Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Statistical analysis of a Phase II study of AMG 386 versus AMG 386 combined with anti-VEGF therapy in patients with advanced renal cell carcinoma
(USC Thesis Other)
Statistical analysis of a Phase II study of AMG 386 versus AMG 386 combined with anti-VEGF therapy in patients with advanced renal cell carcinoma
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
1 Statistical Analysis of a Phase II Study of AMG 386 versus AMG 386 combined with anti-VEGF Therapy in Patients with Advanced Renal Cell Carcinoma Harini Raghunathan Thesis Presented to the Faculty of the Graduate School University of Southern California In Partial Fulfillment of the Requirements for the Degree Master of Science in Applied Biostatistics and Epidemiology May 2019 2 Table of Contents Abstract ............................................................................................................ 3 1. Introduction ......................................................................................................................................................... 4 1.1 Completed Trial………………………………………………………………………………………………………..5 1.1.1 Endpoints in the Completed Trial…………………………………………………………………………..6 1.2 Purpose of this Project………………………………………………………………………………………….…...7 2. Methods ......................................................................................................................... 7 2.1 Baseline Characteristics……………………………………………………………………………………………7 2.2 Toxicity…………………………………………………………………………………………………………………….8 2.2.1 Predictors of Toxicities: Logistic Regression……………….……………………………………..…9 2.3 Tumor Response to Treatment………………………………………………………………………………..10 2.4 Progression-Free Survival (PFS)………………………….…………………………………………………11 3. Results .......................................................................................................................... 11 3.1 Baseline Characteristics……………………………………………….………………………………………...11 3.2 Toxicity……………………………………………………………………..……..……………………………………15 3.2.1 Predictors of Toxicities: Logistic Regression……………………………………………………...18 3.3 Response………………………………………………………………………………….…………………………….20 3.4 PFS…………………..…………………………………………………………………………………………………...22 4. Discussion ........................................................................................................................................................ …23 5. Conclusion ................................................................................................................... 24 6. References……………………………………………………………….………………………………....………….25 7. Appendix…………..………………………………………….….…………………...….…………………...………26 3 Abstract The VEGF pathway is associated with angiogenesis and is an important target for molecular therapies for treating Renal Cell Carcinoma. Existing anti-VEGF therapies, including Bevacizumab, Pazopanib, Sorafenib, and Sunitinib, are available to treat this cancer. However, the presence of alternative pathways, such as the angiopoietin-Tie pathway, is hypothesized to lead to anti-VEGF therapy resistance. AMG 386 was developed to treat renal cell carcinoma through the angiopoietin-Tie pathway. In a Phase II randomized clinical trial, the overall response rate of AMG 386 alone and in combination with an anti-VEGF therapy was assessed. We analyzed the data from this trial to evaluate the efficacy of AMG 386. For this Phase II randomized trial, AMG 386 alone and in combination with anti-VEGF therapy were evaluated separately. The primary endpoint was overall response. For this thesis, we compared the two arms, although this was not done for the original trial. A total of 35 evaluable patients were available in this study, where we evaluated baseline characteristics, toxicity, response, and progression-free survival. Toxicity was summarized by grade and toxicity system; also, using logistic regression, odds ratios were calculated to analyze the association between hematologic and non-hematologic toxicities and baseline explanatory variables. Response was summarized in terms of RECIST Criteria and a waterfall plot. Fisher’s Exact test and Wilcoxon Rank Sum were used to test the difference between baseline explanatory variables in the two treatments. A Kaplan Meier plot and log-rank test were used to analyze progression-free survival. The primary conclusion is that neither AMG 386 alone or in combination with an anti-VEGF therapy was effective in treating advanced renal cell carcinoma. Only two patients - both in the combination arm, experienced an objective response to treatment, and 37% of the patients in both arms had stable disease as their best response. 4 Figure 1: Angiogenesis pathway and role of 4 established anti-VEGF therapies. Reference sited 3* 1. Introduction Renal Cell Carcinoma (RCC) represents 85% of all renal cancer cases and is the most common type of kidney cancer among adults 1 . For most patients diagnosed with RCC, the cancer has already metastasized, making this cancer hard to treat, as it is not localized in one area. There are therapeutic options for this advanced disease – primarily medicinal therapies. Many of these medicinal therapies target the vascular endothelial growth factor (VEGF) molecular pathway. However, these treatments are not effective in completely eliminating RCC as can be seen by the recurrence of the disease or vast proliferation of the cancer throughout the body 2 . The VEGF molecular pathway is associated with angiogenesis or the creation of new blood vessels involving endothelial cells. In this pathway, the Von Hippel Lindau (VHL) is a tumor suppressor protein that degrades the Hypoxia Inducible Factor (HIF) protein. Degradation of HIF prevents other factors