of time in study is 1 year in 40 requesters and a random number of 0.5–1 year in the remaining 40 requesters. 2. Minimum days between shipments x j follows a truncated log-normal distribution, with median (range) of 21 (7, 243) days. 3. Supplier preference vector {y j (i)|1iI} follows a density distribution such that the number of acceptable pro- ducers is 1, 2–5 and 8 in 16%, 29% and 55% of requesters, respectively. 4. IdealnumberofIEQspershipmentq j,ideal followsatruncatedlog-normaldistributionwithmedian(range)of2×10 4 (10 3 ,5×10 5 )IEQs.ThecorrespondingminimumnumberofIEQspershipmentq j,min equalsq j,ideal ,0.75×q j,ideal and 0.5 × q j,ideal in 25%, 35% and 40% of requesters, respectively. 5. Ideal purity p j,ideal follows a truncated normal distribution with median (range) of 0.85 (0.50, 0.90). The corre- spondingminimumpurityp j,min equalsp j,ideal ,0.9×p j,ideal and0.8×p j,ideal in33%,33%and34%ofrequesters, respectively. 6. Ideal viability v j,ideal follows a truncated normal distribution with median (range) of 0.90 (0.50, 0.99). The corre- spondingminimumviabilityv j,min equalsv j,ideal ,0.9×v j,ideal and0.8×v j,ideal in33%,33%and34%ofrequesters, respectively. 7. Priority status: 20% of requesters have preferred status and the remaining 80% have standard status. 8. Delivery distance: 37% of requesters are within same day delivery distance from 1 producer. All the remaining deliveries require next day shipment. 9. Funding support: 69% of requesters have peer-reviewed funding status. The variables for each islet isolation u were simulated as follows: 1. Producer i is a random number selected from 1 to 8. 2. IsolationdatetfollowsaPoissondistributionconditionalonanaverageisolationfrequencydeterminedbyaprefixed islet supply versus demand ratio. 3. Number of produced IEQs Q follows a truncated log-normal distribution with median (range) of 7.7 × 10 4 (8 × 10 3 ,10 6 ) IEQs. 4. Purity P and viability V follow truncated normal distributions with median (range) of 0.90 (0.50, 0.95) and 0.92 (0.70, 0.99), respectively. 115 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 5501 We performed 10 replicated 1-year simulations under each of the three isolation frequency scenarios corresponding to the low, moderate and high islet supply versus demand ratios of 0.3, 0.6 and 0.9, respectively. The 10 replicated simulations represented a total of 7.0 ×10 7 IEQs in 657 isolations, or 1.4 ×10 8 IEQs in 1494 isolations, or 2.0 ×10 8 IEQs in 2001 isolations when the supply versus demand ratio was 0.3, 0.6, or 0.9, respectively. 3. Results 3.1. Application of MAID to actual ICR consortium data By comparison to the LDM approach in place during the time period tested, the MAID-derived results revealed several advantages over those in the actual distribution data (Table 1 and Fig. 3). First, in actual distributions, 24% of islets (1,576,498 IEQs) in 35 shipments met the four matching criteria on producer preference, minimum IEQs per shipment,minimumpurityandminimumviability,24%ofislets(1,578,656IEQs)in70shipmentsdidnotmeetatleast 1 criterion, and 53% of islets (3,498,790 IEQs) in 112 shipments contained missing data on islet characteristics and/or receivers’criteria.When the proposed MAID algorithm was applied to this data set after missing data were imputed as described above and under an estimated 30% rejection rate, 91% of islets (6,038,048 IEQs) in 194 shipments met all the four criteria, and the remaining 9% (615,896 IEQs) were either unmatched or rejected islets.As separate analyses, the total matched islets increased to 97% under both rejection rates of 15% and 0%. Second, of the 62 requesters in the study, 16 (26%) did not receive any islets in the actual distribution activities, and only 3 (5%) did not received any matchedisletsintheMAID-derivedresults(p =0.001byFisher’sexacttest,Fig.3).Forthe56requestersonstudyfor at least 3 months, the above numbers became 11 (20%) and 1 (2%) in actual and MAID-derived results, respectively (p = 0.004). Lastly, the number of requesters who received more islets than desired ideal amounts was 13 (21%) in the actual activities, and 0 (0%) in the MAID-derived results (p = 0.001, Fig. 3). Table 1 Actual and MAID-derived results for the ICR consortium data from January–September 2005 Variable a Actual data MAID under MAID under 15% rejections 30% rejections For 62 requesters No. of shipments received 2 (0, 17) 2 (1, 5) 3 (0,8) Proportion of requested shipments received 0% 16 (26%) 0 (0%) 3 (5%) 1–49% 29 (47%) 58 (93%) 38 (61%) 50–99% 10 (16%) 3 (5%) 17 (27%) 100% 1 (2%) 1 (2%) 4 (6%) >100% 5 (8%) 0 (0%) 0 (0%) Unable to determine 1 (2%) — — For 68 isolations No. of algorithm runs per isolation — 1.8 ± 1.22.3 ± 1.5 No. of qualified requesters per algorithm run — 10.1 ± 10.210.6 ± 9.5 Total shipments for all isolations 217 202 194 Criteria not fully specified 3,498,790 (53%) — — Criteria fully specified Matched 1,576,498 (24%) 6,438,944 (97%) 6,038,048 (91%) Unmatched 1,578,656 (24%) 215,000 (3%) 615,896 (9%) a Descriptive statistics are median (range), mean ± standard deviation, or number (%), as appropriate. “Criteria not fully specified” corresponds to the shipments with missing data for evaluation of all four matching criteria on supplier preference, minimum IEQs per shipment, minimum purity and minimum viability. “Matched” represents the shipped islets meeting all of the above four matching criteria. “Unmatched” denotes the shipped islets that did not meet at least one of the four criteria in actual data, or the combination of unmatched and rejected islets in the last algorithm run in MAID-derived results. “–” = not applicable. 116 5502 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 0.3 0.6 0.9 0 20 40 60 80 Supply vs. demand ratio No. of qualified requesters 0.3 0.6 0.9 0.0 0.2 0.4 0.6 0.8 1.0 Supply vs. demand ratio Prop. of unmatched islets Isolation groups: Both P & V ≥ 0.75 At least one < 0.75 0.3 0.6 0.9 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Supply vs. demand ratio Prop. of received islets Requester groups: Standard status Preferred status Fig.4.Simulationresultsbyisletsupplyversusdemandratio.Eachbox-plotineachsub-figurerepresentsarelationshipin10replicatedsimulations. Each simulation corresponded to a consortium setting of 8 producers, 80 requesters, and a varied number of isolations distributed in a 1-year period and under a rejection rate of 30%. The 10 replicated simulations represented a total of 657, 1494 and 2001 islet isolations generated under islet supply versus demand ratios of 0.3, 0.6 and 0.9, respectively. (a) Number of qualified requesters in each algorithm run. (b) Proportion of unmatched islets, stratified by isolation quality groups on purity and viability values. (c) Proportion of received islets, among requested ideal quantity, stratified by requester groups on priority status. 3.2. Application of MAID to simulated data Several statistics in the combination of 10 replicated 1-year simulations are displayed under islet supply versus demand ratios of 0.3, 0.6 and 0.9, respectively, to show both the patterns and the variations in the simulated data sets (Fig. 4). Fig. 4a indicates that the number of qualified requesters in all algorithm runs for all isolations had a negative correlationwiththemagnitudeofisletsupplyversusdemandratio(Pearsoncorrelation r=−0.30±0.08(range −0.14 to −0.43),p<0.0001ineachofthe10simulations).Inparticular,fiveormorequalifiedrequesterswerefoundin82%, 69% and 62% isolations when islet supply versus demand ratios were 0.3, 0.6 and 0.9, respectively. For islet supply versus demand ratio of 0.6 or less, the number of qualified requesters was large enough in more than 67% algorithm runs to allow a successful search of OOL with no leftovers, and was zero in less than 1% of runs that stopped the algorithm for further searches. Fig. 4b indicates that the proportion of unmatched islets was increased from 5% to 11% to 18% as a result of an increaseinisletsupplyversusdemandratiofrom0.3to0.6to0.9,respectively.Theproportionofunmatchedisletshad a 16–32% difference when comparing isolations with both purity and viability of 0.75 or higher to the remaining ones with low purity and/or viability. In particular, the proportion of unmatched islets was less than 7% in isolations with both purity and viability of 0.75 or higher when the supply versus demand ratio was 0.6 or less. Fig.4cindicatesrequesterswithpreferredprioritystatusreceivedhigherportionsofrequestedisletsthanotherswith standardprioritystatus.Whenisletsupplyversusdemandratiowasatamoderatelevelof0.6,theproportionsofreceived islets among requested ideal quantity in requesters with and without preferred priority status were 0.56 ± 0.24 and 0.43±0.24,respectively.Additionally,requestersacceptinglowerqualityislets(e.g.,p j,min <0.75and/orv j,min <0.75) received a 10–16% higher portion of requested ideal quantity than others only accepting high quality islets. Similarly, requesters asking for less than 10 6 IEQs per year received a 13–23% higher portion of requested ideal quantity than others requesting 10 6 or more IEQs per year. For islet supply versus demand ratio of 0.6, higher rejection rates resulted in higher proportions of unmatched islets (Table 2 and Fig. 5b). Numerically, the proportions of unmatched islets were 4.5% (0.68 ±0.53 ×10 6 IEQs per year), 7.8% (1.16 ±0.53 ×10 6 IEQs per year) and 11.4% (1.67 ±0.53 ×10 6 IEQs per year) under rejection rates of 0, 0.15 and 0.30, respectively (p<0.0001 for the difference in 10 simulations by one-way analysis of variance [ANOVA], Table 2). The number of qualified requesters in the initial algorithm runs in all isolations was 17.0 ± 10.0 under a rejection rate of 0, and dropped to 14.4 ± 9.1 and 13.8 ± 9.0 under rejection rates of 0.15 and 0.30, respectively (p = 0.01 ± 0.01 with range from <0.0001 to 0.04 in 10 simulations, Fig. 5a). The number of qualified requesters 117 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 5503 Table 2 MAID-derived results by rejection rate for simulated data Variable a Rejection rate p-value 0% 15% 30% Total matched shipments in 1-year period 532 ± 62 600 ± 65 610 ± 64 0.02 Unmatched islets in 1-year period Prop. among total produced 0.045 ± 0.030 0.078 ± 0.026 0.114 ± 0.025 <0.0001 Total IEQs (10 6 ) 0.68 ± 0.53 1.16 ± 0.53 1.67 ± 0.53 0.001 Prop. of requesters by % received islets #1 0% 0.030 ± 0.020 0.035 ± 0.021 0.038 ± 0.023 0.73 1–49% 0.411 ± 0.068 0.331 ± 0.069 0.318 ± 0.045 0.004 50–99% 0.356 ± 0.054 0.376 ± 0.049 0.388 ± 0.029 0.31 100% 0.014 ± 0.012 0.028 ± 0.020 0.033 ± 0.021 0.06 >100% 0.189 ± 0.064 0.230 ± 0.055 0.224 ± 0.055 0.25 Prop. of requesters by % received islets #2 0% 0.030 ± 0.020 0.035 ± 0.021 0.038 ± 0.023 0.73 1–49% 0.633 ± 0.079 0.543 ± 0.093 0.528 ± 0.079 0.02 50–99% 0.316 ± 0.064 0.396 ± 0.093 0.396 ± 0.064 0.03 100% 0.021 ± 0.022 0.026 ± 0.021 0.039 ± 0.029 0.27 >100% 0.000 ± 0.000 0.000 ± 0.000 0.000 ± 0.000 — a The reported mean ± standard deviation statistics were based on 10 replicated 1-year simulations generated under islet supply versus demand ratio of 0.6. The % received islets “#1” and “#2” denote the requester-specific proportions of received IEQs among the requested minimum and ideal IEQs, respectively. p-values were obtained from one-wayANOVA comparing statistics under rejection rates of 0%, 15% and 30% in 10 replicated simulations. “–” = not applicable. 0.00 0.15 0.30 0 20 40 60 80 Rejection rate No. of qualified requesters Algorithm run number: 1st run 2nd or later runs 0.00 0.15 0.30 0.0 0.1 0.2 0.3 Rejection rate Prop. of unmatched islets 0.00 0.15 0.30 0.0 0.2 0.4 0.6 0.8 1.0 Rejection rate Prop. of received islets Fig. 5. Simulation results by rejection rate. Each box-plot represents a relationship in 10 replicated simulations generated under islet supply versus demand ratio of 0.6, and the y-axis titles have the same meanings as in Fig. 4. in second and later algorithm runs further dropped to 7.9 ± 6.9 and 7.7 ± 7.0 under rejection tables of 0.15 and 0.30, respectively. Interestingly, the average proportion of received IEQs among requested ideal quantity in all requesters was 0.41 ± 0.23 under a rejection rate of 0, and slightly higher at 0.45 ± 0.24 and 0.46 ± 0.24 under rejection rates of 0.15 and 0.30, respectively (p = 0.43 ± 0.21 with range from 0.10 to 0.69 in 10 simulations, Fig. 5c). Requesters askingforsmallnumberofIEQspershipmenttendedtohavemoreopportunitytoreceiveoffersfromtheredistribution of rejected islets in second and later algorithm runs. 118 5504 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 4. Discussion In1998,theJuvenileDiabetesResearchFoundationInternational(JDRFI)createdtheJDRFhumanisletdistribution program.Asthefirstofitskindatthetime,theprogramsponsoredninepancreasprocessingfacilitiesworldwideinthe United States, Europe, and Canada to distribute human pancreatic islets to approved investigators conducting research in(a)preventionoftype1diabetes,(b)restorationandmaintenanceofnormalbloodglucose,and/or(c)preventionand treatmentofcomplications.EachfacilitywasresponsibleforindependentlymanagingacentrallyprovidedJDRFlistof approvedinvestigatorsandensuringthatrequestersonthatlistreceivedisletsinatimelyfashion.Upondiscontinuation of that program, the ICR distribution program was created shortly thereafter to address the need for human islets in the diabetes research community. This program adopted the JDRF’s LDM approach of allowing each pancreas processing facility to determine who receives islets, using an ABCC provided list of investigators conducting ICR approved research. The LDM approach to islet distribution has advantages and disadvantages. The immediate benefit to this approach is that producers can quickly make their own decisions about who receives the islets.These decisions will also include determining(a)theamountandtypeofisletseachrequesterreceives,(b)reasonsforexcludingorincludingrequesters, and (c) how the islets are packaged and shipped. The LDM approach also has serious limitations. First, because each producer does not know the islet offer and shipment history of requesters by other producers, certain requesters may receive more islet offers than others, while some may be excluded altogether from being contacted. Next, preferential treatmentofcertainrequestersbyaproducerisalwaysapossibility,andespeciallyaconcerntothoserequesterslocated outside of the producers institution. Finally, requesters who need sequential islet shipments in a compressed period of time to perform a series of experiments must coordinate each request with multiple producers to fill the demand. The development of MAID was intended to address the shortcomings of the LDM approach, and to obtain an optimal solution for matching producer islets to requester needs by using all available information. MAIDisdesignedtooptimizetheisletdistributionprocessinasettingofmultipleproducerswithmultipleapproved requesters eagerly awaiting islets. The algorithm searches for an OOL of requesters for a given amount of islets using a screening analysis to identify all qualified requesters, a sorting analysis to rank qualified requesters in descending order of priority scores, and a semi-exhaustive search procedure for the selection of optimal requesters that minimizes the amount of unmatched islets and maximizes the average priority score. A screening analysis may also include other criteria not evaluated in this report, such as those based on matching for type of islets (e.g., fresh, cultured, cryopreserved), and other penalties for rejections of previous offers. In particular, if the number of requesters on a waiting list is constantly large (e.g., 30 or more) and some requesters have not received islet offers for excessively long periods, we may impose a suspension period for rejecting matched offers in order to control the size of waiting list. The considerations for implementing the semi-exhaustive search procedure are two-fold. First, exhaustive search in the full space of 2 K − 1 subsets is computationally infeasible when the number of qualified requesters K is large. Second, the OOL obtained from the semi-exhaustive approach is guaranteed to represent the global optimum if the following two common conditions are met: (a) an exhaustive search in reduced space is performed without resorting to the subroutines of importance sampling and extended local search, and (b) the resulting OOL corresponds to zero unmatched islets. Therefore, an estimate of more than 80% OOLs identified in various matching analyses reported in Sections 4 and 5 were representing the global optimal solutions. Woodruff and Reiners (2004) have discussed the utilization of heuristic optimization to find good solutions in a reasonable amount of time, applying data mining for algorithm development. We have taken a similar approach here to more optimally allocate the limited supply of islets to multiple requesters, using a semi-exhaustive search procedure for combinatorial optimization. Other forms of optimization algorithms could be considered for approaching this real world problem of matching isletcellisolationstorequesters,whileincorporatingboththerequiredsubjectiveandobjectivecriteria.Forexample,a treepruning(TP)algorithmcouldbeusedtosearchforanoptimalcombinationofrequestersbyremovingnon-optimal branches in a systematic fashion.TheTP technique is efficient, well developed, and expected to have an equal or better performance than our brute force “local search” approach. However, we believe that the possibility of obtaining an improved OOL solution using a TP search is very low for the following reasons. First, because more than 50% of the requestersarewillingtohaveaflexiblerangeforamountofIEQspershipment(i.e.,q j,min typicallyis20–50%lessthan their q j,ideal ), the chance of finding a global optimal solution among the top 10 scoring requesters is high (normally >90%). Therefore, the extended searches via TP or other algorithms generally would be unnecessary. Second, on the same basis (i.e., q j,min <q j,ideal ), our local search approach is easy to implement, and is expected, although not 119 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 5505 guaranteed, to rapidly reach the optimal solution in most cases. If the amounts of IEQs per shipment were uniform (i.e., q j,min = q j,ideal ) for all requesters, then a TP search would represent a clear improvement over our proposed algorithm; however, the unmatched islets (i.e., leftovers) would be expected to occur much more often for all search procedures.Third,ouralgorithmdoesnotallowacombinationwith“N over10qualifiedmatches”.Forisletsproduced in one isolation, we want to find a combination with N not to exceed 10 offers. Therefore the benefit of applying a TP routine would be trivial, unless (a) the criterion of q j,min =q j,ideal is true in most requesters, and/or (b) N =20 or more offers are allowed for some “large” isolations, neither of which hold true in our experience. We have demonstrated by application to real/imputed and simulated data that our islet matching algorithm improves the ability of the ICR consortium to rapidly and efficiently match available islets to the list of approved requesters. When applying MAID to a real data set consisting of 68 isolations, fewer unmatched islets remained than when using the LDM approach. We note that a comparison of actual and MAID-derived results in this data set may be biased due to the existence of missing data, but the observed statistics clearly favor the MAID approach over the LDM model. SimulationstudiesalsoconfirmedthebenefitsoftheMAIDapproach,asreflectedinboththeproportionofunmatched islets in all isolations and the proportion of received among requested islets in all requesters.Additionally, MAID can incorporate new rules in each computational component, including the screening analysis, priority score calculations, andsearchforOOL.IsletofferrejectionscanberedistributedtootherqualifiedrequestersbyrunningMAIDrepeatedly. If, however, the amount of available islets at the time of shipment is less than the initial amount used in obtaining the OOL, some requesters in the original offer list will not receive islets. Such exclusions changes should be reflected in future islet distribution analyses. Implementation of the matching algorithm presents remaining challenges, including improving the supply versus demand ratio and the uniform quality of islets produced. Furthermore, as it was seen that half of all rejections are due to lack of ability to contact the approved requester, invoking a reliable rapid mode of response from requesters when an islet isolation offering is available is essential. Creation of a web-based application is currently underway to further evaluate the feasibility and efficiency of this approach. In conclusion, our results indicate that a centralized consortium of multiple producers is able to match more islets to more requesters than in the situation when islets are distributed separately, and the proposed matching algorithm is a useful tool for islet distribution activities under a consortium setting. Acknowledgments Islet distribution data were obtained through the Islet Cell Resource Center (ICR) Basic Science Human Islet Distribution Program. Helpful discussions about the rules for the algorithm were provided by ICR Steering Committee members. The authors are also grateful to Nora Ruel for help with preparation of the manuscript and Dr. Doug Stahl and Dr. Rebecca Nelson for early discussions on the algorithm. This work was supported by a grant from the National CenterforResearchResources(1U42-RR17673-01),acomponentoftheUSNationalInstitutesofHealth,Department of Health and Human Services. References Ahren, B., Taborsky, G.J., 2003. Beta-Cell Function and Insulin Secretion. sixth ed. McGraw-Hill, NewYork. Berney, T., Buhler, L.H., Morel, P., 2005. Pancreas allocation in the era of islet transplantation. Transplant Internat. 18, 763–767. TheDiabetesControlandComplicationsTrialResearchGroup,1993.Theeffectofintensivetreatmentofdiabetesonthedevelopmentandprogression of long-term complications in insulin-dependent diabetes mellitus. N. Engl. J. Med. 329, 977–986. Gaglia, J.L., Shapiro,A.M., Weir, G.C., 2005. Islet transplantation: progress and challenge.Arch. Med. Res. 36, 273–280. Hardikar,A.A., 2004. Generating new pancreas from old. Trends Endocrinol. Metab. 15, 198–203. Knazek, R.A., 2002. The human pancreatic islet cell resource consortium. Diabetes Technol. Therapy 4, 551–552. Martello, S., Toth, P., 1990. Knapsack Problems:Algorithms and Computer Implementations. first ed. Wiley, NewYork, NY, USA. Nanji, S.A., Shapiro,A.M., 2006.Advances in pancreatic islet transplantation in humans. Diabetes Obes. Metab. 8, 15–25. Nathan, D.M., 1993. Long-term complications of diabetes mellitus. N. Engl. J. Med. 328, 1676–1685. Nemhauser, G.L., Wolsey, L.A., 1988. Integer and Combinatorial Optimization. first ed. Wiley-Interscience, NewYork, NY, USA. Pinto, A., Tuttolomondo, A., Di Raimondo, D., Fernandez, P., Licata, G., 2004. Cerebrovascular risk factors and clinical classification of strokes. Semin.Vasc. Med. 4, 287–303. Ramiya,V., Schatz, D., 2004. Islet replacement vs. regeneration: hope or hype? Pediatr. Diabetes 5 (Suppl. 2), 45–56. Ridgway, D.M., White, S.A., Kimber, R.M., Nicholson, M.L., 2005. Current practices of donor pancreas allocation in the UK: future implications for pancreas and islet transplantation. Transplant Internat. 18, 828–834. 120 5506 D. Qian et al. / Computational Statistics & Data Analysis 51 (2007) 5494–5506 Shapiro,A.M., Lakey, J.R., Ryan, E.A., Korbutt, G.S., Toth, E., Warnock, G.L., Kneteman, N.M., Rajotte, R.V., 2000. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N. Engl. J. Med. 343, 230–238. Taylor, G.W., Manz, M.C., Borgnakke,W.S., 2004. Diabetes, periodontal diseases, dental caries, and tooth loss: a review of the literature. Compend. Contin. Educ. Dent. 25, 179–192. Trucco, M., 2005. Regeneration of the pancreatic beta cell. J. Clin. Invest. 115, 5–12. Weir, G.C., Bonner-Weir, S., 2004. Five stages of evolving beta-cell dysfunction during progression to diabetes. Diabetes 53 (Suppl. 3), S16–S21. Woodruff, D.L., Reiners, T., 2004. Experiments with, and on, algorithms for maximum likelihood clustering. Comput. Statist. Data Anal. 47, 237–253. 121 Cell Transplantation, Vol. 19, pp. 1133–1142, 2010 0963-6897/10 $90.00 + .00 Printed in the USA. All rights reserved. DOI: 10.3727/096368910X505486 Copyright © 2010 Cognizant Comm. Corp. E-ISSN 1555-3892 www.cognizantcommunication.com Effectiveness of a Web-Based Automated Cell Distribution System Joyce C. Niland, Tracey Stiller, James Cravens, Janice Sowinski, John Kaddis, and Dajun Qian Department of Information Sciences, City of Hope National Medical Center, Duarte, CA, USA In recent years, industries have turned to the field of operations research to help improve the efficiency of production and distribution processes. Largely absent is the application of this methodology to biological materials, such as the complex and costly procedure of human pancreas procurement and islet isolation. Pancreatic islets are used for basic science research and in a promising form of cell replacement therapy for a subset of patients afflicted with severe type 1 diabetes mellitus. Having an accurate and reliable system for cell distribution is therefore crucial. The Islet Cell Resource Center Consortium was formed in 2001 as the first and largest cooperative group of islet production and distribution facilities in the world. We pre- viously reported on the development of a Matching Algorithm for Islet Distribution (MAID), an automated web-based tool used to optimize the distribution of human pancreatic islets by matching investigator requests to islet characteristics. This article presents an assessment of that algorithm and compares it to the manual distribution process used prior to MAID. A comparison was done using an investigator’s ratio of the number of islets received divided by the number requested pre- and post-MAID. Although the supply of islets in- creased between the pre- versus post-MAID period, the median received-to-requested ratio remained around 60% due to an increase in demand post-MAID. A significantly smaller variation in the received-to-requested ratio was achieved in the post- versus pre-MAID period. In particular, the undesirable outcome of providing users with more islets than requested, ranging up to four times their request, was greatly reduced through the algorithm. In conclusion, this analysis demonstrates, for the first time, the effectiveness of using an automated web-based cell distribution system to facilitate efficient and consistent delivery of human pancre- atic islets by enhancing the islet matching process. Key words: Human islets; Algorithm; Automated cell distribution system; Islet Cell Resources (ICR); Islet equivalents (IEQs) INTRODUCTION therapy for a subset of patients afflicted with type 1 dia- betes mellitus (1). Moreover, human pancreatic islets have been shown to be a crucial basic science research In the last 30 years, the world has increasingly em- ployed an integrated analysis approach to optimize pro- and clinical resource in diabetes research, with at least a 15% increase in the annual supply each year from duction/distribution systems. The goal is to simultane- ously consider variables associated with the planning, 2004 to 2008, but a 25% annual increase in the demand of islets during those years, thereby exceeding the sup- manufacturing, inventory tracking, and distribution pro- cess in order to generate optimization models capable of ply each year (4). Similar to other domains, where the use of intermedi- effectively and efficaciously facilitating the delivery of a product (12). The use of an integrated analysis ap- ate inventory is not practical or desirable (2), the pro- duction of human pancreatic islets for long-term storage proach has been shown to improve production efficiency in a number of industries (3,7,13). However, the applica- and distribution purposes is not possible, as these cell clusters have a finite life span ex vivo. In addition, there tion of these methods for the production and distribution of biological materials is largely absent. is a limited supply of donated pancreata, such that care- ful isolation and utilization of the available islets are Recently, we proposed an algorithm to optimize the production and distribution processes for the isolation of critical to advancing this field (5). The process of human pancreas procurement and islet isolation is highly com- human pancreatic islets (10). Transplantation of these cell clusters, isolated from pancreata of deceased organ plex, and must be conducted under extremely careful conditions to yield this valuable resource (11). Once iso- donors, represents a promising form of cell replacement Received July 17, 2009; final acceptance April 21, 2010. Online prepub date: May 4, 2010. Address correspondence to Joyce C. Niland, Ph.D., Chair, Department of Information Sciences, City of Hope National Medical Center, 1500 East Duarte Road, Duarte, CA 91010-3000, USA. Tel: 626-256-4673, ext. 63032; Fax: 626-301-8802; E-mail: jniland@coh.org 1133 CHAPTER 5: IMPROVING AVAILABILITY OF ISLET TRANSPLANTATION IN TYPE 1 DIABETES Section 5.2: Effectiveness of a Web-Based Automated Cell Distribution System 122 1134 NILAND ET AL. lated, it is crucial that the resulting islets are distributed was highly ineffective, resulting in overdistribution to certain investigators, with low or no distribution to other in the most efficient and timely manner (within 24–48 h) to scientists who are able to receive and utilize these researchers, and poor alignment of the investigators’ needs in terms of the quantity and quality of islets re- islets, thereby minimizing waste of the islet product. The Islet Cell Resources (ICR) consortium was for- quired. Further, when it became difficult to readily place available basic science islets with awaiting approved re- mulated in 2001 as a group of current Good Manufactur- ing Practice (cGMP) qualified production centers for the searchers in a rapid time frame, the islets were either used locally, stored for a short period of time, or dis- isolation of human pancreatic islets for use in clinical transplantation in type 1 diabetes, as well as the isolation carded in some cases. In an attempt to improve the matching of available and distribution of human pancreatic islets for use in basic science research regarding this life-threatening dis- basic science islets to approved investigators, the Ad- ministrative and Bioinformatics Coordinating Center ease (6). While 14 distinct ICRs have been funded for islet production throughout the life of the consortium, (ABCC) of the ICR Consortium proposed the develop- ment of a Matching Algorithm for Islet Distribution over time the ICRs evolved to include eight highly pro- ductive islet distribution sites. Over the past 8 years, the (MAID) (10). This algorithm was developed with input from the ICR Steering Committee, composed of the number of studies approved to receive basic science is- lets from the ICRs grew from 23 in 2001 to 162 by the Principal Investigators and funding agency representa- tives for the ICR program. After development and test- end of 2008, representing over 200 researchers and 100 institutions across the US (4). This growth in demand ing, the matching algorithm was incorporated into the ICR web-based database system, also created by the for islets in basic science research created new chal- lenges in maintaining a fair and equitable distribution ABCC and deployed in February 2007. The purpose of this article is to evaluate the effectiveness of this islet across all approved basic science studies. A total of 323.2 million islet equivalents (IEQs) were isolated by distribution algorithm across approved investigators. The ability to match islets to waiting investigators pre- the ICRs from August 1, 2001 through December 31, 2008, with 68% of IEQs being used for basic science versus postimplementation of MAID is presented. research, 30% utilized in clinical transplantation into se- verely diabetic patients, and a small proportion (2%) un- MATERIALS AND METHODS able to be used either due to low quality/viability of the Development and Testing of the MAID Algorithm islets, lack of formal consent, or inability to distribute The goal of the MAID system is to ensure efficient them in a timely manner to the investigators with ap- fair distribution of islets to ICR approved investigators, proved research. From 2004 thru 2008, 2,810 shipments by best matching available islets to approved requestors, of islets totaling 78.6 million IEQs were distributed to to: 1) optimize the receipt of islets based on waiting time the investigators for basic science research. and priority of the requester, and 2) more closely match For the first 3 years of ICR consortium basic science the requestor’s criteria for islets than under the manual islet distribution, the islet-producing centers were re- distribution process, while 3) minimizing the number of quired to access a list of the growing number of investi- wasted islets. The core computations of the MAID sys- gators approved to receive the islets, and manually at- tem have been presented in detail elsewhere (10). Data tempt to place the islets with a researcher who could accumulated on ICR production/distribution behavior utilize them at that point in time. This manual distribu- from 2001 to 2006 were used to model patterns of islet tion method was fraught with numerous complicating distribution and investigator profiles, and to develop the factors and challenges, such as the need to appropriately islet matching algorithm. The MAID algorithm invokes match the number, quality, and type of available islets three steps for any broadcasted production batch of islets with the most suitable investigator; identifying an inves- in the following order: tigator whose lab was prepared to receive those islets within several hours; locating receiving institutions in 1. Screen approved islet requestors to identify all those whose minimum criteria for islet purity, viability, and close geographic proximity to the production site to minimize the impact of shipping on the islet viability; quantity would be met by the islet offer. 2. Rank those requests that pass the screening step, to determining the highest priority recipient based on who had been waiting the longest for islets, or had the highest determine the requestors who should be offered islets based on length of time waiting to receive islets and imperative study in progress; and meeting all of these challenges within a time frame that is conducive to islet other factors for favoring the ranking score. 3. Select the optimal combination of requestors, using quality and survival. Due to these complex factors, the manual matching a hybrid optimization procedure of exhaustive search coupled with importance sampling, depending on the of available basic science islets to approved recipients 123 AN AUTOMATED CELL DISTRIBUTION SYSTEM 1135 amount of available islets and the number of quali- the ABCC; and 4) direct online access to user feedback forms completed by the islet recipients after islet receipt fied investigators (10). and viewable by the islet production centers. Approved The ranking process takes into account the investiga- investigators are notified of available islets with auto- tor’s waiting time, in days, with the wait time set back mated messages via email, pages, and cell phone text to 0 days any time an investigator rejects a matched islet messages. To further ensure that the offer will be no- offer targeted to his/her criteria. The weighting coeffi- ticed, an islet offer “widget” that automatically pops up cients for favorable factors were established in consulta- on the selected investigator’s desktop is available to all tion with the ICR Steering Committee (SC), and con- investigators for download (Fig. 1). This widget displays sisted of the following: islet parameters such as available IEQ amount, batch a) Multiply the priority score by: 1.1 if the islets could type, time left, purity, viability, and broadcasting center. be delivered to the recipient within the same day; 1.1 Full data on the islet offers, acceptance/rejection, and if NIH or other peer-reviewed funding is available shipments are stored in the ICR database for further for the investigator’s study; 1.5 if islet purity is analysis and evaluation. within 5% of the ideal purity requested by the inves- tigators; 1.5 if islet viability is within 5% of ideal Islet Distribution Mechanisms of MAID viability requested by the investigators. Two distinct types of offers are administered by b) Add a large bonus score if the SC designates a high- MAID to promote the placement of all islets offered by priority study and it is important to expedite islet dis- the participating islet production centers (Fig. 2). In the tribution to complete this study rapidly. The priority first type of offer, targeted offers are created by MAID bonus score is equal to the largest score in all quali- when: a) there are one or more eligible investigators fied studies so that each high-priority study always whose requested islet criteria match the parameters of has a higher score than all studies with standard pri- the available islets, and b) there are more than 30 min orities. remaining before the shipment deadline established by The optimal combination of investigators who could the ICR center. Targeted offers are noncompetitive; the utilize all the islets produced with minimal waste are recipient of a targeted offer has been predetermined to then offered the islets; if an investigator accepts, the is- be part of the optimal match and thus has an inherent lets are shipped overnight by the production ICR site to priority to receive the offered islets, if they accept the the selected investigator(s). If the islets are rejected, or offer. Islets are reserved for individual investigators for there is no response from an investigator, then the algo- a duration of time equal to half of the remaining time rithm is rerun and the islets are offered to the next high- prior to the shipment deadline, set by the islet produc- est ranking selected investigator(s), until all the islets are tion site at the time of broadcasting the islet availability distributed. (up to the last 0.5 h). During this time, each offer recipi- ent will have the opportunity to review the available iso- Deployment of the MAID Algorithm Via the ICR lation data and accept or reject the offer. If one or more Database System of the targeted offer recipients rejects or fails to respond to the targeted offer, MAID will reevaluate the two tar- One of the key responsibilities of the ABCC was to develop a web-based database system for the ICRs, to geted offer criteria above, and send additional targeted offers, or proceed to the next offer type if necessary. allow them to remotely enter high-quality data detailing the process and the resulting characteristics of their islet When one or both of the targeted offer criteria listed above are false, MAID creates open offers as the second isolation process for every batch processed. These data are invaluable to assess the patterns and results of islet type of offer. Open offers are competitive; the islets are not reserved for particular investigators. Instead, more processing among the ICRs nationwide, and to correlate islet isolation data with results of basic science research than one investigator is notified of the same available islets, and all are given the opportunity to accept the and clinical transplantation for type 1 diabetes. Once developed and thoroughly tested, MAID was open offer. Once the shipment deadline has been reached and the offer is closed to additional offers, incorporated into the ICR database system, to support: 1) broadcasting islets available for approved investiga- MAID prioritizes the investigators that accepted the open offer using optimization techniques similar to those tors via the Internet, with automated immediate notifica- tion to investigators when they have been selected to used to determine priority for targeted offers. As a result of this competitive nature, it is possible that one or more receive an islet offer; 2) online investigator response re- garding their acceptance or rejection of the islet offer; of the investigators accepting an open offer will receive a lower priority score than the other accepting investiga- 3) online entry of islet shipment data, and automated shipping form creation for tracking islet shipments by tors, and therefore be denied by the system, if there is 124 1136 NILAND ET AL. Figure 1. Online web-based islet offer broadcasts and notification widget. In this example, the Islet Notification Service (INS) widget is being used by hypothetical user John Doe to monitor his three approved studies (usernames JDoe, JDoe2, and JDoe3). All three studies were contacted for the open offer corresponding to the least pure islet fraction of 40% purity from the hypothetical center, University of Islets. In addition, the second study (username JDoe2) also qualified for a targeted offer corresponding to the most pure islet fraction (85% purity). The INS displays 1 h, 49 min, and 30 s remaining for the open offers and 53 min and 34 s remaining for the targeted offer. Figure 2. Islet allocation process via targeted and open islet offers. 125 AN AUTOMATED CELL DISTRIBUTION SYSTEM 1137 an insufficient sum of islets to meet all accepted offers. the number of monthly shipments, the number of IEQs shipped per month, and type of islets shipped. Ship- In this instance, the investigators with the optimal open offer score are notified by MAID that they have been ments made after the implementation of MAID that did not use the system were excluded from the monthly val- selected to receive the islets from the open offer, and the investigators with the suboptimal open offer scores ues. Islet type was determined by high quality (viability and purity over 80%), medium quality (viability and pu- are notified that they were not confirmed to receive is- lets from this offer. Open offers provide the valuable rity between 50% and 80%), and low quality (either via- bility or purity less than 50% or unknown). utility of contacting a large group of investigators for the purpose of placing all available islets before the islet Box plots were created for the received-to-requested ratio pre- versus post-MAID, and the median and range isolation offer deadline. The number of open offer recip- ients is proportional to the number of available islets to of the ratios were calculated. Ratio values were highly skewed and not normally distributed; therefore, the Wil- ensure that a sufficient, but not excessive, number of investigators are contacted to place the remaining islets. coxon rank sum test was used to test for a difference in the received-to-requested ratio between pre- versus post- Evaluating the Effectiveness of MAID MAID time periods (9). However, because the average received-to-requested ratio is a measure of supply and We utilized data gathered pre- versus postdeployment of the automated algorithm, to determine whether the demand, it is not a great measure of how MAID may have improved islet distribution; we were therefore more deployment of MAID had a positive impact on the abil- ity of the ICRs to effectively place their isolated islets interested in whether the variability of the received-to- requested ratios differed between the two time periods. with investigators requiring such islets for research or transplantation. Over a 43-month period, from February The Levene’s Test of Homogeneity was used to deter- mine if the variability of ratios was similar between the 2004 to August 2007, a total of 1,296 shipments consist- ing of 39,369,909 IEQs were distributed pre-MAID im- two time periods (8). In addition, the ratios were catego- rized into intervals, and the percent of studies falling plementation (63 shipments distributed using the manual process during the first 6 months of initial deployment into each of the ratio intervals was calculated. The chi- square test of association was applied to evaluate of MAID were also included). During a 22-month pe- riod, from March 2007 to December 2008, 1,514 ship- whether there was an association between the received- to-requested ratio pre- and postalgorithm implementa- ments with a total of 39,314,776 IEQs were distributed post-MAID implementation. Across both time periods tion. We recognize there was clustering present due to investigators having multiple studies; therefore, we also these shipments were made to 154 requesting studies from 14 ICR centers. There were 110 active studies in ran a generalized estimating equation (GEE) analysis clustering studies by investigator and adjusting for the the pre-MAID period, and 118 active in the post-MAID period; 74 of the studies active during both periods, and amount of requested islets (<3,000,000 or ≥3,000,000 islets) to test homogeneity of the variability. Similarly, therefore these studies were considered in both time pe- riods. we also ran the GEE analysis to test the association of the received-to-requested ratio groups pre- and post- To determine the effectiveness of the algorithm in distributing islets for approved studies, a standard mea- MAID implementation. Because these analyses yielded similar results, we are reporting the original methods for surement was created to compare successful placement of islets pre- versus postalgorithm implementation. The ease of interpretation. Although the breadth of the data is limited during the measurement was a ratio of received-to-requested islets, calculated for each approved study of islets. Received prealgorithm period, advances in data collection related to the implementation of the algorithm allow for more islets were tracked manually by the ABCC pre-MAID, and via the web-based system postimplementation. dynamic analyses during the post-MAID period. With this additional data during the postalgorithm period, we There were five possible scenarios of a study’s active status pre- and post-MAID deployment, as shown in further analyzed possible reasons for any observed dif- ferences in the received-to-requested ratio. Post-MAID Figure 3. The scenarios included studies that either only took place in the pre-MAID period, only in the post- ratios were evaluated against two factors: total amount of approved islets needed for the study’s research MAID period, or in both periods. Received-to-requested ratios were calculated for each period a study was ap- (<3,000,000 IEQs vs. ≥3,000,000 IEQs), and an accep- tance rate (i.e., the proportion of offers a study agreed proved. Eight studies were only conditionally approved ICR islet recipients and were therefore excluded from to accept for a given study: 0–24%, 25–49%, or 50– 100%). The Wilcoxon Rank Sum test was used to deter- the analysis. To determine how the average distributions changed mine whether the ratio of received-to-requested islets differed depending on the total amount of islets ap- between the pre- and post-MAID periods we analyzed 126 1138 NILAND ET AL. Figure 3. Possible scenarios for calculation of the received-to-requested ratio pre- and postimplementation of matching algorithm for islet distribution (MAID). proved for a study. A Spearman correlation coefficient ity of the received-to-requested ratios pre- versus post- implementation (p = 0.04) (Fig. 4). Pre-MAID ratios was calculated to determine if there was a significant linear trend between the level of accepted offers and the varied from no islets received to 45.4 times the re- quested amount, while post-MAID these ratios had received-to-requested ratio (9). All analyses used the SAS statistical software (version 9.1; SAS Institute, greatly reduced variability with a range from 0 to 4.4. The proportions of the received-to-requested ratio by Cary, NC). intervals are shown in Table 2. A very similar propor- RESULTS tion of studies received less than the desired quantities (up to 75% of their requests) for 54% of studies in the The average production of islet distribution activities was increased in pre- versus post-MAID as shown by pre-MAID period, and 58% of studies in the post-MAID period, reflecting that only two thirds of the demand both number of monthly shipments made and monthly IEQs shipped (Table 1). The number of monthly ship- could be fully met by production in either time frame. The proportion of studies in the most desirable received- ments increased from 33 to 69 and the number of IEQs shipped per month increased from 1 million to 1.8 mil- to-requested ratio range (from 76% to 150% of their de- sired amount) more than doubled after implementation lion. High-quality islets were more frequently distrib- uted in post-MAID compared to the pre-MAID period. of the islet distribution algorithm, from 13% pre-MAID to 31% post-MAID. Additionally, the undesirable out- The box plots for the pre- versus post-MAID period are shown in Figure 4. It can be seen that while production come of providing more islets than required (from 151% to over 400%) was cut by two thirds through implemen- increased, so too did demand, such that the median re- ceived-to-requested ratios were almost identical pre- tation of the algorithm, with 33% of studies in this over- supplied range pre-MAID, and only 11% post-MAID. versus postimplementation of MAID (0.63 vs. 0.61, re- spectively, p = 0.13), indicating that in both time periods The association between implementing the algorithm and the ratio of received-to-requested was statistically approximately two thirds of the desired islets could be delivered to the requesting investigators. It can also be significant by chi-square test (p = 0.001). Table 3 summarizes our investigation into possible seen that there is a significant difference in the variabil- 127 AN AUTOMATED CELL DISTRIBUTION SYSTEM 1139 Table 1. Summary of Pre- and Post-Web Deployment of the Matching Algorithm for Islet Distribution (MAID) Variable Pre-Web Post-Web Monthly shipments* 33.3 ± 20.9 (3, 82) 69.0 ± 29.7 (21, 128) Monthly shipped IEQs* (10 6 ) 1.0 ± 0.8 (0.05, 3.3) 1.8 ± 0.7 (0.4, 3.4) Shipments by purity & viability Both 80% or higher 553 (43%) 765 (51%) Both 50% or higher 218 (17%) 341 (22%) Either below 50%, or unknown 525 (40%) 408 (27%) The reported statistics are mean ± SD (range), frequency (%). *Note that 63 (5% of pre-MAID) shipments distributed using manual process were excluded from the pre-web monthly shipments and monthly shipped IEQs because they were made after the inception of the MAID system. Figure 4. Distribution of received-to-requested islet ratios of pre- versus post-MAID for all approved studies (N = 154). Variability demonstrated in the box plots was significantly different between the pre-MAID and post-MAID period using Levene’s Test for Homogeneity (p = 0.04). 