that are involved in angiogenesis, including VEGF, to be transcribed. In RCC, the VHL is often inactive, causing an increase in HIF and as a result, an increase in factors involved in angiogenesis, such as VEGF (Figure 1). This ultimately causes an increase in blood vessel formation in the tumor creating a favorable environment for tumor growth and metastasis. Thus, the VEGF pathway is an important molecular target for RCC therapies. Although there have been several advancements in anti-VEGF therapies over the years, the presence of other pro- angiogenic molecular pathways have led to a resistance to anti-VEGF therapies. One alternative pathway is the angiopoietin- Tie pathway. Tie 2 is a tyrosine kinase receptor on endothelial cells and Ang 1 and Ang 2 are ligands that bind to Tie 2. The exact molecular signaling pathway of Ang-Tie2 is unknown. However, it is understood that Ang 1 promotes vascular stabilization, while Ang 2 causes endothelial cell proliferation by allowing the cells to be more reactive to VEGF signaling. Therapies are being developed to treat this alternate angiogenic pathway along with the VEGF pathway. 5 There are a few small molecule therapies and an antibody drug therapy that are used to treat RCC. All these drugs treat RCC by blocking further molecular signaling of the VEGF pathway. Pazopanib and Sunitinib are small molecules that inhibit tyrosine kinase receptors on the VEGF receptors (VEFRs). Sorafenib is a small molecule tyrosine kinase inhibitor that uses autophagy to suppress tumor growth. Lastly, Bevacizumab, an antibody drug, binds to VEGF and prevents interaction with the VEGFR (Figure 1). It is known that in RCC, during anti-VEGF treatment, Ang 2 levels are elevated, thus leading to anti-VEGF resistance. This resistance leads to alternative pro-angiogenic pathways and eventually cancer progression. To target the angiopoietin-Tie pathway, AMG 386 has been developed. AMG 386 binds and separates Ang1 and Ang2, thus preventing these ligands to interact with Tie2. This causes the endothelial cells not to proliferate and thus inhibits both the angiopoietin-Tie pathway and cancer progression 2 . 1.1 Completed Trial A Phase II parallel arm clinical trial had been conducted to evaluate the effect of AMG 386 alone and AMG 386 in combination with a previously received anti-VEGF separately for the treatment of RCC (clinicaltrial.gov Trial ID: NCT01664182). This was a ‘signal seeking’ trial, and a response rate of 15% or greater was considered potentially interesting. The primary endpoint in this trial was objective response (defined below). In this Phase II clinical trial, Dr. Semrad and colleagues wanted to assess the overall response rate of AMG 386 alone and in combination with the patient’s prior anti-VEGF therapy. Additionally, they evaluated the progression free survival (PFS) and described the toxicity experienced by these patients 2 . They hypothesized that AMG 386 alone and in combination with the prior anti-VEGF therapy would have anti-tumor activity in patients with metastatic RCC, by inhibiting the angiopoietin-Tie pathway. For this trial, each arm was evaluated separately, and the primary efficacy endpoint was objective response. Using a one-sample binomial test, 32 patients would be required for each arm. However, the study investigators used the Simon Optimum test, which required 39 patients with an interim analysis after 17 patients had been enrolled. The study parameters were: 90% power when true response rate is 15%, and alpha of 0.1 when the true response rate is 3%. For this trial, the investigators set the null hypothesis to be H 0 : p≤3%, where p is equal to the true response rate. Technically, for a one-sided test, the alternative hypothesis test would be H a : p>3%. However, for this trial, a true response rate of 15% was clinically interesting and the alternative hypothesis was set to be H a : p≥15%. With this design, when 1 to 2 patients out of 39 patients experienced an objective response, the null hypothesis was not rejected. If 3 or more patients out of 39 patients experienced an objective response, then the null hypothesis was rejected. All patients in this trial had metastatic renal cell carcinoma that had been previously treated with one of the four anti-VEGF therapies. The patients had been randomly assigned to two groups (A and B). Group A was given a monotherapy of AMG 386 while group B was given AMG 386 in combination with the patient’s prior anti- 6 VEGF treatment. All patients in both groups were given AMG 386 intravenously (IV) at a dose of 15 mg/kg at each sitting; each cycle was 6 weeks. Patients in group A were administered AMG 386 alone. Those in group B additionally also received the anti- VEGF drug they had previously received. Those who previously took Bevacizumab were given a dosage of 10 mg/kg intravenously (IV) over a minimum period of 90 minutes. If this time length was well tolerated, then the length of each administration was lessened to a minimum of 60 minutes. Patients took Pazopanib orally at 800 mg once a day. Sorafenib and Sunitinib were also taken orally, but those who took Sorafenib had a dosage of 400 mg daily, while patients took Sunitinib at a dosage of 50 mg daily for four weeks and then off for 2 weeks. For both Sorafenib and Sunitinib, if the starting dosage resulted in toxicity, then the dosage was reduced. All these therapies had 6-week cycles, and patients continued on treatment until disease progression, unacceptable toxicity or patient decision (Table 1). Table 1: Summary of Treatment Arms AMG 386 Bevacizumab Pazopanib Sorafenib Sunitinib Arm A 15 mg/kg, through IV, 6 weeks - - - - Arm B* 15 mg/kg, through IV, 6 weeks 10 mg/kg, through IV, 6