128 1140 NILAND ET AL. Table 2. Frequency of Received-to-Requested Islet Ratios by Pre- Versus Post-MAID for All Approved Studies (N = 154) Pre-Website Post-Website Frequency Frequency Interpretation of Ratio Ratio (N = 110) (N = 118) Less Desirable—under supplied 0.00 12 13 >0.00–0.25 19 19 >0.25–0.50 16 19 >0.50–0.75 13 18 Percentage 54% 58% Desirable—supply similar to request >0.75–1.00 7 18 >1.00–1.50 7 18 Percentage 13% 31% Less desirable—over supplied >1.50–2.00 7 8 >2.00–3.00 11 2 >3.00–4.00 6 2 >4.00 12 1 Chi-square test of association was performed to determine if there is an association between received-to-requested ratio and time period (p = 0.001). “Interpretation of Ratio” refers to the categorization of islet supply based on the balance between distribution efficiency, fairness among islet recipients, and feasibility of research resulting from the available supply of islets. causes of a low received-to-requested ratio postimple- ing the lowest median received-to-requested ratio of 0.50, those accepting 25–49% of offers having a more mentation of MAID using the enriched data available postimplementation of MAID. As could be anticipated, favorable median of 0.75, and those accepting 50% or more of offers having a highly desirable median ratio of studies that desire a very large amountof islets (>3,000,000 IEQs) were significantly less likely to receive their de- 0.85 (p < 0.0001). sired amount, with a median received-to-requested ratio DISCUSSION of 0.33 for those requiring greater than 3 million islets, versus a more desirable ratio of 0.71 of their needs for Before the development of an automated method for matching investigator criteria to islets available for dis- studies requesting smaller amounts (p = 0.02). In addition, the influence of a study’s acceptance rate tribution, islet isolation centers needed to manually match the islets to a list of investigators interested in of both targeted and open offers was evaluated for their ratio of received islets. There was a linear trend ob- obtaining them. This was often a time-consuming pro- cess, and resulted in undistributed or wasted islets or an served, with those accepting only 0–24% of offers hav- Table 3. Evaluation of Factors Associated With Received-to-Requested Islet Ratio Postimplementation of MAID Median Received-to- Variable N Requested Ratio 25th–75th Percentile p-Value Total approved islets (N = 117 studies)* Total approved islets <3 million 98 0.71 0.26–1.06 0.02 Total approved islets ≥3 million 19 0.33 0.12–0.55 Acceptance rate (N = 107 studies)† Accepted 0–24% of offers 53 0.50 0.18–0.88 <0.0001 Accepted 25–49% of offers 40 0.75 0.42–1.10 Accepted 50–100% of offers 14 0.85 0.70–1.20 *N = 117 and not 118 because one study did not provide a value for total islets. The p-value was calculated using Wilcoxon rank sum test. †Only studies that received open and/or target offers were included in analysis. The p-value was calculated by Spearman correlation. 129 AN AUTOMATED CELL DISTRIBUTION SYSTEM 1141 oversupply to the same requesters. To ensure the most algorithm in properly allocating supply even in environ- ments where demand greatly exceeds supply. Although efficient distribution of islets, the ABCC implemented an automated matching algorithm to aid in islet distribu- the received-to-requested ratio can be affected by a number of factors outside the investigators’ control, tion. With the experience of close to 2 years of islet distributions using the automated algorithm, the authors such as production levels and quality of islets, the MAID algorithm successfully accounts for all of these have evaluated its impact on successfully placing islets with approved investigators who request them. factors. When investigators respond in a timely manner to the offers provided, MAID achieves a reduction in This evaluation of our automated method for match- ing scientists requiring islets with the islets produced by extreme values of the received-to-requested ratio. Fur- ther, the MAID algorithm closely matches targeted offer the ICRs nationwide demonstrates the efficiency of an integrated analysis approach for a production distribu- characteristics to investigator needs in the post-MAID period. In addition, postimplementation of MAID the al- tion system involving human biological materials. We have shown that the MAID system greatly improved this gorithm was able to place all of islets in an isolation batch 93% of the time, and placed at least 75% of the matching process. The median received-to-requested is- let ratio did not differ between pre- and post-MAID im- islets in 96% of the batches offered for distribution. While this information was not available pre-MAID, our plementation, reflecting the continued demand of islets exceeding the amount of available islets supplied. The experience suggests that the pre-MAID distributions were not as closely matched to investigator requests, or variability in how many islets the investigators received was substantially reduced, indicating a more fair distri- placed as successfully as the post-MAID distributions. This study has a few limitations. The amount of de- bution of islets in accordance with the requests. Fewer investigators received very low proportions of their tailed information is much richer postimplementation of MAID than preimplantation, when only manual records needed islets, and there were many fewer instances of oversupplying some investigators, to the detriment of of islet shipments (and not offers) were kept in this ear- lier time period. Therefore, while we could identify the others waiting for islets. The fact that, in both time periods, only about two some conditional factors captured through the data sys- tem that might influence successful placement of islets, thirds of the desired islets nationwide could be supplied is a reflection of the supply-to-demand issue across the we did not have this information recorded pre-MAID. We were also unable to determine if MAID was success- nation. The demand outstrips the supply due to many factors: the complex process in producing high-quality ful in reducing the number of unused islets because of the lack of information pre-MAID. To avoid bias in one islets, establishing and maintaining a cGMP-approved infrastructure, hiring and training technical staff, and ob- time period or another such investigators were included in both time periods; however, this could alter the results taining viable pancreata for harvesting. Solutions to these constraints need to be reached in order to fully slightly. The supply-to-demand ratio remains the largest prob- satisfy the diabetes research and transplantation needs for islets. lem (4). If this could be resolved, all the requests for islets to speed diabetes research could potentially be re- Our evaluation also showed that the optimal re- ceived-to-requested ratio of 1.00 is achievable. If the to- solved. In the meantime, this evaluation has demon- strated that the MAID automated islet distribution sys- tal amount of islets required is in a practical range of less than 3,000,000 islets, the supply for such studies tem greatly enhances the islet matching process, and delivers closer to the desired amount of islets without can be successfully met by using the MAID system to match their requests with the islet isolations. However, wasting or oversupplying them. Therefore, we will con- tinue to rely upon and continually enhance the MAID as seen, certain experiments may require the use of a greater number of islets than others (4). Hence, this fac- system. Such a system could potentially be modified to fit other matching problems in the scientific world in tor may not be modifiable in experiments requiring such a large amount of islets. future, such as organ donors. An influencing factor that is under the control of the ACKNOWLEDGMENTS: Islet distribution data were obtained investigators requesting islets is their acceptance rate through the Islet Cell Resource Center (ICR) Basic Science when islets are offered to them via the automated MAID Human Islet Distribution Program. Helpful discussions about the rules for the algorithm were provided by ICR Steering system. If the investigators respond to these offers and Committee members. The authors are equally grateful to all accept them at least 25–49% of the time, a very reason- of the ICR principal investigators and staff members at each able received-to-requested ratio of 0.75 offers can be contributing laboratory for providing guidance in the develop- achieved. If the investigators accept the islet offers even ment of the algorithm and feedback during deployment of the half of the time this ratio jumps to 0.85 of their desired ICR islet allocation system. In addition, special thanks is given to Martha Antler for assistance in collecting some of the data islet amount. This illustrates the success of the MAID 130 1142 NILAND ET AL. used in this manuscript. This work was funded by the coopera- transplantation in diabetes mellitus: How to allocate de- ceased donor pancreata? Transplant. Proc. 38:365–367; tive efforts of the National Center for Research Resources (NCRR) and the National Institute of Diabetes and Digestive 2006. 6. Knazek, R. A. The human pancreatic islet cell resource and Kidney Diseases, a component of the US National Insti- tutes of Health (NIH) in conjunction with the generous contri- consortium. Diabetes Technol. Ther. 4:551–552; 2002. 7. Lee, Y. H.; Kim, S. H. Production–distribution planning butions of the Juvenile Diabetes Research Foundation. Drs. Niland and Qian, Messrs. Cravens and Kaddis, and Mss. in supply chain considering capacity constraints. Comput. Ind. Eng. 43:169–190; 2002. Stiller and Sowinski were supported by grant U42 RRR017673 (J.C.N) from the NCRR. 8. Montgomery, D. Design and analysis of experiments, 5th ed. New York: Wiley; 2001. REFERENCES 9. Pagana, M.; Gauvreau, K. Principles of biostatistics, 2nd ed. Pacific Grove, CA: Duxbury; 2000. 1. Alejandro, R.; Barton, F. B.; Hering, B. J.; Wease, S. Up- date from the Collaborative Islet Transplant Registry. 10. Qian, D.; Kaddis, J.; Niland, J. C. A matching algorithm for the distribution of human pancreatic islets. Comput. Transplantation 86:1783–1788; 2008. 2. Chen, Z. L.; Vairaktarakis, G. L. Integrated scheduling of Statist. Data Anal. 51:5494–5506; 2007. 11. Ricordi, C.; Strom, T. M. Clinical islet transplantation: production and distribution operations. Manag. Sci. 51: 614–628; 2005. Advances and immunological challenges. Nat. Rev. Im- munol. 4:259–268; 2004. 3. Cohen, M. A.; Lee, H. L. Strategic analysis of integrated production-distribution systems: Models and methods. 12. Sarmiento, A. M.; Nagi, R. A review of integrated analy- sis of production distribution systems. IIE Trans. 31: Oper. Res. 36:216–228; 1988. 4. Kaddis, J. S.; Olack, B. J.; Sowinski, J.; Contreras, J. L.; 1061–1074; 1999. 13. Vidal, C. J.; Goetschalckx, M. Strategic production-distri- Niland, J. C. Human pancreatic islets and diabetes re- search. JAMA 301:1580–1587; 2009. bution models: A critical review with emphasis on global supply chain models. Eur. J. Oper. Res. 98:1–18; 1997. 5. Kandaswamy, R.; Sutherland, D. E. Pancreas versus islet 131 CHAPTER 5: IMPROVING AVAILABILITY OF ISLET TRANSPLANTATION IN TYPE 1 DIABETES Section 5.3: Standardized Transportation of Human Pancreatic Islets: An Islet Cell Resource (ICR) Center Study of Over 3000 Shipments CHAPTER 5.3 ABSTRACT Mammalian cells, including pancreatic islets, have been shown to be responsive to temperature changes in-vitro and in-vivo, yet protocols that control for thermal fluctuations during cell transport are lacking. An initial assessment of transportation conditions was conducted using standardized materials and operating procedures in 48 shipments sent to a central location by 8 human islet manufacturing laboratories using a single commercial carrier. Optimization of preliminary conditions was then conducted. Evaluation of preparation quality was next performed in 3091 shipments of human pancreatic islets, received and used by diabetes researchers, pre and post implementation of a finalized shipping container and operating procedures. The outside temperature ranged from a mean low of -4.6±10.3°C to a high of 20.9±4.8°C; disposable indicators readings were not available for 3 shipments, recorded a within-container temperature drop to < 15°C in 16 (36%), stable values between 15-29°C in 29 (64%), and no occurrences of fluctuations >29°C in any shipment. Implementation of an optimized transportation container and operating procedure 132 reduced the number of within-container temperature drops to 11.3% (n=39 of 346 corresponding winter shipments), improved the number maintained between 15- 29°C to 88.1% (n=305), but also increased the number reaching or exceeding 29°C to 0.6% (n=2; overall p<0.0001). In 3091 shipments of human islets, quality rankings of excellent to good improved from 81.9% (n=1252) to 91.1% (n=1424) pre vs. post implementation, as did the number of fair to poor from 18.1% (n=275) to 8.9% (n=140; overall p<0.0001). Our results show that extreme temperature fluctuations during transport of human pancreatic islets occur when using a commercial carrier for long distance shipping, but can be controlled using standardized containers, materials, and operating procedures. This protocol, developed for the transportation of human pancreatic islets, can be adapted for use with other mammalian cell systems. CHAPTER 5.3 INTRODUCTION Human pancreatic islets play a vital role in diabetes research 1 . Because the isolation of these preparations are expensive 2 and require the use of specialized facilities staffed with experienced personnel, the need for islet shipping protocols to extend the services of established pancreas processing centers to distant clinical transplant centers and basic science laboratories is growing 3,4 . Rigid control of within-container temperature during transit has not been shown to be necessary in a protocol designed for short term shipments using a charter jet 4 , 133 but changes of at least 20°C during intra- 5 and inter- 6 continental shipments using a commercial carrier have been documented. As the duration and distance of islet shipments increase 7,8 , so to is a concern over adverse temperature effects on preparation quality and function. Exposure of islets to moderate hypothermia at 22-24°C remains the most widely used non- physiological temperature during cell culture. While the benefits and consequences of culture at 37°C vs. 22-24°C have been reviewed by Murdock and colleagues 9 , little is known about the effects of prolonged extreme hypothermia on isolated human islets. In proliferating mammalian cells, culture at 4°C reduces cell growth and viability 10 , but also leads to morphological and nuclear alterations, including bleb formation and DNA fragmentation, and cell apoptosis 11 . Cell cycle progression has been shown to be altered in a number of cell lines grown at <20°C 12 , although a mild reduction of temperature, from 37°C to 32°C, in culture prevented apoptosis by a variety of cellular stressors 13 . Furthermore, exposure of islets to hyperthermic conditions have also been shown to alter cellular survival and function. Brandhorst and colleagues exposed pig islets to mild hyperthermia at 43°C and found that although there was an increase in the resistance to inflammation by in-vitro stimulants when compared to controls at 37°C, there was also an enhancement in the number of apoptotic proteins detected in vitro as well as a reduction in the early survival of 134 xenografts 14-16 . In proliferating mammalian cells, mild hyperthermia was also shown to diminish the inflammatory response of cytokine-stimulated cell lines 17 . However, islet graft angiogenesis and revascularization was reduced in heat shock preconditioned hamster islets, vs. controls, following transplantation 18 . In heat shocked human islets, in vitro protection against cytokine induced damage and nitric oxide radicals has been reported 19 . Although variation in response to hyperthermia may be explained, in part, by different levels of species-specific basal expression of heat shock and other cellular stress response proteins 20 , the effects of exposure to non-physiological temperature increases cannot be ignored. A number of published articles have reviewed the cold 21 and heat 22 shock responses of mammalian cells, including studies on gene expression changes during periods of thermal fluctuation 23,24 . As the appreciation of thermal effects on cell biology increase, so to is the need to minimize any undesirable influence of extreme temperature on human islets during transportation. We therefore conducted a series of experiments to design, optimize, and implement a method for long-distance transport of human pancreatic islets using a commercial air- carrier and standard operating procedures. The goal was to stabilize the temperature within the shipping container during long distance transport and minimize extreme fluctuations exceeding ±7°C of the selected reference value at 22°C. Several recent studies have employed the use of 22°C following a short 135 initial culture period of 37°C 25-27 ; however, the concept of reducing human islet culture temperature was introduced more than 2 decades ago 28,29 . This study was performed by the National Islet Cell Resource Center (ICR) Consortium, a multi-center initiative from 2001-2009 focused on improving human islet isolation and transplantation technologies. A total of 3091 human pancreatic islet shipments, made by 14 different laboratories to investigators in North America and Europe, were assessed pre- and post-implementation of a standardized shipping container and operating procedures to evaluate the quality of the preparation. CHAPTER 5.3 MATERIALS AND METHODS Establishing Standardized Shipping Conditions. Islet containers were used to generate and assess temperature, pressure, and transportation data collected during shipment of packages prepared following a standard operating procedure (SOP). Fed-Ex overnight service was used by each participating laboratory twice a week for three consecutive weeks. A total of 8 facilities from 7 states completed 6 shipments each, all delivered centrally to the City of Hope in Duarte, CA. A reusable temperature and pressure data logger, a disposable temperature indicator strip, 2 ambient temperature stabilization gel packs, a 240mL cell culture bag, an absorbent cotton pad, and packing peanuts were all packaged inside a polystyrene foam inner container with a corrugated cardboard outer box. Ambient temperature stabilization gel packs were prepared according to SEBRA 136 manufacturer protocol for temperature protection of platelet concentrate units at 20-22°C (instruction manual available by request through SEBRA website). Reliability assessments of the temperature and pressure data loggers were performed prior to use in shipment. Readings for these experiments were logged every 2 minutes for temperature and 8 minutes for pressure for the duration of each shipment. Pressure data is not included in this report, as it was not focus of the study. Temperature Control. The use of an ambient temperature stabilization gel pack inside an islet shipping container to control cold and hot weather changes has been previously reported 4 . Using this same product, optimal temperature control inside a standardized islet shipping container was assessed using 0, 2, 4, 6, and/or 8 gel packs under fixed and alternating external temperature environments. Fixed conditions used included exposure to -20°C, 4°C, room temperature (16-25°C), and 37°C. Alternating conditions used included temperature variations from -20°C to 37°C and vise versa. Temperature readings for these experiments were logged every 10 seconds for a minimum of at least 18 hours. Islet Distribution Data. From February 12, 2004 – December 31, 2010, 3678 shipments of human islets were provided to investigators by the ICR Basic Science Human Islet Distribution Program. Of those, 522 records were excluded 137 because they did not occur within the defined study periods. An additional 65 records were excluded due to missing information. Data on 3091 shipments were used for this analysis. Supplies. Expanded polystyrene foam containers enclosed in corrugated cartons (model # 355) were purchased from ThermoSafe (Hayward, CA). Ambient temperature stabilizing gel packs (model # 1290) were purchased from SEBRA (Haemoentics Corp.; Braintree, MA). DURASORB underpads (model # 1093) were purchased from COVIDIEN (Mansfield, MA). Bubble wraps (model #S-214) and biodegradable cornstarch peanuts (model #1564) were both purchased from ULINE (Waukegan, IL). Tempasure plus disposable temperature indicators (model # TLCSEN 364A) were purchased from Tip TEMPerature Products (Burlington, NJ). Reusable HOBO temperature and pressure data loggers, with software and adapters, were purchased from Onset (model #s UA- 004-64, HPA-0015, BCP4.3-ON, and USB232 respectively; Bourne, MA). Permalife 240 ml FEP cell culture bags were supplied by Origen (Austin, TX) and filled with 240 ml of CMRL 1066 supplemented media purchased from Mediatech, Inc. (Manassas, VA). Kodak digital cameras (model #C530) were purchased from Office Depot (Delray Beach, FL). Study Outcome. The primary study endpoint was the within-container transport temperature, measured using either a reusable temperature data logger 138 (continuous variable) or disposable indicator (categorical data). Human islet quality ranking was used as a secondary study endpoint to qualitatively measure the usability of and satisfaction with the cellular preparation for laboratory experimentation post-shipment by each investigator. Researchers are asked to rank the shipped islets as either excellent, good, fair, or poor, using any method(s) they choose. Re-coding of this variable into excellent or good and fair or poor reduces the number of categories from 4 to 2 and potentially diminishes the subjectivity of the data by eliminating distinctions between excellent vs. good or fair vs. poor. Univariate and Multivariable Analysis. Percentages are reported for all categorical variables. The measure of central tendency was described using both the mean (± 1 SD) and median (min, max) for all continuous variables, unless otherwise noted. Coding of continuous variables to categorical groupings is noted when used. The reliability of reusable temperature monitors was evaluated using sensitivity and precision statistics. The sensitivity statistic is defined as a measure of linearity and quantified using the Pearson’s correlation coefficient for between-monitor observations. The precision statistic is quantified by comparing the maximum absolute difference of the mean value across all tested monitors at each observation time. The variation of temperature during transportation was evaluated using the change range during shipment and the fastest rate of change per minute. Assessment of the agreement between the 139 reusable and disposable temperature monitors was evaluated using the data frequency, accuracy rate, and kappa statistic 30 . Use of pearson’s chi-square test was employed to examine the differences in islet quality rankings prior to and after implementation of a standardized islet shipping protocol. Because values in all cells did not exceed 5, Fisher’s exact test was used, in place of the chi- square, to examine the differences in temperature probe activations across seasonal variations. Statistical significance for all tests was defined as p<0.05, unless otherwise noted. An analysis of shipping conditions was performed using univariate logistic regression (LR) model testing. Chi-square p-values, the corresponding odds ratio (OR), and the 95% confidence interval (95% CI) are reported for all categorical variables. The profile likelihood method was used to generate the reported statistical measures. To test for simultaneous effects of multiple predictors, all variables found to be significant from univariate LR analysis with a p-value of <0.20 were considered in defining an optimal multivariable model. A stepwise LR approach was used to choose independent factors of significance. All statistical analysis was performed using SAS software version 9.1.3 12 (SAS Institute, Cary, NC). CHAPTER 5.3 RESULTS Equipment Testing. The reliability of temperature data loggers used for 140 monitoring values inside the shipping container during transportation was evaluated prior to initiation of this study (Supplemental Table 1). The data logger demonstrated high sensitivity in both trial runs, as measured by the linear agreement between all pairs of probes tested (r=0.97±0.01 and 0.93±0.05 for temperature; r=0.99±0.01 and 1.00±0.00 for pressure). The maximum variation seen in temperature monitor readings was less than 1°C from the mean value, signifying moderate precision in all trial runs (0.49°C ± 0.32 in run 1; 0.98°C ± 0.49 in run 2). The coefficient of variation for the temperature monitors did not exceed a mean value of 3.74%. Preliminary Shipping Evaluation. Eight ICR laboratories completed a total of 48 shipments (Table 1). Complete data for all parameters of interest was obtained in 44 of 48 shipments (92%), with partial data available on the remaining 4 shipments (8%). Compliance to the shipping SOP was documented via a) digital photography of the islet container during different stages of the package preparation process (data not shown), and b) monitoring of the instructed packaging temperature (Supplemental Table 2). Twenty six of 46 shipments (57%) shipments were packaged within the instructed temperature range of 21- 26 °C, with an additional 20 shipments at or below 21°C (43%). Fluctuations in temperature and pressure were seen within containers during shipment (Table 2). Temperature changes inside the shipping container varied 141 by center and ranged from an overall low of 4.7°C to a high of 26.6°C. The mean temperature change inside the box during shipment was 7.5°C ± 3.8 (range: 2.1- 18.9). The maximum rate of temperature change in the container was found to be 0.32°C/minute ± 0.24 (range: 0.05-1.2). Ambient temperature at departure, arrival, and hub cities during shipment ranged from -19.4 to 31.7 °C. Forty two of 46 shipments (91%) were delivered on time, with a mean shipping distance of 1851 miles ± 726 (range: 38-2544). Detailed temperature and pressure profile line plots by ICR center and shipment are also reported (Supplemental Figure 1). The performance of disposable temperature indicators relative to reusable probes was also evaluated (Supplemental Table 3). We found statistically significant moderate agreement between disposable temperature indicators and reusable temperature probes (Kappa=0.4974; 95% CI 0.251 to 0.744; p<0.0001). To determine if packaging temperature affected temperature changes inside the shipping container during transportation, we compared changes in shipping parameters by packing temperature (Supplemental Table 4). We found that the lowest temperature reached inside the container was statistically significantly higher if the shipment was packed at >20-25°C vs. <20°C (16.5°C ± 1.9 vs.13.5°C ± 4.5; p=0.01). Likewise, the highest temperature reached inside the container was statistically significantly higher if the shipment was packed at >20- 25°C vs. <20°C (23.5°C ± 1.4 vs.21.6°C ± 2.3; p=0.03). 142 Several shipping conditions were examined to determine their influence on temperature drops within the shipping container (Table 3). Ambient, i.e. outside, city temperature was shown to be important (overall p=0.04); in particular, a temperature drop <15°C inside the shipping container was less likely when the outside city temperature was either -13°C to - 1°C(OR=0.1) or >1°C (OR=0.5) when compared to colder environments of -13°C (OR=0.1). The number of Fed- Ex hub stops during the course of a shipment was also found to be statistically significantly (p=0.04); that is, shipments that passed through 2 or 3 hubs were more likely to see within container temperature drops <15°C compared to those only stopping at one hub station (OR=4.5).To determine if a multifactorial influence on drops in container temperature exist, multivariable logistic regression analysis was performed on all univariate factors that were found to be significant at p<0.20. The lowest outside city temperature (p=0.04) and number of Fed-Ex hub stops (p=0.051) were independently shown to be of significance in a multivariable model. Optimization of Gel-Pack Number. Because of the large drop and range in temperature values observed during the preliminary shipping evaluation, a series of optimization experiments were carried out to identify the minimum number of gel packs that could be used to prevent an undesirable decrease in the climate conditions within the shipping container (Supplemental Figure 2). Only the use of 143 6 or 8 gel-packs was able to prevent an undesirable temperature drop below 15°C in both cold conditions tested. Likewise, the use of 6 or 8 gel-packs prevented the undesirable temperature drop in both warm conditions tested. Finally, when 2 different alternating temperature conditions were evaluated using 6 gel-packs, the temperature drop never fell below 15°C. In all optimization experiments performed, the within-container temperature never exceeded 37°C. Validation of Standardized Shipping Container. From February 12, 2004 to December 31, 2010, 3091 qualified shipments of human pancreatic islets were made by 14 laboratories participating in the ICR Basic Science Human Islet Distribution Program (Table 4). In 2 time periods examined, there was a statistically significant change in islet quality rankings (p<0.0001). First, the percentage of shipments ranked as excellent to good increased from 81.9% (n=1254) in the pre-study period to 91.3% (n=1424) after implementation of a standardized shipping protocol. Likewise, the percentage of shipments ranked as fair to poor decreased from 18.1% (n=275) to 8.7% (n=140). Moreover, temperature fluctuations inside the shipping container were also monitored after implementation of a standardized shipping protocol (Table 5). First, when using 6 gelpacks, there was not a statistically significant change in the number of shipments reaching a within-container temperature of <15°C (cold- triggered), >29°C (heat-activated), or maintained at 15-29°C across winter, 144 spring, summer, or fall (p=0.4989, fishers exact test). Next, when comparing the use of 2 vs. 6 gelpacks in the winter season, there was a statistically significant change in the temperature during shipment (p<0.0001). The percentage of shipments that were cold-triggered was reduced from 35.6% (n=16) to 11.3% (n=39). Likewise, the percentage of shipments that maintained a stable temperature improved from 64.4% (n=29) to 88.1% (n=305). The percentage of shipments that were heat-activated increased from 0% (n=0) to 0.6% (n=2). CHAPTER 5.3 DISCUSSION The availability of a standardized human islet shipping protocol was important in demonstrating the feasibility of utilizing remote pancreas processing laboratories for clinical transplantation studies 4 . Clearly, the need for such protocols is growing, necessitated by the increasing number of long distance basic science and clinical collaborations 1,31-41 . However, the use of different transportation methods, such as commercial carriers 32,34 , charter jet 4 , or ambulance 37,38 , makes the widespread adaptation of any single protocol impractical. This report details the development and validation of a standardized shipping protocol used by the ICR Consortium in 1602 human islet shipments delivered to diabetes investigators using a single commercial carrier. Our preliminary evaluation of 48 shipments using standardized materials and operating procedures revealed a number of issues that needed to be addressed 145 prior to widespread implementation of a final shipping protocol. First, we observed that packing preparations affected differences in extreme within- container temperatures during shipment. In particular, depending on if the packing temperature was <21°C vs >21°C, as much as a 3°C difference existed in the mean high and low temperatures. Each package contained 2 ambient temperature stabilization gel-packs. A phase change in each gel-pack, from a frozen solid to liquid slushy consistency, occurs at 19.6°C±1°C (personal communication with SEBRA). It is therefore likely that when the gel packs remained in the frozen state, the contents of the container remained cooler in hot and cold conditions, a suggestion consistent with our data. It should be noted, however, that packing temperature alone was not associated with a drop in within-container temperature during shipment. Another factor identified was the number of gel-packs included inside the container, as the SEBRA manual states that ambient temperatures below 10°C or above 38°C may require additional stabilizers. In fact, ambient city temperatures outside of the container ranged from -19.4°C to 31.7°C, with 36 of 48 shipments (75%) exposed to lows of <1°C. Temperature highs above 38°C were not observed, as our initial experiments were performed in the winter months only, a potential limitation of this study. Next, the use of a commercial carrier resulted in as many as 3 hub-stops prior to package delivery and that temperature inside the shipping container was 4.5 times more likely to drop <15°C if greater than 1 hub-stop was made. This may be explained, in part, by the fact that packages shipped using a commercial 146 carrier are placed in a pressurized, but not temperature regulated, cargo carrier (personal communication with Fed-Ex) and therefore vulnerable to extreme changes in ambient temperature. Indeed, we found that if the lowest departing, arriving, and hub-stop city temperatures were warmer, there was a progressive decline in the odds that a temperature drop under or at 15°C will occur. Finally, although the mean overall, minimum, and maximum temperatures during shipment (Table 2) were within 1.5°C of the values shown to be acceptable by Ichii and colleagues 4 , the maximum change range seen was 18.9°C, similar to the undesirable extreme values reported by Ikemoto 6 and Rozak 5 . Moreover, depending on whether reusable loggers or disposable indicators were used, a total of 30% to 36% of all shipments, respectively, reached a temperature low that dropped at or below 15°C. We therefore sought to abolish temperature drop occurrences <15°C by optimizing the number of gel packs included inside the standardized shipping container, as it had been previously reported that the use of additional packs stabilized temperature during blood product transport 42 . This was done in a controlled environment using a number of different testing conditions. It was determined that a minimum of 6 gel-packs were required to eliminate undesirable temperature drops. 147 Having now developed a standardized shipping protocol, implementation of this procedure occurred on November 1, 2007. When we compared islet quality in the 3-year period prior vs. post implementation, we found statistically significant improvements in the excellent to good quality rankings, and reductions in fair to poor outcomes (Table 4). Although this suggests that the use of this protocol positively affected islet quality at the receiving laboratory, rankings may have also been attributable to ICR shipping experience over time, such that improvements may have occurred independently of standard materials and operating procedures. While our data cannot rule out this possibility, a number of participating islet laboratories were shipping islets before the creation of the ICR program and thus deemed experienced manufacturers. Nonetheless, staff turnover at each center was not evaluated; therefore, stability of personnel could not be accessed as a surrogate marker of experience. Next, because the islet ranking was subjective, it is conceivable that in the period prior to standard protocol implementation, investigators receiving human islets were less critical of the quality because there was no cost associated with the shipment, i.e. they were paid for by the ICR program. Oppositely, islet rankings in the post- implementation period may represent a more critical review of the quality because a nominal islet subscription fee was instituted by the ICR program in 2009. However, an analysis of rankings by free or paid shipments was not possible due to a major reduction in sample size. 148 Finally, although we observed over a 3 fold reduction in the number of cold- activations when comparing the use of our initial vs. final shipping protocol (Table 5), elimination of extreme temperature drops <15°C was not possible. Moreover, we observed temperature rises >29°C. Despite the fact that there was no statistical difference in the number of hold, cold, or temperature stable shipments throughout the 4 major seasons when using the final shipping protocol, technical factors may have influenced these numbers. For example, under-warming of the gel-packs in the summer months may have falsely activated the disposable cold- indicator tab. Likewise, inappropriate handling of the disposable temperature indicator may have also caused false readings, as it must be refrigerated prior to and warmed immediately before use. We did not examine several other parameters that might also be important in improving the quality of transported pancreatic islets. The choice of shipping culture vessel, such as bag, flask, or tube, has been previously examined 4,6,43 . Active control devises to regulate temperature and pressure fluctuations during shipping of human pancreatic islets have also been proposed 5 . A number of physiological parameters have been suggested 44 . CHAPTER 5.3 ACKNOWLEDGMENTS This work was funded by the cooperative efforts of the NCRR and the NIDDK, a component of the US NIH, in conjunction with the generous contributions of the 149 JDRF. Participating ICR centers included: (1) University of Minnesota, Minneapolis, MN (2001–2009; U42 RR 016598 to B.J. Hering), (2) University of Pennsylvania, Philadelphia, PA (2001–2009; U42 RR 016600 to A. Naji), (3) University of Miami, Miami, FL (2001–2009; U42 RR 016603 to C. Ricordi), (4) City of Hope National Medical Center, Duarte, CA (2001–2009; U42 RR 016607 to F. Kandeel), (5) University of Wisconsin, Madison, WI (2006–2009; U42 RR 023240 to L.A. Fernandez), (6) Chicago Consortium (University of Illinois at Chicago and Northwestern University), Chicago, IL (2006–2009; U42 RR 023245 to J. Oberholzer, including subcontract to Northwestern, sub-PI, D. Kaufman) and (7) University of Alabama, Birmingham, AL (2006–2009; U42 RR 023246 to J. Contreras). The ICR-ABCC is located at the City of Hope National Medical Center (2001–2009; U42 RR 017673 to J.C. Niland). ICR Islet Shipping Standardization Subcommittee members conceived and designed the study. Members included representatives from each ICR laboratory and the ABCC, Alvin Powers, Daniel Rosenblum, and Mike Appel. Experiments were carried out by members of each ICR laboratory and the ABCC. 150 CHAPTER 5.3 REFERENCES 1. Kaddis JS, Olack BJ, Sowinski J, Cravens J, Contreras JL, Niland JC. Human Pancreatic Islets and Diabetes Research. JAMA: The Journal of the American Medical Association. 04/15 2009;301(15):1580-1587. 2. Guignard AP, Oberholzer J, Benhamou P-Y, et al. Cost Analysis of Human Islet Transplantation for the Treatment of Type 1 Diabetes in the Swiss-French Consortium GRAGIL. Diabetes Care. 04/01 2004;27(4):895- 900. 3. Goss JA, Goodpastor SE, Brunicardi FC, et al. Development of a human pancreatic islet-transplant program through a collaborative relationship with a remote islet-isolation center. Transplantation. Feb 15 2004;77(3):462-466. 4. Ichii H, Sakuma Y, Pileggi A, et al. Shipment of human islets for transplantation. Am J Transplant. Apr 2007;7(4):1010-1020. 5. Rozak PR, Weegman BP, Avgoustiniatos ES, et al. Devices and methods for maintenance of temperature and pressure during islet shipment. Transplant Proc. Mar 2008;40(2):407-410. 6. Ikemoto T, Matsumoto S, Itoh T, et al. Assessment of islet quality following international shipping of more than 10,000 km. Cell Transplant. 2010;19(6):731-741. 7. Qi M, Strand BL, Morch Y, et al. Encapsulation of human islets in novel inhomogeneous alginate-ca2+/ba2+ microbeads: in vitro and in vivo function. Artif Cells Blood Substit Immobil Biotechnol. 2008;36(5):403-420. 8. Vaithilingam V, Barbaro B, Oberholzer J, Tuch BE. Functional Capacity of Human Islets After Long-Distance Shipment and Encapsulation. Pancreas. Oct 13 2010. 9. Murdoch TB, McGhee-Wilson D, Shapiro AM, Lakey JR. Methods of human islet culture for transplantation. Cell Transplant. 2004;13(6):605- 617. 151 10. Hunt L, Hacker DL, Grosjean F, et al. Low-temperature pausing of cultivated mammalian cells. Biotechnology and Bioengineering. 2005;89(2):157-163. 11. Rauen U, Polzar B, Stephan H, Mannherz HG, Groot HD. Cold-induced apoptosis in cultured hepatocytes and liver endothelial cells: mediation by reactive oxygen species. The FASEB Journal. January 1, 1999;13(1):155- 168. 12. Rieder CL, Cole RW. Cold-shock and the Mammalian cell cycle. Cell cycle (Georgetown, Tex. May-Jun 2002;1(3):169-175. 13. Sakurai T, Itoh K, Liu Y, et al. Low temperature protects mammalian cells from apoptosis initiated by various stimuli in vitro. Experimental Cell Research. 2005;309(2):264-272. 14. Brandhorst D, Hammes HP, Brandhorst H, et al. Pretransplant induction of HSP-70 in isolated adult pig islets decreases early islet xenograft survival. Cell Transplant. May-Jun 2000;9(3):423-430. 15. Brandhorst D, Brandhorst H, Kumarasamy V, et al. Hyperthermic preconditioning protects pig islet grafts from early inflammation but enhances rejection in immunocompetent mice. Cell Transplant. 2003;12(8):859-865. 16. Brandhorst H, Olbrich M, Neumann A, Jahr H, Brandhorst D. Effect of pretransplant preconditioning by whole body hyperthermia on islet graft survival. Cell Transplant. 2007;16(7):707-715. 17. Hagiwara S, Iwasaka H, Matsumoto S, Noguchi T. Changes in cell culture temperature alter release of inflammatory mediators in murine macrophagic RAW264.7 cells. Inflammation Research. 2007;56(7):297- 303. 18. Schramm R, Scheuer C, Yamauchi J, Vollmar B, Menger MD. Heat shock preconditioning impairs revascularization of freely transplanted pancreatic islets1. Transplantation. 2002;74(11):1544-1550. 152 19. Scarim AL, Heitmeier MR, Corbett JA. Heat Shock Inhibits Cytokine- Induced Nitric Oxide Synthase Expression by Rat and Human Islets. Endocrinology. December 1, 1998;139(12):5050-5057. 20. Welsh N, Margulis B, Borg LA, et al. Differences in the expression of heat- shock proteins and antioxidant enzymes between human and rodent pancreatic islets: implications for the pathogenesis of insulin-dependent diabetes mellitus. Mol Med. Nov 1995;1(7):806-820. 21. Al-fageeh MB, Smales CM. Control and regulation of the cellular responses to cold shock: the responses in yeast and mammalian systems. Biochem J. July 15, 2006;397(2):247-259. 22. Richter K, Haslbeck M, Buchner J. The Heat Shock Response: Life on the Verge of Death. Molecular Cell. 2010;40(2):253-266. 23. Sonna LA, Fujita J, Gaffin SL, Lilly CM. Invited Review: Effects of heat and cold stress on mammalian gene expression. Journal of Applied Physiology. April 1, 2002;92(4):1725-1742. 24. Spriggs KA, Bushell M, Willis AE. Translational Regulation of Gene Expression during Conditions of Cell Stress. Molecular Cell. 2010;40(2):228-237. 25. Hering BJ, Kandaswamy R, Harmon JV, et al. Transplantation of Cultured Islets from Two-Layer Preserved Pancreases in Type 1 Diabetes with Anti-CD3 Antibody. American Journal of Transplantation. 2004;4(3):390- 401. 26. Hering BJ, Kandaswamy R, Ansite JD, et al. Single-Donor, Marginal-Dose Islet Transplantation in Patients With Type 1 Diabetes. JAMA: The Journal of the American Medical Association. February 16, 2005 ;293(7):830-835. 27. Froud T, Ricordi C, Baidal DA, et al. Islet Transplantation in Type 1 Diabetes Mellitus Using Cultured Islets and Steroid-Free Immunosuppression: Miami Experience. American Journal of Transplantation. 2005;5(8):2037-2046. 153 28. Scharp DW, Lacy PE, Finke E, Olack B. Low-temperature culture of human islets isolated by the distention method and purified with Ficoll or Percoll gradients. Surgery. Nov 1987;102(5):869-879. 29. Ricordi C, Lacy PE, Sterbenz K, Davie JM. Low-temperature culture of human islets or in vivo treatment with L3T4 antibody produces a marked prolongation of islet human-to-mouse xenograft survival. Proc Natl Acad Sci U S A. Nov 1987;84(22):8080-8084. 30. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. Mar 1977;33(1):159-174. 31. Tibell A, Bolinder J, Hagström-Toft E, et al. Experience with human islet transplantation in Sweden. Transplantation Proceedings. 2001;33(4):2535-2536. 32. Rabkin JM, Olyaei AJ, Orloff SL, et al. Distant processing of pancreas islets for autotransplantation following total pancreatectomy. Am J Surg. May 1999;177(5):423-427. 33. Rabkin JM, Leone JP, Sutherland DE, et al. Transcontinental shipping of pancreatic islets for autotransplantation after total pancreatectomy. Pancreas. Nov 1997;15(4):416-419. 34. Mathe Z, Langer RM, Jaray J, et al. Human Pancreatic Islet Transplantation in International Collaboration With A Distant Islet Isolation Center: Preliminary Results From the Budapest-Geneva Experience. Transplantation. 2004;78(2):353. 35. Lundgren T, Korsgren O, Foss A, et al. The Nordic Network for Clinical Islet Transplantation - Islet After Kidney (Iak) - Results and Experiences From the First 44 Transplantations. Transplantation. 2004;78(2):178. 36. Langer RM, Mathe Z, Doros A, et al. Successful islet after kidney transplantations in a distance over 1000 kilometres: Preliminary results of the Budapest-Geneva collaboration. Transplant Proc. Dec 2004;36(10):3113-3115. 154 37. Kessler L, Bucher P, Milliat-Guittard L, et al. Influence of islet transportation on pancreatic islet allotransplantation in type 1 diabetic patients within the Swiss-French GRAGIL network. Transplantation. Apr 27 2004;77(8):1301-1304. 38. Kempf MC, Andres A, Morel P, et al. Logistics and transplant coordination activity in the GRAGIL Swiss-French multicenter network of islet transplantation. Transplantation. May 15 2005;79(9):1200-1205. 39. Goss JA, Schock AP, Brunicardi FC, et al. Achievement of insulin independence in three consecutive type-1 diabetic patients via pancreatic islet transplantation using islets isolated at a remote islet isolation center. Transplantation. Dec 27 2002;74(12):1761-1766. 40. Brunicardi FC, Atiya A, Stock P, et al. Clinical islet transplantation experience of the University of California Islet Transplant Consortium. Surgery. Dec 1995;118(6):967-971; discussion 971-962. 41. Benhamou PY, Oberholzer J, Toso C, et al. Human islet transplantation network for the treatment of Type I diabetes: first data from the Swiss- French GRAGIL consortium (1999-2000). Groupe de Recherche Rhin Rhjne Alpes Geneve pour la transplantation d'Ilots de Langerhans. Diabetologia. Jul 2001;44(7):859-864. 42. Jansen J, Nolan PL, Reeves MI, et al. Transportation of peripheral blood progenitor cell products: effect of ambient temperature. Cytotherapy. 2010;12(7):919-923. 43. Avgoustiniatos ES, Hering BJ, Rozak PR, et al. Commercially available gas-permeable cell culture bags may not prevent anoxia in cultured or shipped islets. Transplant Proc. Mar 2008;40(2):395-400. 44. Fraker C, Montelongo J, Szust J, Khan A, Ricordi C. The use of multiparametric monitoring during islet cell isolation and culture: a potential tool for in-process corrections of critical physiological factors. Cell Transplant. 2004;13(5):497-502. 155 Run 1 A Run 2 Temperature Monitors Number tested 3 12 Testing period ( in hours) 94.5 45.9 Number of readings per monitor 2836 (8,508 total) 1378 (16,536 total) Time between readings (in min.) 2 2 Mean ±SD Min, Max Mean ±SD Min, Max Between-monitor Pearson, r B 0.97±0.01 0.95, 0.98 0.93±0.05 0.82,1.00 Max difference from mean (°C) C 0.49±0.32 0.00, 1.35 0.98±0.49 0.19, 2.13 Coefficient of Variation (%) D 3.20±2.22 0.00, 9.73 3.74±2.01 0.56, 8.87 A Two temperature monitors had incomplete data and were not included. B Reproducibility statistic, defined in materials and methods C Precision statistic, defined in materials and methods D Calculated as standard deviation over the mean for each reading across all monitors used in a run Supplemental Table 1. Reproducibility and Precision of Temperature Monitors Used in Study. Reliability of probes used in this study was determined over 2 trial runs for 15 of 17 temperature monitors. 156 Parameters of Interest A No. B (%) Reusable Temperature Data Logger Complete data obtained 46 (96%) Delayed shipment 2 (4%) Disposable Temperature Indicator Complete data obtained 45 (94%) Indicator not included 2 (4%) Indicator not set up correctly 1 (2%) Fed-EX Transportation Information C Number of hub stops prior to delivery 1 37 (77%) 2 10 (21%) 3 1 (2%) Ambient City Temperature D Complete data obtained 48 (100%) A Total of 9 participating sites, i.e. 8 laboratories and 1 coordinating center B Number of Shipments (No.) C Other variables collected included departure/arrival times and cities, and total miles traveled. D Departing, arriving, and all hub-stop city temperatures collected using data provided by www.weather.com Table 1. Description of Shipment Data Obtained. Detailed information on temperature and transportation of standardized islet shipping containers was collected for this study. 157 Packed At B Temperature Range (°C) During Packing of Shipment A ICR Centers 1 2 34 56 <21.0°C No. (%) >21.0- 26.0°C No. (%) Northwestern U. 18.4- 20.5 20.4- 21.3 14.6- 19.5 6.9- 19.4 18.8- 19.2 18.4- 19.0 5 (83%) 1 (17%) U. Miami C 15.5- 17.5 ND D 17.7- 18.3 15.7- 17.1 14.0- 16.1 18.9- 19.6 5 (100%) 0 (0%) U. Wisconsin 22.2- 26.3 24.4- 26.6 18.7- 18.7 19.9- 21.2 15.4- 16.8 21.6- 22.9 2 (33%) 4 (67%) SCIC C 10.6- 11.6 ND D 10.7- 23.6 18.7- 20.9 18.0- 19.3 18.9- 22.7 3 (60%) 2 (40%) U. Alabama 22.1- 24.0 21.4- 22.3 22.7- 23.4 21.9- 22.2 22.7- 24.6 22.8- 23.0 0 (0%) 6 (100%) U. Minnesota 24.5- 24.9 20.2- 24.2 18.9- 19.5 25.5- 25.8 21.7- 22.2 22.6- 23.1 1 (17%) 5 (83%) UIC 21.9- 22.9 21.7- 21.9 20.1- 20.9 21.5- 22.0 19.9- 20.9 16.9- 18.9 3 (50%) 3 (50%) U. Pennsylvania 20.1- 20.6 21.7- 22.8 20.9- 22.6 20.9- 21.9 20.8- 22.4 20.6- 21.5 1 (17%) 5 (83%) Total -- -- -- -- -- -- 20 (43%) 26 (57%) A Packing period defined as the hour prior to Fed-Ex pickup. B Determined using median value when temperature range did not exclusively fall within either category. C Percentages based on 5 shipments. D No data (ND) obtained for 2 delayed shipments. Supplemental Table 2. Temperature During Packing of Islet Shipping Containers. Instructions to ICR laboratories specified containers to be packaged at room temperature. Packing temperatures within the 21-26°C range met this criterion. 158 No. A Mean SD Median Min Max Temperature w/in Shipping Container B Overall (°C) 46 Lowest value reached (°C) 46 15.2 3.6 16.4 4.7 18.7 Highest value reached (°C) 46 22.7 2.1 22.9 15.7 26.6 Change range (°C) 46 7.5 3.8 6.8 2.1 18.9 Fastest change rate (°C/minute) 46 0.32 0.24 0.29 0.05 1.29 Ambient (Outside) City Temperature Lowest value reached (°C) 48 -4.6 10.3 -6.1 -19.4 16.1 Highest value reached (°C) 48 20.9 4.8 20.6 15.0 31.7 Change range (°C) 48 25.6 12.0 23.1 2.2 48.3 Shipping Time C 46 27.3 8.5 25.3 21.7 71.1 On time delivery (20-28 hours) 42 25.1 1.4 25.2 21.7 27.9 Delayed (>28 hours) 4 49.9 17.5 50.0 28.5 71.1 Shipping Distance (miles) 48 1851 726 2016 38 2544 A Number of Shipments (No.) with data B Monitored using reusable temperature data loggers B Defined from the estimated time of packing, one hour prior to Fed-Ex pickup, to arrival at the data coordinating center in Duarte, CA. Table 2. Changes in Shipping Parameters During Transportation of Islet Containers. 159 160 Supplemental Figure 1. Temperature Profiles During Cross Country Shipments Using Standardized Islet Shipping Containers. All shipments originated from geographically distinct Universities, as indicated above, and centrally delivered to the data coordinating center at City of Hope in Duarte, CA. Shipments from the Southern California Islet Consortium (SCIC), at City of Hope, were held at the central Fed-Ex hub in Los Angeles, CA and did not involve cross country transportation. First dotted line at -1 hour represents packaging begin time. Second dotted line at 0 hours signifies Fed-Ex pick up time. University of Illinois at Chicago abbreviated UIC. 161 Reusable Temperature Data Logger B Disposable Temperature Indicator A <15 °C reached Between 15 to 29 °C >29 °C reached No data Total <15 °C reached 11 4 0 1 16 Between 15 to 29 °C 3250 129 >29 °C reached 000 00 No data 030 03 Total 14 32 0 2 48 A Disposable temperature indicators are cold activated at or below 15°C, not activated between 15-29°C, and warm activated at or greater than 29°C B Coding of continuous to categorical data for reusable temperature data logger was based on disposable indicator groupings. For values falling between 15- 16°C, < 15.44°C was rounded down to 15 and >15.44°C rounded up to 16. Supplemental Table 3. Two-Way Agreement Between Temperature Monitoring Methods. Assessment was performed from the time of Fed-Ex pickup to arrival at the data coordinating center in Duarte, CA. 162 No. Mean SD Med Min Max p-value A Temperature w/in Container Lowest temp. reached (°C) if packed >20-25 °C 25 16.5 1.9 17.1 12.2 18.7 0.01 if packed <20 °C 20 13.5 4.5 15.2 4.7 18.6 Highest temp. reached (°C) if packed >20-25 °C 25 23.5 1.4 23.1 20.6 26.6 0.03 if packed <20 °C 20 21.6 2.3 21.8 15.7 24.5 Range of temp. change (°C) if packed >20-25 °C 25 7 2.4 6.7 3.5 13.8 0.42 if packed <20 °C 20 8 5.2 6.8 2.1 18.9 Fastest change rate (°C/min) if packed >20-25 °C 25 0.28 0.15 0.29 0.05 0.58 0.23 if packed <20 °C 20 0.37 0.32 0.26 0.05 1.29 A Obtained using two-group t-tests Supplemental Table 4. Effects of Packing Islet Shipment Containers at Different Temperatures. Shipping parameters were stratified by packing temperature and analyzed. 163 OR A (95% CI) p-value Temperature at packing (°C) 20-25 (n=25) -- 0.08 <20 (n=20) 3.1 (0.9, 12.0) Lowest outside city temperature (°C) < -13 (n=12) -- 0.04 -13 to 1 (n=24) 0.1 (0.0, 0.7) >1 (n=10) 0.5 (0.1, 2.6) Highest outside city temperature (°C) < 17 (n=10) -- 0.6 17-22 (n=27) 2.4 (0.5, 117.7) > 22 (n=9) 2.0 (0.3, 19.2) Fed-Ex hubs used per shipment 1 (n=36) -- 0.04 2 or 3 (n=10) 4.5 (1.1, 21.3) Shipping distance (miles) < 1500 (n=5) -- 0.47 1500-2300 (n=33) 0.9 (0.1, 7.4) > 2300 (n=11) 0.3 (0.0, 3.7) Shipping time (hours) < 24 -- 0.47 24-28 0.5 (0.1, 2.3) >28 1.3 (0.1, 14.8) A Odds ratios, confidence intervals and p-values were calculated using univariate logistic regression. Dashes indicate baseline category. Table 3. Univariate Logistic Regression Analysis of Shipping Conditions Associated with a Temperature Drop <15°C Inside the Container. 164 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 2 4 6 8 Temp. Outside of Box: -20°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 2 4 6 8 Temp. Outside of Box: 4°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 Temp. Outside of Box: -20 to 37°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 Temp. Outside of Box: 37 to -20°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 8 Temp. Outside of Box: 37°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 2 4 6 8 Temp. Outside of Box: -20°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 2 4 6 8 Temp. Outside of Box: 4°C A: Cold -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 8 Temp. Outside of Box: 37°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 8 Temp. Outside of Box: Room B: Warm -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 Temp. Outside of Box: -20 to 37°C -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40 0 6 Temp. Outside of Box: 37 to -20°C C: Alternating Supplemental Figure 2. Optimization of Internal Temperature Control During Mock 18-Hour Islet Shipments. Standardized islet shipping containers were exposed to A) cold, B) warm, and C) alternating external temperatures for at least 18 hours. Number of ambient temperature stabilization gel packs inside the containers based on the temperature condition being tested. Red dashed line represents the target minimum threshold value of 15°C. 165 Pre-Study Shipments B Standardized Shipments C p-value D Islet Quality A n % n % Excellent to Good 1252 81.9 1424 91.1 Fair to Poor 275 18.1 140 8.9 <0.0001 TOTALS 1527 100 1564 100 A A qualitative assessment performed post-shipment by the islet-receiving laboratory; additional details in materials and methods. B From Feb. 12, 2004 to Oct. 31, 2007; 27 shipments w/missing data excluded C From Nov. 01, 2007 to Dec. 31, 2010; 38 shipments w/missing data excluded D Determined using chi-square test Table 4. Islet Quality Rankings in over 3000 shipments of Human Pancreatic Islets. Perceived islet quality was compared pre and post implementation of a standardized shipping protocol for shipments made by the ICR Basic Science Human Islet Distribution Program. 166 Standardized Shipments B 2 Gelpacks 6 Gelpacks Winter Winter Spring Summer Fall Temperature During Shipment (°C) A n% n % n % n % n % p-value C <15 °C reached 16 35.6 39 11.3 52 12.1 35 9.0 52 13.1 Always 15-29 °C 29 64.4 305 88.1 375 87.0 351 89.7 342 86.4 >29 °C reached 0 0 2 0.6 4 0.9 5 1.3 2 0.5 <0.0001 TOTALS 45 100 346 100 431 100 391 100 396 100 A Monitored using disposable temperature indicators B Shipments were binned into seasons, depending on package pick-up date. Seasons defined using equinox and solstice dates, for 2007-2010, taken from United States Naval Observatory website at: http://www.usno.navy.mil/USNO/astronomical-applications/data-services/earth-seasons C Determined using fisher’s exact test for comparison made between the use of 2 vs. 6 gelpacks in the winter; Significant if p<0.025 using bonferroni correction, see text for additional detail. Table 5. Temperature Control Improvements in Over 1500 Shipments of Human Pancreatic Islets. Temperature fluctuations inside the islet shipping container were evaluated using a standardized shipping protocol and either 2 or 6 ambient temperature gelpacks. 167 CHAPTER 6: CONCLUSION The primary objective of islet-based research is to one day cure diabetes. Transplantation of human pancreatic islets is one form of cellular therapy currently being used experimentally to treat a subset of type 1 diabetics who have unstable disease. Relevant molecular, cellular, and genetic data have been collected from these studies, along with other clinical and laboratory investigations that seek to fill in gaps in our understanding of human islet physiology. Reliable markers of pancreatic beta-cell function in the healthy and diseased state are being intensively investigated to improve therapeutic strategies, including transplantation outcomes. A solid islet manufacturing infrastructure has been developed and is increasingly in demand by diabetes investigators worldwide for both basic science research and clinical transplantation. Responding to this need requires the development and implementation of unbiased and efficient mechanisms for islet matching and distribution to ensure a maximal utilization rate. Some of the major questions raised by our data remain unanswered. First, it was surprising to see that the standard acquisition charge (SAC) for a pancreas used in islet isolation ranged from a low of $600 to a high of $39,800. A SAC fee represents the average cost associated with retrieving each type of organ following donation, and can be influenced by surgeons fees, hospital operating room time, reagents and equipment used, number of organs removed, and the 168 intent of use for each organ 1 . In the case of pancreas, evolving regulatory guidelines may also have added to the variability observed 2 . Second, of 45 variables we examined that might influence the odds of obtaining successful human islet isolation, several identified factors were novel and will require further investigation. The administration of fluid/electrolyte medications in a donor prior to transplant negatively impacted outcome in both a univariate and logistic regression analysis; likewise, hormonal medications received also played a negative role, but only in the univariate analysis. Our analysis of the number and type of medication administered to the donor less than 24 hours prior to donation is the first of its kind in the literature. We regrouped 295 medications into categories based on mechanisms of action. It is therefore possible that certain drugs have antagonistic or synergistics effects not represented in our data. Nonetheless, it will be important to understand the contributions of the liver and overall electrolyte balance in the donor. Moreover, the statistically significant factors we identified have not been shown to be associated with successful transplantation outcome. Next, as we moved to examine a subset of donor islets by microarray, we found that genomic variability existed between samples that could not be correlated with indicators of successful human islet isolation outcome. Unsupervised analysis revealed two distinct cluster groups, whose genomics data have yet to 169 be fully explored, particularly in relation to favorable outcomes. Although a wealth of genetic data exist 3 , a careful analysis will need to be undertaken to apply our current understanding of beta-cell biology to transplant biology. Finally, some have concluded that because insulin independence rates following transplantation are in the long-term lower than that of whole pancreas replacement, the procedure is of limited use. Robertson interestingly demonstrates that when compared to the success rates of pancreas transplantation, islet transplantation is exactly where pancreas replacement was in an earlier era, strongly suggesting room for improvement in this form of cellular replacement therapy 4 . Indeed, the next generation of clinical trials are now underway and are approaching the problems of innate and adaptive immunity and limited engraftment in exciting and new ways 5 . Prior to transplantation, reliable markers of graft performance are needed to better standardize the procedure and have demonstrated therapeutic potential 6 . Ultimately, a deeper understanding of the technical and biological challenges in islet transplantation may one day help to eliminate the disease and its serious complications. 170 CHAPTER 6 REFERENCES 1. Medicare and Medicaid programs; conditions for coverage for organ procurement organizations (OPOs). Final rule. Fed Regist. May 31 2006;71(104):30981-31054. 2. Markmann JF, Kaufman DB, Ricordi C, Schwab PM, Stock PG. Financial issues constraining the use of pancreata recovered for islet transplantation: a white paper. Am J Transplant. Aug 2008;8(8):1588- 1592. 3. Oliver-Krasinski JM, Stoffers DA. On the origin of the β cell. Genes & Development. August 1, 2008;22(15):1998-2021. 4. Robertson RP. Islet Transplantation a Decade Later and Strategies for Filling a Half-Full Glass. Diabetes. 06/01 2010;59(6):1285-1291. 5. Van Belle TL, Coppieters KT, Von Herrath MG. Type 1 Diabetes: Etiology, Immunology, and Therapeutic Strategies. Physiological Reviews. January 1, 2011;91(1):79-118. 6. Caiazzo R, Gmyr V, Kremer B, et al. Quantitative In Vivo Islet Potency Assay in Normoglycemic Nude Mice Correlates With Primary Graft Function After Clinical Transplantation. Transplantation. 2008;86(2):360- 363. 171 BIBLIOGRAPHY 2007 update on allogeneic islet transplantation from the Collaborative Islet Transplant Registry (CITR). Cell Transplant. 2009;18(7):753-767. Abdueva D, Wing M, Schaub B, Triche T, Davicioni E. Quantitative Expression Profiling in Formalin-Fixed Paraffin-Embedded Samples by Affymetrix Microarrays. J Mol Diagn. July 1, 2010;12(4):409-417. Ach RA, Floore A, Curry B, et al. Robust interlaboratory reproducibility of a gene expression signature measurement consistent with the needs of a new generation of diagnostic tools. BMC Genomics. 2007;8:148. Ach RA, Wang H, Curry B. Measuring microRNAs: comparisons of microarray and quantitative PCR measurements, and of different total RNA prep methods. BMC Biotechnol. 2008;8:69. Ader M, Ni TC, Bergman RN. Glucose effectiveness assessed under dynamic and steady state conditions. Comparability of uptake versus production components. J Clin Invest. Mar 15 1997;99(6):1187-1199. Adewola A, Mage R, Hansen M, et al. Comparing cooling systems for the COBE 2991 cell separator used in the purification of human pancreatic islets of Langerhans. Cryo Letters. Jul-Aug 2010;31(4):310-317. Adewola AF, Lee D, Harvat T, et al. Microfluidic perifusion and imaging device for multi-parametric islet function assessment. Biomed Microdevices. Jun 2010;12(3):409-417. Agrawal A, Gurusamy K, Powis S, Gray DW, Fuller B, Davidson BR. A meta- analysis of the impact of the two-layer method of preservation on human pancreatic islet transplantation. Cell Transplant. 2008;17(12):1315-1322. Aguayo-Mazzucato C, Bonner-Weir S. Stem cell therapy for type 1 diabetes mellitus. Nat Rev Endocrinol. 2010;6(3):139-148. 172 Ahren B, Taborsky GJ. Beta-Cell Function and Insulin Secretion. 6th ed. New York: McGraw-Hill; 2003. Alan K, Brian O. [11] Sample Labeling: An Overview. DNA Microarrays, Part A: Array Platforms and Wet-Bench Protocols. Volume 410: Academic Press; 2006:222-237. Alejandro R, Barton FB, Hering BJ, Wease S, Investigators C. 2008 Update From the Collaborative Islet Transplant Registry. Transplantation. 2008;86(12):1783- 1788 American Diabetes A. Standards of Medical Care in Diabetes - 2011. Diabetes Care. January 1, 2011;34(Supplement 1):S11-S61. Amiel SA. "Brittle" diabetes. Bmj. Aug 3 1991;303(6797):260-261. Andersen CL, Jensen JL, Orntoft TF. Normalization of Real-Time Quantitative Reverse Transcription-PCR Data: A Model-Based Variance Estimation Approach to Identify Genes Suited for Normalization, Applied to Bladder and Colon Cancer Data Sets. Cancer Res. August 1, 2004;64(15):5245-5250. Antonello Pileggi LCLICR. Overcoming the Challenges Now Limiting Islet Transplantation. Annals of the New York Academy of Sciences. 2006;1079(Immunology of Diabetes IV: Progress in Our Understanding):383-398. Application Note TaqMan MicroRNA Assays: Endogenous Controls for Real- Time Quantitation of miRNA Using TaqMan MicroRNA Assays. Applied Biosystems Publication 127AP11-01. August 2007:1-8. Avila JG, Tsujimura T, Oberholzer J, et al. Improvement of pancreatic islet isolation outcomes using glutamine perfusion during isolation procedure. Cell Transplant. 2003;12(8):877-881. Badiee A, Eiken HG, Steen VM, Lovlie R. Evaluation of five different cDNA labeling methods for microarrays using spike controls. BMC Biotechnol. Dec 11 2003;3:23. 173 Baidal DA, Faradji RN, Messinger S, et al. Early metabolic markers of islet allograft dysfunction. Transplantation. Mar 15 2009;87(5):689-697. Bak M, Conley L, Hedegaard J, et al. Evaluation of two methods for generating cRNA for microarray experiments from nanogram amounts of total RNA. Anal Biochem. Nov 1 2006;358(1):111-119. Balamurugan AN, Breite AG, Anazawa T, et al. Successful human islet isolation and transplantation indicating the importance of class 1 collagenase and collagen degradation activity assay. Transplantation. Apr 27 2010;89(8):954-961. Balamurugan AN, Chang Y, Bertera S, et al. Suitability of human juvenile pancreatic islets for clinical use. Diabetologia. Aug 2006;49(8):1845-1854. Ballian N, Brunicardi FC. Islet vasculature as a regulator of endocrine pancreas function. World Journal of Surgery. Apr 2007;31(4):705-714. Ballinger WF, Lacy PE. Transplantation of intact pancreatic islets in rats. Surgery. Aug 1972;72(2):175-186. Bammler T, Beyer RP, Bhattacharya S, et al. Standardizing global gene expression analysis between laboratories and across platforms. Nat Methods. May 2005;2(5):351-356. Barbaro B, Salehi P, Wang Y, et al. Improved human pancreatic islet purification with the refined UIC-UB density gradient. Transplantation. Nov 15 2007;84(9):1200-1203. Barnett MJ, Zhai X, LeGatt DF, Cheng SB, Shapiro AM, Lakey JR. Quantitative assessment of collagenase blends for human islet isolation. Transplantation. Sep 27 2005;80(6):723-728. Barry R, Tadayyon M, Green IC. Reproducibility of targeted gene expression measurements in human islets of Langerhans. Biochem Biophys Res Commun. Nov 1 2002;298(3):350-356. 174 Barshes NR, Wyllie S, Goss JA. Inflammation-mediated dysfunction and apoptosis in pancreatic islet transplantation: implications for intrahepatic grafts. J Leukoc Biol. May 1, 2005 2005;77(5):587-597. Becker C, Hammerle-Fickinger A, Riedmaier I, Pfaffl MW. mRNA and microRNA quality control for RT-qPCR analysis. Methods. Jan 15 2010; 50(4):237-43. Benes V, Castoldi M. Expression profiling of microRNA using real-time quantitative PCR, how to use it and what is available. Methods. Apr 2010;50(4):244-249. Benhamou PY, Oberholzer J, Toso C, et al. Human islet transplantation network for the treatment of Type I diabetes: first data from the Swiss-French GRAGIL consortium (1999-2000). Groupe de Recherche Rhin Rhjne Alpes Geneve pour la transplantation d'Ilots de Langerhans. Diabetologia. Jul 2001;44(7):859-864. Benhamou PY, Watt PC, Mullen Y, et al. Human islet isolation in 104 consecutive cases. Factors affecting isolation success. Transplantation. Jun 27 1994;57(12):1804-1810. Bergholdt R, Karlsen AE, Hagedorn PH, et al. Transcriptional profiling of type 1 diabetes genes on chromosome 21 in a rat beta-cell line and human pancreatic islets. Genes Immun. Apr 2007;8(3):232-238. Berney T, Bühler LH, Morel P. Pancreas allocation in the era of islet transplantation. Transplant International. 2005;18(7):763-767. Bertuzzi A, Salinari S, Mingrone G. Insulin granule trafficking in Î²-cells: mathematical model of glucose-induced insulin secretion. American Journal of Physiology - Endocrinology And Metabolism. July 1, 2007;293(1):E396-E409. Bertuzzi F, Cainarca S, Marzorati S, et al. Collagenase isoforms for pancreas digestion. Cell Transplant. 2009;18(2):203-206. Bertuzzi F, Ricordi C. Prediction of Clinical Outcome in Islet Allotransplantation. Diabetes Care. February 2007;30(2):410-417. 175 Bertuzzi F, Verzaro R, Provenzano V, Ricordi C. Brittle type 1 diabetes mellitus. Curr Med Chem. 2007;14(16):1739-1744. Bhonde R, Shukla RC, Kanitkar M, Shukla R, Banerjee M, Datar S. Isolated islets in diabetes research. The Indian journal of medical research. Mar 2007;125(3):425-440. Bikopoulos G, da Silva Pimenta A, Lee SC, et al. Ex vivo transcriptional profiling of human pancreatic islets following chronic exposure to monounsaturated fatty acids. J Endocrinol. Mar 2008;196(3):455-464. Blondet JJ, Carlson AM, Kobayashi T, et al. The role of total pancreatectomy and islet autotransplantation for chronic pancreatitis. Surg Clin North Am. Dec 2007;87(6):1477-1501, x. Bolstad BM, Irizarry RA, Astrand M, Speed TP. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics. Jan 22 2003;19(2):185-193. Bonner-Weir S, Weir GC. New sources of pancreatic beta-cells. Nature Biotechnology. Jul 2005;23(7):857-861. Boutayeb A, Chetouani A. A critical review of mathematical models and data used in diabetology. Biomed Eng Online. 2006;5:43. Brandhorst H, Brandhorst D, Hering BJ, Federlin K, Bretzel RG. Body mass index of pancreatic donors: a decisive factor for human islet isolation. Exp Clin Endocrinol Diabetes. 1995;103 Suppl 2:23-26. Brandhorst H, Brandhorst D, Hesse F, et al. Successful human islet isolation utilizing recombinant collagenase. Diabetes. May 2003;52(5):1143-1146. Brandhorst H, Hering BJ, Brandhorst D, et al. Comparison of histidine- tryptophane-ketoglutarate (HTK) and University of Wisconsin (UW) solution for pancreas perfusion prior to islet isolation, culture and transplantation. Transplant Proc. Dec 1995;27(6):3175-3176. 176 Bravo V, Rosero S, Ricordi C, Pastori RL. Instability of miRNA and cDNAs derivatives in RNA preparations. Biochem Biophys Res Commun. Feb 23 2007;353(4):1052-1055. Brendel MD, Hering BJ, Schultz AO, Bretzel RG. International Islet Transplant Registry, Newsletter 9, Vol 8 (No 1), June 2001. Accessed online December 16, 2010 at: http://www.med.uni-giessen.de/itr/newsletter/no_9/news_9.html. Brissova M, Fowler MJ, Nicholson WE, et al. Assessment of human pancreatic islet architecture and composition by laser scanning confocal microscopy. J Histochem Cytochem. Sep 2005;53(9):1087-1097. Brissova M, Shostak A, Shiota M, et al. Pancreatic islet production of vascular endothelial growth factor--a is essential for islet vascularization, revascularization, and function. Diabetes. Nov 2006;55(11):2974-2985. Brunicardi FC, Atiya A, Stock P, et al. Clinical islet transplantation experience of the University of California Islet Transplant Consortium. Surgery. Dec 1995;118(6):967-971; discussion 971-962. Buchanan TA, Watanabe RM, Xiang AH. Limitations in Surrogate Measures of Insulin Resistance. J Clin Endocrinol Metab. November 1, 2010 2010;95(11):4874-4876. Bucher P, Oberholzer J, Bosco D, et al. Microbial surveillance during human pancreatic islet isolation. Transpl Int. May 2005;18(5):584-589. Butler AE, Janson J, Bonner-Weir S, Ritzel R, Rizza RA, Butler PC. Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes. Jan 2003;52(1):102-110. Caballero-Corbalan J, Eich T, Lundgren T, et al. No beneficial effect of two-layer storage compared with UW-storage on human islet isolation and transplantation. Transplantation. Oct 15 2007;84(7):864-869. 177 Cabrera O, Berman DM, Kenyon NS, Ricordi C, Berggren PO, Caicedo A. The unique cytoarchitecture of human pancreatic islets has implications for islet cell function. Proc Natl Acad Sci U S A. Feb 14 2006;103(7):2334-2339. Caiazzo R, Gmyr V, Kremer B, et al. Quantitative In Vivo Islet Potency Assay in Normoglycemic Nude Mice Correlates With Primary Graft Function After Clinical Transplantation. Transplantation. 2008;86(2):360-363 310.1097/TP.1090b1013e31817ef31846. Calin GA, Croce CM. MicroRNA signatures in human cancers. Nat Rev Cancer. 2006;6(11):857-866. Capurso G, Crnogorac-Jurcevic T, Milione M, et al. Peanut-like 1 (septin 5) gene expression in normal and neoplastic human endocrine pancreas. Neuroendocrinology. 2005;81(5):311-321. Cartwright A, Wallymahmed M, Macfarlane IA, Wallymahmed A, Williams G, Gill GV. The outcome of brittle type 1 diabetes - a 20 year study. QJM. 2011. Centers for Disease Control and Prevention. National diabetes fact sheet: general information and national estimates on diabetes in the United States, 2007. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2008. Chambers KT, Unverferth JA, Weber SM, Wek RC, Urano F, Corbett JA. The role of nitric oxide and the unfolded protein response in cytokine-induced beta- cell death. Diabetes. Jan 2008;57(1):124-132. Chen J, Lozach J, Garcia EW, et al. Highly sensitive and specific microRNA expression profiling using BeadArray technology. Nucleic Acids Res. Aug 2008;36(14):e87. Chen Y, Gelfond J, McManus L, Shireman P. Reproducibility of quantitative RT- PCR array in miRNA expression profiling and comparison with microarray analysis. BMC Genomics. 2009;10(1):407. 178 Chen ZL, Vairaktarakis GL. Integrated Scheduling of Production and Distribution Operations. Management Science. 2005;51(4):614-628. Christensen K, Murray JC. What Genome-wide Association Studies Can Do for Medicine. N Engl J Med. March 15, 2007 2007;356(11):1094-1097. Cobelli C, Toffolo G, Ferrannini E. A model of glucose kinetics and their control by insulin, compartmental and noncompartmental approaches. Math Biosci. 1984;72:291-315. Cohen MA, Lee HL. Strategic Analysis of Integrated Production-Distribution Systems: Models and Methods. OPERATIONS RESEARCH. March 1, 1988 1988;36(2):216-228. Cole K, Truong V, Barone D, McGall G. Direct labeling of RNA with multiple biotins allows sensitive expression profiling of acute leukemia class predictor genes. Nucleic Acids Res. 2004;32(11):e86. Collaborative Islet Transplant Registry (CITR) 6th Annual Report, 2009. https://web.emmes.com/study/isl/reports/CITR%206th%20Annual%20Data%20R eport%20120109.pdf. Last accessed 03/11/2011: Conti DV, Cortessis V, Molitor J, Thomas DC. Bayesian modeling of complex metabolic pathways. Hum Hered. 2003;56(1-3):83-93. Cope L, Hartman SM, Gohlmann HW, Tiesman JP, Irizarry RA. Analysis of Affymetrix GeneChip data using amplified RNA. Biotechniques. Feb 2006;40(2):165-166, 168, 170. Cortessis V, Thomas DC. Toxicokinetic genetics: an approach to gene- environment and gene-gene interactions in complex metabolic pathways. IARC Sci Publ. 2004(157):127-150. Couri CE, Oliveira MC, Stracieri AB, et al. C-peptide levels and insulin independence following autologous nonmyeloablative hematopoietic stem cell transplantation in newly diagnosed type 1 diabetes mellitus. Jama. Apr 15 2009;301(15):1573-1579. 179 Couto FM, Minn AH, Pise-Masison CA, et al. Exenatide blocks JAK1-STAT1 in pancreatic beta cells. Metabolism. Jul 2007;56(7):915-918. Cras-Meneur C, Inoue H, Zhou Y, et al. An expression profile of human pancreatic islet mRNAs by Serial Analysis of Gene Expression (SAGE). Diabetologia. Feb 2004;47(2):284-299. Creighton CJ, Reid JG, Gunaratne PH. Expression profiling of microRNAs by deep sequencing. Brief Bioinform. Sep 2009;10(5):490-497. Cryer PE, Davis SN, Shamoon H. Hypoglycemia in Diabetes. Diabetes Care. June 1, 2003;26(6):1902-1912. Cure P, Pileggi A, Froud T, et al. Improved metabolic control and quality of life in seven patients with type 1 diabetes following islet after kidney transplantation. Transplantation. Mar 27 2008;85(6):801-812. Curtis RK, Oresic M, Vidal-Puig A. Pathways to the analysis of microarray data. Trends in Biotechnology. 2005;23(8):429-435. David FN. The moments of the z and F distributions. Biometrika. 1949;36(3- 4):394 - 403. DCCT. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N Engl J Med. Sep 30 1993;329(14):977-986. Doleshal M, Magotra AA, Choudhury B, Cannon BD, Labourier E, Szafranska AE. Evaluation and validation of total RNA extraction methods for microRNA expression analyses in formalin-fixed, paraffin-embedded tissues. J Mol Diagn. May 2008;10(3):203-211. Domínguez-Gil B, Delmonico FL, Shaheen FAM, et al. The critical pathway for deceased donation: reportable uniformity in the approach to deceased donation. Transplant International. 2011;24(4):373-378. 180 Dor Y, Brown J, Martinez OI, Melton DA. Adult pancreatic beta-cells are formed by self-duplication rather than stem-cell differentiation. Nature. May 6 2004;429(6987):41-46. Dufrane D, D'Hoore W, Goebbels RM, Saliez A, Guiot Y, Gianello P. Parameters favouring successful adult pig islet isolations for xenotransplantation in pig-to- primate models. Xenotransplantation. May 2006;13(3):204-214. Easton DF, Pooley KA, Dunning AM, et al. Genome-wide association study identifies novel breast cancer susceptibility loci. Nature. Jun 2007;447(7148):1087-U1087. Efrat S. Beta-cell replacement for insulin-dependent diabetes mellitus. Advanced drug delivery reviews. Jan 14 2008;60(2):114-123. Eizirik DL, Pipeleers DG, Ling Z, Welsh N, Hellerstrom C, Andersson A. Major species differences between humans and rodents in the susceptibility to pancreatic beta-cell injury. Proc Natl Acad Sci U S A. Sep 27 1994;91(20):9253- 9256. Elliott RB, Escobar L, Garkavenko O, et al. No evidence of infection with porcine endogenous retrovirus in recipients of encapsulated porcine islet xenografts. Cell Transplant. Nov-Dec 2000;9(6):895-901. Feldman AL, Costouros NG, Wang E, et al. Advantages of mRNA amplification for microarray analysis. Biotechniques. 2002;33(4):906 - 912. Ferrannini E, Smith JD, Cobelli C, Toffolo G, Pilo A, DeFronzo RA. Effect of insulin on the distribution and disposition of glucose in man. J Clin Invest. Jul 1985;76(1):357-364. Fiedor P, Goodman ER, Sung RS, Czerwinski J, Rowinski W, Hardy MA. The effect of clinical and biochemical donor parameters on pancreatic islet isolation yield from cadaveric organ donors. Ann Transplant. 1996;1(1):59-62. Fiorina P, Shapiro AM, Ricordi C, Secchi A. The clinical impact of islet transplantation. Am J Transplant. Oct 2008;8(10):1990-1997. 181 Fisher RA. The correlation between relatives on the supposition of Mendelian inheritance. Trans Royal Soc Edinburgh. 1918;52(399-433). Flamez D, Roland I, Berton A, et al. A genomic-based approach identifies FXYD domain containing ion transport regulator 2 (FXYD2)gammaa as a pancreatic beta cell-specific biomarker. Diabetologia. Jul 2010;53(7):1372-1383. Fraker CA, Alejandro R, Ricordi C. Use of oxygenated perfluorocarbon toward making every pancreas count. Transplantation. Dec 27 2002;74(12):1811-1812. Francois P, Bento M, Vaudaux P, Schrenzel J. Comparison of fluorescence and resonance light scattering for highly sensitive microarray detection of bacterial pathogens. J Microbiol Methods. Dec 2003;55(3):755-762. Francois P, Garzoni C, Bento M, Schrenzel J. Comparison of amplification methods for transcriptomic analyses of low abundance prokaryotic RNA sources. J Microbiol Methods. Feb 2007;68(2):385-391. Frank A, Deng S, Huang X, et al. Transplantation for type I diabetes: comparison of vascularized whole-organ pancreas with isolated pancreatic islets. Ann Surg. Oct 2004;240(4):631-640; discussion 640-633. Franke-Whittle IH, Klammer SH, Mayrhofer S, Insam H. Comparison of different labeling methods for the production of labeled target DNA for microarray hybridization. J Microbiol Methods. Apr 2006;65(1):117-126. Gaglia JL, Shapiro AM, Weir GC. Islet transplantation: progress and challenge. Arch Med Res. May-Jun 2005;36(3):273-280. Gao X, Gulari E, Zhou X. In situ synthesis of oligonucleotide microarrays. Biopolymers. 2004;73(5):579 - 596. Gill GV. The spectrum of brittle diabetes. J R Soc Med. May 1992;85(5):259-261. Gill GV, Lucas S, Kent LA. Prevalence and characteristics of brittle diabetes in Britain. QJM. Nov 1996;89(11):839-843. 182 Git A, Dvinge H, Salmon-Divon M, et al. Systematic comparison of microarray profiling, real-time PCR, and next-generation sequencing technologies for measuring differential microRNA expression. RNA. May 2010;16(5):991-1006. Goss JA, Goodpastor SE, Brunicardi FC, et al. Development of a human pancreatic islet-transplant program through a collaborative relationship with a remote islet-isolation center. Transplantation. Feb 15 2004;77(3):462-466. Goss JA, Schock AP, Brunicardi FC, et al. Achievement of insulin independence in three consecutive type-1 diabetic patients via pancreatic islet transplantation using islets isolated at a remote islet isolation center. Transplantation. Dec 27 2002;74(12):1761-1766. Goto M, Eich TM, Felldin M, et al. Refinement of the automated method for human islet isolation and presentation of a closed system for in vitro islet culture. Transplantation. Nov 15 2004;78(9):1367-1375. Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ. miRBase: tools for microRNA genomics. Nucl. Acids Res. January 11, 2008;36(suppl_1):D154-158. Groth CG, Korsgren O, Tibell A, et al. Transplantation of porcine fetal pancreas to diabetic patients. Lancet. Nov 19 1994;344(8934):1402-1404. Gudmundsson J, Sulem P, Manolescu A, et al. Genome-wide association study identifies a second prostate cancer susceptibility variant at 8q24. Nature Genetics. May 2007;39(5):631-637. Guignard AP, Oberholzer J, Benhamou P-Y, et al. Cost Analysis of Human Islet Transplantation for the Treatment of Type 1 Diabetes in the Swiss-French Consortium GRAGIL. Diabetes Care. 04/01 2004;27(4):895-900. Gunton JE, Kulkarni RN, Yim S, et al. Loss of ARNT/HIF1beta mediates altered gene expression and pancreatic-islet dysfunction in human type 2 diabetes. Cell. Aug 12 2005;122(3):337-349. 183 Guthrie JL, Seah C, Brown S, Tang P, Jamieson F, Drews SJ. Use of Bordetella pertussis BP3385 to establish a cutoff value for an IS481-targeted real-time PCR assay. J Clin Microbiol. 2008;46(11):3798 - 3799. Halban PA, German MS, Kahn SE, Weir GC. Current Status of Islet Cell Replacement and Regeneration Therapy. J Clin Endocrinol Metab. March 1, 2010 2010;95(3):1034-1043. Hammerle-Fickinger A, Riedmaier I, Becker C, Meyer HH, Pfaffl MW, Ulbrich SE. Validation of extraction methods for total RNA and miRNA from bovine blood prior to quantitative gene expression analyses. Biotechnol Lett. Jan 2010;32(1):35-44. Hanley NA, Hanley KP, Miettinen PJ, Otonkoski T. Weighing up beta-cell mass in mice and humans: self-renewal, progenitors or stem cells? Molecular and cellular endocrinology. Jun 25 2008;288(1-2):79-85. Hanley SC, Paraskevas S, Rosenberg L. Donor and isolation variables predicting human islet isolation success. Transplantation. Apr 15 2008;85(7):950-955. Hanson MS, Park EE, Sears ML, et al. A simplified approach to human islet quality assessment. Transplantation. May 27 2010;89(10):1178-1188. Hardikar AA. Generating new pancreas from old. Trends Endocrinol Metab. Jul 2004;15(5):198-203. Hare JM, Traverse JH, Henry TD, et al. A randomized, double-blind, placebo- controlled, dose-escalation study of intravenous adult human mesenchymal stem cells (prochymal) after acute myocardial infarction. J Am Coll Cardiol. Dec 8 2009;54(24):2277-2286. Hering B, Parkay J, Kandaswamy R, et al. Long-Term Survival of Islet Allografts in Type 1 Diabetes. American Journal of Transplantation. 2007;Supp s2 (7) (s2):205. 184 Hering BJ, Matsumoto I, Sawada T, et al. Impact of two-layer pancreas preservation on islet isolation and transplantation. Transplantation. Dec 27 2002;74(12):1813-1816. Hohmeier HE, Newgard CB. Cell lines derived from pancreatic islets. Molecular and Cellular Endocrinology. Dec 30 2004;228(1-2):121-128. Hohmeier HE, Newgard CB. Islets for all? Nature biotechnology. Oct 2005;23(10):1231-1232. Hopper JL. Variance components for statistical genetics: applications in medical research to characteristics related to human diseases and health. Stat Methods Med Res. 1993;2(3):199-223. Hosmer DW, Hosmer T, Cessie SL, Lemeshow S. A Comparison of Goodness- of-Fit Tests for the Logistic Regression Model. Statistics in Medicine. 1997;16(9):965-980. Hosmer DW, Lemeshow S. Applied Logistic Regression. 2nd ed. New York: Wiley; 2000. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat. Protocols. 2008;4(1):44-57. Huang X, Moore DJ, Ketchum RJ, et al. Resolving the Conundrum of Islet Transplantation by Linking Metabolic Dysregulation, Inflammation, and Immune Regulation. Endocr Rev. August 1, 2008;29(5):603-630. Hubert T, Gmyr V, Arnalsteen L, et al. Influence of preservation solution on human islet isolation outcome. Transplantation. Feb 15 2007;83(3):270-276. Hubert T, Strecker G, Gmyr V, et al. Acute insulin response to arginine in deceased donors predicts the outcome of human islet isolation. Am J Transplant. Apr 2008;8(4):872-876. 185 Hucking K, Watanabe RM, Stefanovski D, Bergman RN. OGTT-derived Measures of Insulin Sensitivity Are Confounded by Factors Other Than Insulin Sensitivity Itself. Obesity. 2008;16(8):1938-1945. Hughes SJ, McShane P, Contractor HH, Gray DW, Clark A, Johnson PR. Comparison of the collagen VI content within the islet-exocrine interface of the head, body, and tail regions of the human pancreas. Transplant Proc. Oct 2005;37(8):3444-3445. Hui H, Wang C, Li H, et al. Gene expression profiling of cultured human islet preparations. Diabetes Technology & Therapeutics. Aug 2004;6(4):481-492. Hultcrantz M, Huhn MH, Wolf M, et al. Interferons induce an antiviral state in human pancreatic islet cells. Virology. Oct 10 2007;367(1):92-101. Hung RJ, Brennan P, Malaveille C, et al. Using hierarchical modeling in genetic association studies with multiple markers: application to a case-control study of bladder cancer. Cancer Epidemiol Biomarkers Prev. Jun 2004;13(6):1013-1021. Hunter DJ, Kraft P. Drinking from the Fire Hose -- Statistical Issues in Genomewide Association Studies. N Engl J Med. July 18, 2007: 357(5):436-9. Ibberson D, Benes V, Muckenthaler MU, Castoldi M. RNA degradation compromises the reliability of microRNA expression profiling. BMC Biotechnol. 2009;9:102. Ichii H, Inverardi L, Pileggi A, et al. A novel method for the assessment of cellular composition and beta-cell viability in human islet preparations. Am J Transplant. Jul 2005;5(7):1635-1645. Ichii H, Pileggi A, Molano RD, et al. Rescue purification maximizes the use of human islet preparations for transplantation. American Journal of Transplantation. Jan 2005;5(1):21-30. Ichii H, Ricordi C. Current status of islet cell transplantation. J Hepatobiliary Pancreat Surg. 2009;16(2):101-112. 186 Ichii H, Sakuma Y, Pileggi A, et al. Shipment of human islets for transplantation. Am J Transplant. Apr 2007;7(4):1010-1020. Ichii H, Wang X, Messinger S, et al. Improved human islet isolation using nicotinamide. Am J Transplant. Sep 2006;6(9):2060-2068. Igoillo-Esteve M, Marselli L, Cunha DA, et al. Palmitate induces a pro- inflammatory response in human pancreatic islets that mimics CCL2 expression by beta cells in type 2 diabetes. Diabetologia. Jul 2010;53(7):1395-1405. Ihm SH, Matsumoto I, Sawada T, et al. Effect of donor age on function of isolated human islets. Diabetes. May 2006;55(5):1361-1368. Institute for Scientific Information (ISI) Web of Knowledge Website. Web of Science and Journal Impact Report Tools. Available through paid subscription at http://www.isiwebofknowledge.com/. Accessed February 18, 2009 International Diabetes Federation. IDF Diabetes Atlas, 4th edn. Brussels, Belgium: International Diabetes Federation, 2009. International Society of Cellular Therapy Website. Risk of Bovine Spongiform Encephalopathy (BSE) in Collagenase Enzymes. Available at http://www.celltherapysociety.org/files/PDF/Resources/Risk_BSE_in_Collagenas e_Enzymes.pdf. Accessed September 17, 2008. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res. Feb 15 2003;31(4):e15. Irizarry RA, Hobbs B, Collin F, et al. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. Apr 2003;4(2):249-264. Jackson A, McWilliams C, Kaizer E, et al. Gene expression profiling of human pancreatic islets undergoing a simulated process of instant blood-mediated inflammatory reaction. Transplantation proceedings. Mar 2008;40(2):430-432. 187 Jindal RM, Ricordi C, Shriver CD. Autologous Pancreatic Islet Transplantation for Severe Trauma. New England Journal of Medicine. 2010;362(16):1550-1550. Johansson U, Olsson A, Gabrielsson S, Nilsson B, Korsgren O. Inflammatory mediators expressed in human islets of Langerhans: implications for islet transplantation. Biochem Biophys Res Commun. Aug 29 2003;308(3):474-479. Jung M, Schaefer A, Steiner I, et al. Robust microRNA stability in degraded RNA preparations from human tissue and cell samples. Clin Chem. Jun 2010;56(6):998-1006. Juvenile Diabetes Research Foundation International Website. JDRF Funded Research. Available at http://onlineapps.jdfcure.org/AbstractSearchEngine.cfm. Accessed November 17, 2008. Kaddis JS, Danobeitia JS, Niland JC, Stiller T, Fernandez LA. Multicenter Analysis of Novel and Established Variables Associated with Successful Human Islet Isolation Outcomes. American Journal of Transplantation. 2010;10(3):646- 656. Kaddis JS, Olack BJ, Sowinski J, Cravens J, Contreras JL, Niland JC. Human Pancreatic Islets and Diabetes Research. JAMA: The Journal of the American Medical Association. Apr 15 2009;301(15):1580-1587. Kaestner KH, Lee CS, Scearce LM, et al. Transcriptional program of the endocrine pancreas in mice and humans. Diabetes. Jul 2003;52(7):1604-1610. Kahvejian A, Quackenbush J, Thompson JF. What would you do if you could sequence everything? Nat Biotech. 2008;26(10):1125-1133. Kandaswamy R, Sutherland DE. Pancreas versus islet transplantation in diabetes mellitus: How to allocate deceased donor pancreata? Transplant Proc. Mar 2006;38(2):365-367. Kauppinen S, Vester B, Wengel J. Locked nucleic acid: high-affinity targeting of complementary RNA for RNomics. Handb Exp Pharmacol. 2006(173):405-422. 188 Kempf MC, Andres A, Morel P, et al. Logistics and transplant coordination activity in the GRAGIL Swiss-French multicenter network of islet transplantation. Transplantation. May 15 2005;79(9):1200-1205. Kennedy J, Katsuta H, Jung MH, et al. Protective unfolded protein response in human pancreatic beta cells transplanted into mice. PLoS One. 2010;5(6):e11211. Kent LA, Gill GV, Williams G. Mortality and outcome of patients with brittle diabetes and recurrent ketoacidosis. Lancet. Sep 17 1994;344(8925):778-781. Kessler L, Bucher P, Milliat-Guittard L, et al. Influence of islet transportation on pancreatic islet allotransplantation in type 1 diabetic patients within the Swiss- French GRAGIL network. Transplantation. Apr 27 2004;77(8):1301-1304. Kimura N, Tamura TA, Murakami M. Evaluation of the performance of two carbodiimide-based cyanine dyes for detecting changes in mRNA expression with DNA microarrays. Biotechniques. May 2005;38(5):797-806. Kin T, Mirbolooki M, Salehi P, et al. Islet isolation and transplantation outcomes of pancreas preserved with University of Wisconsin solution versus two-layer method using preoxygenated perfluorocarbon. Transplantation. Nov 27 2006;82(10):1286-1290. Kin T, Murdoch TB, Shapiro AM, Lakey JR. Estimation of pancreas weight from donor variables. Cell Transplant. 2006;15(2):181-185. Kin T, Rosichuk S, Shapiro AM, Lakey JR. Detection of microbial contamination during human islet isolation. Cell Transplant. 2007;16(1):9-13. Kin T, Senior P, O'Gorman D, Richer B, Salam A, Shapiro AM. Risk factors for islet loss during culture prior to transplantation. Transpl Int. Nov 2008;21(11):1029-1035. Kin T, Zhai X, Murdoch TB, Salam A, Shapiro AMJ, Lakey JRT. Enhancing the Success of Human Islet Isolation Through Optimization and Characterization of 189 Pancreas Dissociation Enzyme. American Journal of Transplantation. 2007;7(5):1233-1241. Kinasiewicz A, Juszczak M, Pachecka J, Fiedor P. Pancreatic islets isolation using different protocols with in situ flushing and intraductal collagenase injection. Physiol. Res. 2004;53(3):327-333. Kirchhof N, Shibata S, Wijkstrom M, et al. Reversal of diabetes in non- immunosuppressed rhesus macaques by intraportal porcine islet xenografts precedes acute cellular rejection. Xenotransplantation. 2004;11(5):396-407. Kissler HJ, Niland JC, Olack B, et al. Validation of methodologies for quantifying isolated human islets: an islet cell resources study. Clin Transplant. Aug 30 2009; 24(2):236-42. Klein D, Timoneri F, Ichii H, Ricordi C, Pastori RL. CD40 activation in human pancreatic islets and ductal cells. Diabetologia. Oct 2008;51(10):1853-1861. Knazek RA. The human pancreatic islet cell resource consortium. Diabetes Technol Ther. 2002;4(4):551-552. Korsgren O, Lundgren T, Felldin M, et al. Optimising islet engraftment is critical for successful clinical islet transplantation. Diabetologia. Feb 2008;51(2):227- 232. Koul PB. Diabetic Ketoacidosis: A Current Appraisal of Pathophysiology and Management. Clinical Pediatrics. March 1, 2009; 48(2):135-144. Kruger DF, Martin CL, Sadler CE. New Insights Into Glucose Regulation. The Diabetes Educator. March 1, 2006; 32(2):221-228. Kutlu B, Kayali AG, Jung S, et al. Meta-analysis of gene expression in human pancreatic islets after in vitro expansion. Physiol Genomics. Sep 9 2009;39(1):72-81. 190 Lakey JR, Mirbolooki M, Shapiro AM. Current status of clinical islet cell transplantation. Methods Mol Biol. 2006;333:47-104. Lakey JR, Rajotte RV, Warnock GL, Kneteman NM. Human pancreas preservation prior to islet isolation. Cold ischemic tolerance. Transplantation. Mar 15 1995;59(5):689-694. Lakey JR, Ricordi C, Hering BJ. Update and New Features in cGMP Facility, Equipment and Structure of the Clinical Islet Transplant Program. Paper presented at: Human Islet Isolation and Transplantation Techniques (HIITT) 6th Training Workshop; April 9-12, 2006; Snowbird, Utah. Lakey JR, Tsujimura T, Shapiro AM, Kuroda Y. Preservation of the human pancreas before islet isolation using a two-layer (UW solution-perfluorochemical) cold storage method. Transplantation. Dec 27 2002;74(12):1809-1811. Lakey JR, Warnock GL, Rajotte RV, et al. Variables in organ donors that affect the recovery of human islets of Langerhans. Transplantation. Apr 15 1996;61(7):1047-1053. Lakey JR, Warnock GL, Shapiro AM, et al. Intraductal collagenase delivery into the human pancreas using syringe loading or controlled perfusion. Cell Transplant. May-Jun 1999;8(3):285-292. Lakey JRT, Burridge PW, Shapiro AMJ. Technical aspects of islet preparation and transplantation. Transplant International. 2003;16(9):613-632. Langer RM, Mathe Z, Doros A, et al. Successful islet after kidney transplantations in a distance over 1000 kilometres: Preliminary results of the Budapest-Geneva collaboration. Transplant Proc. Dec 2004;36(10):3113-3115. Lau NC, Lim LP, Weinstein EG, Bartel DP. An abundant class of tiny RNAs with probable regulatory roles in Caenorhabditis elegans. Science. 2001;294(5543):858 - 862. Lee TC, Barshes NR, Brunicardi FC, et al. Procurement of the human pancreas for pancreatic islet transplantation. Transplantation. Aug 15 2004;78(3):481-483. 191 Lee YH, Kim SH. Production-distribution planning in supply chain considering capacity constraints. Computers and Industrial Engineering. 2002;43:169-190. Li J, Smyth P, Flavin R, et al. Comparison of miRNA expression patterns using total RNA extracted from matched samples of formalin-fixed paraffin-embedded (FFPE) cells and snap frozen cells. BMC Biotechnol. 2007;7:36. Liechti R, Csardi G, Bergmann S, et al. EuroDia: a beta-cell gene expression resource. Database (Oxford). 2010;2010:baq024. Linetsky E, Bottino R, Lehmann R, Alejandro R, Inverardi L, Ricordi C. Improved human islet isolation using a new enzyme blend, liberase. Diabetes. Jul 1997;46(7):1120-1123. Liu CG, Spizzo R, Calin GA, Croce CM. Expression profiling of microRNA using oligo DNA arrays. Methods. 2008;44(1):22-30. Liu X, Matsumoto S, Okitsu T, et al. Analysis of donor- and isolation-related variables from non-heart-beating donors (NHBDs) using the Kyoto islet isolation method. Cell Transplant. 2008;17(6):649-656. Liu YQ, Tornheim K, Leahy JL. Fatty acid-induced beta cell hypersensitivity to glucose. Increased phosphofructokinase activity and lowered glucose-6- phosphate content. J Clin Invest. May 1 1998;101(9):1870-1875. Liuwantara D, Elliot M, Smith MW, et al. Nuclear factor-kappaB regulates beta- cell death: a critical role for A20 in beta-cell protection. Diabetes. Sep 2006;55(9):2491-2501. Lonergan W, Whistler T, Vernon SD. Comparison of target labeling methods for use with Affymetrix GeneChips. BMC Biotechnol. 2007;7:24. Lundgren T, Korsgren O, Foss A, et al. The Nordic Network for Clinical Islet Transplantation - Islet After Kidney (Iak) - Results and Experiences From the First 44 Transplantations. Transplantation. 2004;78(2):178. 192 Lynch A, Hadfield J, Dunning M, Osborne M, Thorne N, Tavare S. The cost of reducing starting RNA quantity for Illumina BeadArrays: A bead-level dilution experiment. BMC Genomics. 2010;11(1):540. Ma C, Lyons-Weiler M, Liang W, et al. In vitro transcription amplification and labeling methods contribute to the variability of gene expression profiling with DNA microarrays. J Mol Diagn. May 2006;8(2):183-192. Macchiarini P, Jungebluth P, Go T, et al. Clinical transplantation of a tissue- engineered airway. Lancet. Dec 13 2008;372(9655):2023-2030. Maffei A, Liu Z, Witkowski P, et al. Identification of tissue-restricted transcripts in human islets. Endocrinology. Oct 2004;145(10):4513-4521. Magombedze G, Nduru P, Bhunu CP, Mushayabasa S. Mathematical modelling of immune regulation of type 1 diabetes. Biosystems. 2010;102(2-3):88-98. Mahler R, Franke FE, Hering BJ, et al. Evidence for a significant correlation of donor pancreas morphology and the yield of isolated purified human islets. J Mol Med. Jan 1999;77(1):87-89. Maitra A, Hansel DE, Argani P, et al. Global expression analysis of well- differentiated pancreatic endocrine neoplasms using oligonucleotide microarrays. Clin Cancer Res. Dec 1 2003;9(16 Pt 1):5988-5995. Manduchi E, Scearce LM, Brestelli JE, Grant GR, Kaestner KH, Stoeckert CJ, Jr. Comparison of different labeling methods for two-channel high-density microarray experiments. Physiol Genomics. Sep 3 2002;10(3):169-179. Mardis ER. The impact of next-generation sequencing technology on genetics. Trends Genet. Mar 2008;24(3):133-141. Mari A, Camastra S, Toschi E, et al. A model for glucose control of insulin secretion during 24 h of free living. Diabetes. February 1, 2001;50(suppl 1):S164. 193 Mari A, Ferrannini E. β-cell function assessment from modelling of oral tests: an effective approach. Diabetes, Obesity and Metabolism. 2008;10:77-87. Markmann JF, Deng SP, Desai NM, et al. The use of non-heart-beating donors for isolated pancreatic islet transplantation. Transplantation. May 2003;75(9):1423-1429. Markmann JF, Kaufman DB, Ricordi C, Schwab PM, Stock PG. Financial issues constraining the use of pancreata recovered for islet transplantation: a white paper. Am J Transplant. Aug 2008;8(8):1588-1592. Marselli L, Thorne J, Ahn YB, et al. Gene expression of purified beta-cell tissue obtained from human pancreas with laser capture microdissection. The Journal of clinical endocrinology and metabolism. Mar 2008;93(3):1046-1053. Marshall D, Sabek O, Fraga D, Kotb M, Gaber AO. Examination of the molecular signature associated with islet dysfunction. Transplantation proceedings. Mar 2005;37(2):1311-1312. Martello S, Toth P. Knapsack problems: algorithms and computer implementations. 1 ed. New York, NY USA: John Wiley & Sons, Inc; 1990. Mathe Z, Langer RM, Jaray J, et al. Human Pancreatic Islet Transplantation in International Collaboration With A Distant Islet Isolation Center: Preliminary Results From the Budapest-Geneva Experience. Transplantation. 2004;78(2):353. Matsumoto I, Sawada T, Nakano M, et al. Improvement in islet yield from obese donors for human islet transplants. Transplantation. Sep 27 2004;78(6):880-885. Matsumoto S, Okitsu T, Iwanaga Y, et al. Successful islet transplantation from nonheartbeating donor pancreata using modified Ricordi islet isolation method. Transplantation. Aug 27 2006;82(4):460-465. McPherson R, Pertsemlidis A, Kavaslar N, et al. A common allele on chromosome 9 associated with coronary heart disease. Science. Jun 2007;316(5830):1488-1491. 194 Medicare and Medicaid programs; conditions for coverage for organ procurement organizations (OPOs). Final rule. Fed Regist. May 31 2006;71(104):30981- 31054. Meivar-Levy I, Ferber S. New organs from our own tissues: liver-to-pancreas transdifferentiation. Trends in Endocrinology and Metabolism. 2003;14(10):460- 466. Mohammed JS, Wang Y, Harvat TA, Oberholzer J, Eddington DT. Microfluidic device for multimodal characterization of pancreatic islets. Lab Chip. Jan 7 2009;9(1):97-106. Montgomery D. Design and Analysis of Experiments, 5th ed. New York: Wiley; 2001. Mora JR, Getts RC. High-sensitivity detection methods for low-abundance RNA species: applications for functional genomics research. Expert Review of Molecular Diagnostics. 2007;7(6):775-785. Movahedi B, Gysemans C, Jacobs-Tulleneers-Thevissen D, Mathieu C, Pipeleers D. Pancreatic duct cells in human islet cell preparations are a source of angiogenic cytokines interleukin-8 and vascular endothelial growth factor. Diabetes. Aug 2008;57(8):2128-2136. Mraz M, Malinova K, Mayer J, Pospisilova S. MicroRNA isolation and stability in stored RNA samples. Biochem Biophys Res Commun. Dec 4 2009;390(1):1-4. Murdoch TB, McGhee-Wilson D, Shapiro AM, Lakey JR. Methods of human islet culture for transplantation. Cell Transplant. 2004;13(6):605-617. Nabokikh A, Ilhan A, Bilban M, et al. Reduced TGF-beta1 expression and its target genes in human insulinomas. Experimental and Clinical Endocrinology & Diabetes : Official Journal, German Society of Endocrinology [and] German Diabetes Association. Nov 2007;115(10):674-682. 195 Nagy ZB, Kelemen JZ, Feher LZ, Zvara A, Juhasz K, Puskas LG. Real-time polymerase chain reaction-based exponential sample amplification for microarray gene expression profiling. Anal Biochem. Feb 1 2005;337(1):76-83. Nanji SA, Shapiro AM. Advances in pancreatic islet transplantation in humans. Diabetes Obes Metab. Jan 2006;8(1):15-25. Nano R, Clissi B, Melzi R, et al. Islet isolation for allotransplantation: variables associated with successful islet yield and graft function. Diabetologia. May 2005;48(5):906-912. Narushima M, Kobayashi N, Okitsu T, et al. A human beta-cell line for transplantation therapy to control type 1 diabetes. Nature Biotechnology. Oct 2005;23(10):1274-1282. Nathan DM. Long-term complications of diabetes mellitus. N Engl J Med. Jun 10 1993;328(23):1676-1685. National Institutes of Health. The Research Portfolio Online Reporting Tool (RePORT). Available at http://report.nih.gov/index.aspx. Accessed October 30, 2008. National Islet Cell Resource Center Consortium Website. Available at http://icr.coh.org/. Accessed December 3, 2008. National Science Foundation Website. Award Search. Available at http://www.nsf.gov/awardsearch/. Accessed November 9, 2008. . Neidlinger NA, Odorico JS, Sollinger HW, Fernandez LA. Can 'extreme' pancreas donors expand the donor pool? Curr Opin Organ Transplant. Feb 2008;13(1):67- 71. Nelson PT, Wang WX, Wilfred BR, Tang G. Technical variables in high- throughput miRNA expression profiling: much work remains to be done. Biochim Biophys Acta. Nov 2008;1779(11):758-765. 196 Nemhauser GL, Wolsey LA. Integer and Combinatorial Optimization. 1 ed. New York, NY, USA Wiley-Interscience 1988 Niland J, Chen J, Paul W, Ottesen R. A Web-Based Database for Conducting Outcomes Research via the Internet: the National Comprehensive Cancer Network System. Scientific and Statistical Database Management. 1998:233- 237. Niland J. NCCN Outcomes Research Database: The First Two Years of Data Collection and Analysis. Oncology (Williston Park, N.Y. Nov 1999;13(11A). Niland J, Rouse L. Clinical Research Needs. In: Lehman HP AP, Roderer NK, Rothschild A, Mandell S, Ferrer JA, Miller RE, Ball MJ, ed. Aspects of Electronic Health Record Systems: Spring Science + Business Media Inc. ; 2006. Niland JC. NCCN Internet-based data system for the conduct of outcomes research. Oncology (Williston Park, N.Y. Nov 1998;12(11A):142-146. Niland JC. NCCN outcomes research database: data collection via the Internet. Oncology (Williston Park, N.Y. Nov 2000;14(11A):100-103. Niland JC. Internet-based data system for Outcomes Research. In: Silva J, Ball MJ, Chute C, et al., eds. Cancer Informatics: Essential Technologies for Clinical Trials 2002. Niland JC, Rouse L, Stahl DC. An informatics blueprint for healthcare quality information systems. J Am Med Inform Assoc. Jul-Aug 2006;13(4):402-417. Niland JC, Stahl DC. Increasing clinical trial awareness and accrual via the web. In: Silva J BM, Chute C, Langlotz E, Niland JC, Scerlis W, Douglas JV, ed. Cancer Informatics: Essential Technologies for Clinical Trials: Springer Verlag; 2002. Niland JC, Stiller T, Cravens J, Sowinski J, Kaddis J, Qian D. Effectiveness of a Web-Based Automated Cell Distribution System. Cell Transplant. May 4 2010; 19(9):1133-42. 197 Niland-Weiss J, Azen SP, Odom-Maryon T, Lui F, Hagerty C. A microcomputer- based distributed data management system for a large cooperative study of transfusion associated acquired immunodeficiency syndrome. Computers and Biomedical Research, an International Journal. Jun 1987;20(3):225-243. O'Gorman D, Kin T, Murdoch T, et al. The standardization of pancreatic donors for islet isolations. Transplantation. Sep 27 2005;80(6):801-806. Ohira M, Ishiyama K, Tanaka Y, et al. Adoptive immunotherapy with liver allograft-derived lymphocytes induces anti-HCV activity after liver transplantation in humans and humanized mice. J Clin Invest. Nov 2009;119(11):3226-3235. Ojo AO, Heinrichs D, Emond JC, et al. Organ donation and utilization in the USA. American Journal of Transplantation. 2004;4:27-37. Oliver-Krasinski JM, Stoffers DA. On the origin of the β cell. Genes & Development. August 1, 2008;22(15):1998-2021. Olsson A, Johansson U, Korsgren O, Frisk G. Inflammatory gene expression in Coxsackievirus B-4-infected human islets of Langerhans. Biochem Biophys Res Commun. May 6 2005;330(2):571-576. O'Neil JJ, Tchipashvili V, Parent RJ, et al. A simple and cost-effective method for the isolation of islets from nonhuman primates. Cell Transplant. 2003;12(8):883- 890. Organ Procurement and Transplantation Network (OPTN). View Data Reports: National Data. Available at http://www.optn.org/latestData/step2.asp?. Accessed February 20, 2009. Ostenson CG, Gaisano H, Sheu L, Tibell A, Bartfai T. Impaired gene and protein expression of exocytotic soluble N-ethylmaleimide attachment protein receptor complex proteins in pancreatic islets of type 2 diabetic patients. Diabetes. Feb 2006;55(2):435-440. 198 Ouellet M, Adams PD, Keasling JD, Mukhopadhyay A. A rapid and inexpensive labeling method for microarray gene expression analysis. BMC Biotechnol. 2009;9:97. Pagana M, Gauvreau K. Principles of Biostatistics, 2nd ed. Pacific Grove, CA: Duxbury; 2000. Papas KK, Colton CK, Qipo A, et al. Prediction of marginal mass required for successful islet transplantation. J Invest Surg. Feb 2010;23(1):28-34. Papas KK, Suszynski TM, Colton CK. Islet assessment for transplantation. Current Opinion in Organ Transplantation. 2009;14(6):674-682 Parnaud G, Bosco D, Berney T, et al. Proliferation of sorted human and rat beta cells. Diabetologia. Jan 2008;51(1):91-100. Patel OV, Suchyta SP, Sipkovsky SS, et al. Validation and application of a high fidelity mRNA linear amplification procedure for profiling gene expression. Vet Immunol Immunopathol. May 15 2005;105(3-4):331-342. Peltier HJ, Latham GJ. Normalization of microRNA expression levels in quantitative RT-PCR assays: identification of suitable reference RNA targets in normal and cancerous human solid tissues. RNA. May 2008;14(5):844-852. Pfaffl MW. A new mathematical model for relative quantification in real-time RT- PCR. Nucleic Acids Res. 2001;29(9):e45. Pileggi A, Ricordi C, Kenyon NS, et al. Twenty years of clinical islet transplantation at the Diabetes Research Institute--University of Miami. Clin Transpl. 2004:177-204. Pinto A, Tuttolomondo A, Di Raimondo D, Fernandez P, Licata G. Cerebrovascular risk factors and clinical classification of strokes. Semin Vasc Med. Aug 2004;4(3):287-303. 199 Polacek DC, Passerini AG, Shi C, et al. Fidelity and enhanced sensitivity of differential transcription profiles following linear amplification of nanogram amounts of endothelial mRNA. Physiol Genomics. 2003;13(2):147 - 156. Ponte GM, Pileggi A, Messinger S, et al. Toward maximizing the success rates of human islet isolation: influence of donor and isolation factors. Cell Transplant. 2007;16(6):595-607. Pop M, Salzberg SL. Bioinformatics challenges of new sequencing technology. Trends in Genetics. 2008;24(3):142-149. Pradervand S, Weber J, Lemoine F, et al. Concordance among digital gene expression, microarrays, and qPCR when measuring differential expression of microRNAs. Biotechniques. Mar 2010;48(3):219-222. Pradervand S, Weber J, Thomas J, et al. Impact of normalization on miRNA microarray expression profiling. RNA. Mar 2009;15(3):493-501. Puskas LG, Zvara A, Hackler L, Jr., Van Hummelen P. RNA amplification results in reproducible microarray data with slight ratio bias. Biotechniques. Jun 2002;32(6):1330-1334, 1336, 1338, 1340. Qi M, Barbaro B, Wang S, Wang Y, Hansen M, Oberholzer J. Human pancreatic islet isolation: Part I: digestion and collection of pancreatic tissue. J Vis Exp. 2009(27). Qi M, Barbaro B, Wang S, Wang Y, Hansen M, Oberholzer J. Human pancreatic islet isolation: Part II: purification and culture of human islets. J Vis Exp. 2009(27). Qi M, Strand BL, Morch Y, et al. Encapsulation of human islets in novel inhomogeneous alginate-ca2+/ba2+ microbeads: in vitro and in vivo function. Artificial Cells, Blood Substitutes, and Immobilization Biotechnology. 2008;36(5):403-420. 200 Qian D, Kaddis J, Niland JC. A matching algorithm for the distribution of human pancreatic islets. Computational Statistics & Data Analysis. 2007;51(12):5494- 5506. Qin H, Matsumoto S, Klintmalm GB, De Vol EB. A Meta-analysis for Comparison of the Two-Layer and University of Wisconsin Pancreata Preservation Methods in Islet Transplantation. Cell Transplant. Nov 19 2010. Quayum N, Kutchma A, Sarkar SA, et al. GeneSpeed Beta Cell: an online genomics data repository and analysis resource tailored for the islet cell biologist. Exp Diabetes Res. 2008;2008:312060. Rabkin JM, Leone JP, Sutherland DE, et al. Transcontinental shipping of pancreatic islets for autotransplantation after total pancreatectomy. Pancreas. Nov 1997;15(4):416-419. Rabkin JM, Olyaei AJ, Orloff SL, et al. Distant processing of pancreas islets for autotransplantation following total pancreatectomy. Am J Surg. May 1999;177(5):423-427. Ramiya V, Schatz D. Islet replacement vs. regeneration: hope or hype? Pediatr Diabetes. 2004;5 Suppl 2:45-56. Rao Y, Lee Y, Jarjoura D, et al. A comparison of normalization techniques for microRNA microarray data. Stat Appl Genet Mol Biol. 2008;7(1):Article22. Rebrin K, Steil GM, Getty L, Bergman RN. Free fatty acid as a link in the regulation of hepatic glucose output by peripheral insulin. Diabetes. Sep 1995;44(9):1038-1045. Rees DA, Alcolado JC. Animal models of diabetes mellitus. Diabet Med. Apr 2005;22(4):359-370. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. Jul 1997;20(7):1183-1197. 201 Reynoso J, Gruessner C, Sutherland D, Gruessner R. Short- and long-term outcome for living pancreas donors. Journal of Hepato-Biliary-Pancreatic Sciences. 2010;17(2):92-96. Riachy R, Vandewalle B, Moerman E, et al. 1,25-Dihydroxyvitamin D3 protects human pancreatic islets against cytokine-induced apoptosis via down-regulation of the Fas receptor. Apoptosis. Feb 2006;11(2):151-159. Richter A, Schwager C, Hentze S, Ansorge W, Hentze MW, Muckenthaler M. Comparison of fluorescent tag DNA labeling methods used for expression analysis by DNA microarrays. Biotechniques. Sep 2002;33(3):620-628, 630. Rickels MR, Naji A, Teff KL. Acute insulin responses to glucose and arginine as predictors of beta-cell secretory capacity in human islet transplantation. Transplantation. Nov 27 2007;84(10):1357-1360. Ricordi C. Quantitative and qualitative standards for islet isolation assessment in humans and large mammals. Pancreas. Mar 1991;6(2):242-244. Ricordi C, Alejandro R, Angelico MC, et al. Human islet allografts in patients with type 2 diabetes undergoing liver transplantation. Transplantation. Feb 15 1997;63(3):473-475. Ricordi C, Fraker C, Szust J, et al. Improved human islet isolation outcome from marginal donors following addition of oxygenated perfluorocarbon to the cold- storage solution. Transplantation. May 2003;75(9):1524-1527. Ricordi C, Gray DW, Hering BJ, et al. Islet isolation assessment in man and large animals. Acta diabetologica latina. Jul-Sep 1990;27(3):185-195. Ricordi C, Inverardi L, Kenyon NS, Goss J, Bertuzzi F, Alejandro R. Requirements for success in clinical islet transplantation. Transplantation. May 27 2005;79(10):1298-1300. Ricordi C, Lacy PE, Finke EH, Olack BJ, Scharp DW. Automated method for isolation of human pancreatic islets. Diabetes. Apr 1988;37(4):413-420. 202 Ricordi C, Lacy PE, Scharp DW. Automated islet isolation from human pancreas. Diabetes. Jan 1989;38 Suppl 1:140-142. Ricordi C, Strom TM. Clinical Islet Transplantation: Advances and Immunological Challenges. Nature Reviews Immunology. 2004;4:259-268. Ridgway DM, White SA, Kimber RM, Nicholson ML. Current practices of donor pancreas allocation in the UK: future implications for pancreas and islet transplantation*. Transplant International. 2005;18(7):828-834. Risch N. Linkage strategies for genetically complex traits. I. Multilocus models. Am J Hum Genet. Feb 1990;46(2):222-228. Robertson RP. Islet transplantation as a treatment for diabetes - a work in progress. N Engl J Med. Feb 12 2004;350(7):694-705. Robertson RP. Islet Transplantation a Decade Later and Strategies for Filling a Half-Full Glass. Diabetes. 06/01 2010;59(6):1285-1291. Rose NL, Palcic MM, Shapiro AMJ, Lakey JRT. Endogenous pancreatic enzyme activity levels show no significant effect on human islet isolation yield. Cell Transplant. 2004;13(2):153-160. Ryan EA, Paty BW, Senior PA, Lakey JR, Bigam D, Shapiro AM. Beta-score: an assessment of beta-cell function after islet transplantation. Diabetes Care. Feb 2005;28(2):343-347. Ryan EA, Shandro T, Green K, et al. Assessment of the severity of hypoglycemia and glycemic lability in type 1 diabetic subjects undergoing islet transplantation. Diabetes. Apr 2004;53(4):955-962. Saba R, Booth SA. Target labelling for the detection and profiling of microRNAs expressed in CNS tissue using microarrays. BMC Biotechnol. 2006;6(1):47. 203 Sabek OM, Cowan P, Fraga DW, Gaber AO. The effect of donor factors on human islet yield and their in vivo function. Prog Transplant. Dec 2006;16(4):350- 354. Sabek OM, Marshall DR, Penmetsa R, Scarborough O, Gaber AO. Examination of gene expression profile of functional human pancreatic islets after 2-week culture. Transplantation Proceedings. Dec 2006;38(10):3678-3679. Sah S, McCall MN, Eveleigh D, Wilson M, Irizarry RA. Performance evaluation of commercial miRNA expression array platforms. BMC Res Notes. Mar 18 2010;3(1):80. Salehi P, Hansen MA, Avila JG, et al. Human islet isolation outcomes from pancreata preserved with Histidine-Tryptophan Ketoglutarate versus University of Wisconsin solution. Transplantation. Oct 15 2006;82(7):983-985. Sarkar SA, Kutlu B, Velmurugan K, et al. Cytokine-mediated induction of anti- apoptotic genes that are linked to nuclear factor kappa-B (NF-kappaB) signalling in human islets and in a mouse beta cell line. Diabetologia. Jun 2009;52(6):1092- 1101. Sarkar SA, Wong R, Hackl SI, et al. Induction of indoleamine 2,3-dioxygenase by interferon-gamma in human islets. Diabetes. Jan 2007;56(1):72-79. Sarmiento AM, Nagi R. A review of integreated analysis of production distribution systems. IEE Transactions. 1999;31:1061-1074. Sato F, Tsuchiya S, Terasawa K, Tsujimoto G. Intra-platform repeatability and inter-platform comparability of microRNA microarray technology. PLoS One. 2009;4(5):e5540. Sawada A, Mizufune S, Kaji N, Tokeshi M, Baba Y. Evaluation of amplified cRNA targets for oligonucleotide microarrays. Anal Bioanal Chem. Apr 2007;387(8):2645-2654. 204 Saxena R, Voight BF, Lyssenko V, et al. Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels. Science. Jun 2007;316(5829):1331-1336. Scicchitano MS, Dalmas DA, Bertiaux MA, et al. Preliminary comparison of quantity, quality, and microarray performance of RNA extracted from formalin- fixed, paraffin-embedded, and unfixed frozen tissue samples. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society. Nov 2006;54(11):1229-1237. Scott LJ, Mohlke KL, Bonnycastle LL, et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science. Jun 2007;316(5829):1341-1345. Secrest AM, Becker DJ, Kelsey SF, LaPorte RE, Orchard TJ. All-Cause Mortality Trends in a Large Population-Based Cohort With Long-Standing Childhood- Onset Type 1 Diabetes. Diabetes Care. December 1, 2010;33(12):2573-2579. Selby JV, Karter AJ, Ackerson LM, Ferrara A, Liu J. Developing a prediction rule from automated clinical databases to identify high-risk patients in a large population with diabetes. Diabetes Care. Sep 2001;24(9):1547-1555. Shalev A, Pise-Masison CA, Radonovich M, et al. Oligonucleotide microarray analysis of intact human pancreatic islets: identification of glucose-responsive genes and a highly regulated TGFbeta signaling pathway. Endocrinology. Sep 2002;143(9):3695-3698. Shapiro AM, Lakey JR, Ryan EA, et al. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N Engl J Med. Jul 27 2000;343(4):230-238. Shapiro AM, Ricordi C, Hering BJ, et al. International trial of the Edmonton protocol for islet transplantation. N Engl J Med. Sep 28 2006;355(13):1318-1330. Shapiro ET, Tillil H, Rubenstein AH, Polonsky KS. Peripheral insulin parallels changes in insulin secretion more closely than C-peptide after bolus intravenous glucose administration. J Clin Endocrinol Metab. Nov 1988;67(5):1094-1099. 205 Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Research and Clinical Practice. 2010;87(1):4-14. Shi L, Reid LH, Jones WD, et al. The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. Nature Biotechnology. Sep 2006;24(9):1151-1161. Shippy R, Sendera TJ, Lockner R, et al. Performance evaluation of commercialshort-oligonucleotide microarrays and the impact of noise in makingcross-platform correlations. BMC Genomics. 2004;5(1):61. Silva J, Ball MJ, Chute C, et al. Cancer Informatics: Essential Technologies for Clinical Trials: Spring Verlag; 2002. Singh R, Maganti RJ, Jabba SV, et al. Microarray-based comparison of three amplification methods for nanogram amounts of total RNA. Am J Physiol Cell Physiol. May 2005;288(5):C1179-1189. Sivanandan S, Sinha A, Jain V, Lodha R. Management of Diabetic Ketoacidosis. Indian Journal of Pediatrics. 2010:1-9. Sladek R, Rocheleau G, Rung J, et al. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature. Feb 2007;445(7130):881- 885. Sorenson RL, Brelje TC. Adaptation of islets of Langerhans to pregnancy: beta- cell growth, enhanced insulin secretion and the role of lactogenic hormones. Hormone and metabolic research. Hormon- und Stoffwechselforschung. Jun 1997;29(6):301-307. Soria B, Tuduri E, Gonzalez A, et al. Pancreatic islet cells: a model for calcium- dependent peptide release. HFSP journal. Apr 2010;4(2):52-60. Stahl DC, Stahl CL, Niland JC. Evaluating the Impact of Clinical Trials On-Line on Clinical Trial Awareness and Accrual. Proceedings of the 33rd Annual Hawaii International Conference on System Science. Maui, Hawaii 2000. 206 Stewart ZA, Cameron AM, Singer AL, Dagher NN, Montgomery RA, Segev DL. Histidine-tryptophan-ketoglutarate (HTK) is associated with reduced graft survival in pancreas transplantation. Am J Transplant. Jan 2009;9(1):217-221. Sueblinvong V, Weiss DJ. Stem cells and cell therapy approaches in lung biology and diseases. Transl Res. Sep 2010;156(3):188-205. Szafranska AE, Davison TS, Shingara J, et al. Accurate Molecular Characterization of Formalin-Fixed, Paraffin-Embedded Tissues by microRNA Expression Profiling. J Mol Diagn. September 1, 2008;10(5):415-423. 't Hoen PAC, de Kort F, van Ommen GJB, den Dunnen JT. Fluorescent labelling of cRNA for microarray applications. Nucl. Acids Res. March 1, 2003; 31(5):e20. Tan PK, Downey TJ, Spitznagel EL, et al. Evaluation of gene expression measurements from commercial microarray platforms. NucleicAcids Res. 2003;31(19):5676 - 5684. Tang X, Muniappan L, Tang G, Ozcan S. Identification of glucose-regulated miRNAs from pancreatic {beta} cells reveals a role for miR-30d in insulin transcription. RNA. 2009;15(2):287 - 293. Tattersall R. Brittle diabetes. Clin Endocrinol Metab. Jul 1977;6(2):403-419. Taylor GW, Manz MC, Borgnakke WS. Diabetes, periodontal diseases, dental caries, and tooth loss: a review of the literature. Compend Contin Educ Dent. Mar 2004;25(3):179-184, 186-178, 190; quiz 192. Taylor MJ, Baicu S, Leman B, Greene E, Vazquez A, Brassil J. Twenty-four hour hypothermic machine perfusion preservation of porcine pancreas facilitates processing for islet isolation. Transplant Proc. Mar 2008;40(2):480-482. The Diabetes Control and Complications Trial Research Group. The Effect of Intensive Treatment of Diabetes on the Development and Progression of Long- Term Complications in Insulin-Dependent Diabetes Mellitus. New England Journal of Medicine. 1993;329(14):977-986. 207 Thomas DC. Are We Ready for Genome-wide Association Studies? Cancer Epidemiol Biomarkers Prev. April 1, 2006;15(4):595-598. Tibell A, Bolinder J, Hagström-Toft E, et al. Experience with human islet transplantation in Sweden. Transplantation Proceedings. 2001;33(4):2535-2536. Tissot C. Application Note: Analysis of miRNA content in total RNA preparations using the Agilent 2100 bioanalyzer. Publication Number 5989-787OEN. February 2008:1-8. Todd JA, Walker NM, Cooper JD, et al. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet. 2007;39(7):857-864. Toffolo G, Breda E, Cavaghan MK, Ehrmann DA, Polonsky KS, Cobelli C. Quantitative indexes of Î²-cell function during graded up&down glucose infusion from C-peptide minimal models. American Journal of Physiology - Endocrinology And Metabolism. January 1, 2001;280(1):E2-E10. Toso C, Shapiro AM, Bowker S, et al. Quality of life after islet transplant: impact of the number of islet infusions and metabolic outcome. Transplantation. Sep 15 2007;84(5):664-666. Trout KK, Homko C, Tkacs NC. Methods of Measuring Insulin Sensitivity. Biol Res Nurs. April 1, 2007;8(4):305-318. Trucco M. Regeneration of the pancreatic beta cell. J Clin Invest. Jan 2005;115(1):5-12. Tsujimura T, Kuroda Y, Avila JG, et al. Influence of pancreas preservation on human islet isolation outcomes: impact of the two-layer method. Transplantation. Jul 15 2004;78(1):96-100. Tsujimura T, Kuroda Y, Churchill TA, et al. Short-term storage of the ischemically damaged human pancreas by the two-layer method prior to islet isolation. Cell Transplant. 2004;13(1):67-73. 208 Tsujimura T, Kuroda Y, Kin T, et al. Human islet transplantation from pancreases with prolonged cold ischemia using additional preservation by the two-layer (UW solution/perfluorochemical) cold-storage method. Transplantation. Dec 27 2002;74(12):1687-1691. United Network for Organ Sharing (UNOS) Website. Policy 3.8 Organ Distribution: Pancreas Allocation. Available at http://www.unos.org/PoliciesandBylaws2/policies/pdfs/policy_10.pdf. Accessed February 28, 2009. Valdes-Gonzalez RA, Dorantes LM, Garibay GN, et al. Xenotransplantation of porcine neonatal islets of Langerhans and Sertoli cells: a 4-year study. European journal of endocrinology / European Federation of Endocrine Societies. Sep 2005;153(3):419-427. Van Belle TL, Coppieters KT, Von Herrath MG. Type 1 Diabetes: Etiology, Immunology, and Therapeutic Strategies. Physiological Reviews. January 1, 2011;91(1):79-118. Vandesompele J, De Preter K, Pattyn F, et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. Jun 18 2002;3(7):RESEARCH0034. Viale A, Li J, Tiesman J, et al. Big results from small samples: evaluation of amplification protocols for gene expression profiling. J Biomol Tech. Jul 2007;18(3):150-161. Vidal CJ, Goetschalckx M. Strategic production-distribution models: A critical review with emphasis on global supply chain models. European Journal of Operational Research. 1997;98(1):1-18. Wang H, Ach RA, Curry B. Direct and sensitive miRNA profiling from low-input total RNA. RNA. January 2007;13(1):151-159. Wang K, Li M, Bucan M. PathwayBased Approaches for Analysis of Genomewide Association Studies. The American Journal of Human Genetics. 2007;81(6):1278-1283. 209 Wang WX, Wilfred BR, Baldwin DA, et al. Focus on RNA isolation: obtaining RNA for microRNA (miRNA) expression profiling analyses of neural tissue. Biochim Biophys Acta. Nov 2008;1779(11):749-757. Watanabe RM, Steil GM, Bergman RN. Critical evaluation of the combined model approach for estimation of prehepatic insulin secretion. Am J Physiol. Jan 1998;274(1 Pt 1):E172-183. Watanabe RM, Volund A, Roy S, Bergman RN. Prehepatic beta-cell secretion during the intravenous glucose tolerance test in humans: application of a combined model of insulin and C-peptide kinetics. J Clin Endocrinol Metab. Oct 1989;69(4):790-797. Weber DJ. FDA regulation of allogeneic islets as a biological product. Cell biochemistry and biophysics. 2004;40(3 Suppl):19-22. Weber DJ, McFarland RD, Irony I. Selected Food and Drug Administration review issues for regulation of allogeneic islets of langerhans as somatic cell therapy. Transplantation. Dec 27 2002;74(12):1816-1820. Weir GC, Bonner-Weir S. Five stages of evolving beta-cell dysfunction during progression to diabetes. Diabetes. Dec 2004;53 Suppl 3:S16-21. Westermark GT, Westermark P, Berne C, Korsgren O, the Nordic Network for Clinical Islet T. Widespread Amyloid Deposition in Transplanted Human Pancreatic Islets. N Engl J Med. August 28, 2008;359(9):977-979. Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature. Jun 7 2007;447(7145):661-678. White SA, Shaw JA, Sutherland DE. Pancreas transplantation. Lancet. May 23 2009;373(9677):1808-1817. Wild S, Roglic G, Green A, Sicree R, King H. Global Prevalence of Diabetes. Diabetes Care. May 1, 2004;27(5):1047-1053. 210 Willenbrock H, Salomon J, Sokilde R, et al. Quantitative miRNA expression analysis: comparing microarrays with next-generation sequencing. RNA. Nov 2009;15(11):2028-2034. William K, Jian-Jun Z, Lili H, Jin QC. Strategies for profiling MicroRNA expression. Journal of Cellular Physiology. 2009;218(1):22-25. Williams G, Gill GV, Pickup JC. 'Brittle' diabetes. Bmj. Sep 21 1991;303(6804):714. Woodruff DL, Reiners T. Experiments with, and on, algorithms for maximum likelihood clustering. Computational Statistics and Data Analysis. 2004;47(2 SPEC. ISS.):237-253. Wu Z, Irizarry RA, Gentleman R, Martinez-Murillo F, Spencer F. A Model-Based Background Adjustment for Oligonucleotide Expression Arrays. Journal of the American Statistical Association. 2004;99(468):909. Xi Y, Nakajima G, Gavin E, et al. Systematic analysis of microRNA expression of RNA extracted from fresh frozen and formalin-fixed paraffin-embedded samples. RNA. Oct 2007;13(10):1668-1674. Xie W, Ted Brown W, Denman RB. Translational regulation by non-protein- coding RNAs: different targets, common themes. Biochem Biophys Res Commun. 2008;373(4):462 - 466. Yamamoto T, Horiguchi A, Ito M, et al. Quality control for clinical islet transplantation: organ procurement and preservation, the islet processing facility, isolation, and potency tests. J Hepatobiliary Pancreat Surg. 2009;16(2):131-136. Yamamoto T, Ricordi C, Messinger S, et al. Deterioration and variability of highly purified collagenase blends used in clinical islet isolation. Transplantation. Oct 27 2007;84(8):997-1002. Yang H, Acker J, Chen A, McGann L. In situ assessment of cell viability. Cell Transplant. Sep-Oct 1998;7(5):443-451. 211 Yang Z, Chen M, Deng S, et al. Assessment of human pancreatic islets after long distance transportation. Transplantation Proceedings. 2004;36(5):1532-1533. Yauk CL, Rowan-Carroll A, Stead JD, Williams A. Cross-platform analysis of global microRNA expression technologies. BMC Genomics. May 26 2010;11(1):330. Ylipaasto P, Kutlu B, Rasilainen S, et al. Global profiling of coxsackievirus- and cytokine-induced gene expression in human pancreatic islets. Diabetologia. Aug 2005;48(8):1510-1522. Yu J, Othman MI, Farjo R, et al. Evaluation and optimization of procedures for target labeling and hybridization of cDNA microarrays. Mol Vis. Apr 26 2002;8:130-137. Zakharkin SO, Kim K, Mehta T, et al. Sources of variation in Affymetrix microarray experiments. BMC Bioinformatics. 2005;6:214. Zeng Y, Torre MA, Karrison T, Thistlethwaite JR. The correlation between donor characteristics and the success of human islet isolation. Transplantation. Mar 27 1994;57(6):954-958. Zhang X, Chen J, Radcliffe T, Lebrun DP, Tron VA, Feilotter H. An array-based analysis of microRNA expression comparing matched frozen and formalin-fixed paraffin-embedded human tissue samples. J Mol Diagn. Nov 2008;10(6):513- 519. Zhou Q, Brown J, Kanarek A, Rajagopal J, Melton DA. In vivo reprogramming of adult pancreatic exocrine cells to beta-cells. Nature. Aug 27 2008; 455(7213):627-32. Zhu B, Xu F, Baba Y. An evaluation of linear RNA amplification in cDNA microarray gene expression analysis. Mol Genet Metab. Jan 2006;87(1):71-79. Zhu Q, Hong A, Sheng N, et al. muParafloTM Biochip for Nucleic Acid and Protein Analysis. Microarrays. 2007;2:287 - 312. 212 Zoon KC. Dear colleague letter to transplant centers: allogeneic pancreatic islets for transplantation. Washington DC, US Department of Health and Human Services. 2000. 213

Abstract (if available)

Abstract Destruction or dysfunction in the human pancreatic islet affects at least 23.6 million affected individuals in the U.S. alone. The inability to regulate insulin production and maintain glucose homeostasis leads to a variety of severe diabetic complications at an estimated 2007 US health care cost of $174 billion dollars. Although medical management, lifestyle changes, and pharmacological agents are successful treatment tools for some, they are less effective, and have failed, in those with unstable diabetes, indicating that an urgent need for alternative therapies exist. Pancreatic islet transplantation is a form of cellular replacement therapy that has been shown to restore glycometabolic control and render some patients insulin independent. Our long term goal is to therefore improve human islet survival and transplantation success rates by understanding the factors affecting cell function in-vitro and in-vivo both in the native pancreas and transplant environments.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Development of ECM for the preservation of adult mouse pancreatic islet function in vitro

PDF

Non-invasive live-cell imaging for monitoring and evaluating pancreatic islet and beta cell metabolism

PDF

A novel construct to study the pulsatility of insulin secretion in single cells, islets and whole pancreas

PDF

Quantity versus quality: how adipose tissue accumulation and immune cell profile associate with risk for type 2 diabetes in minority children and adults

PDF

The role of endoplasmic reticulum chaperones in adipogenesis, liver cancer and mammary gland development

Asset Metadata

Creator Kaddis, John Samuel (author)

Core Title Transplantation and genetics of human pancreatic islets in diabetes: Approaches in translational medicine and statistics

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Systems Biology

Publication Date 05/04/2013

Defense Date 03/22/2011

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag bioinformatics,Biostatistics,Diabetes,genetics,genomics,human pancreatic islets,microarray,miRNA,OAI-PMH Harvest,systems biology,Transplantation

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Warburton, David (committee chair), Allayee, Hooman (committee member), Conti, David V. (committee member), Niland, Joyce (committee member)

Creator Email kaddis@usc.edu,mitsrayim@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3887

Unique identifier UC1179296

Identifier etd-Kaddis-4563 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-476566 (legacy record id),usctheses-m3887 (legacy record id)

Legacy Identifier etd-Kaddis-4563.pdf

Dmrecord 476566

Document Type Dissertation

Rights Kaddis, John Samuel

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu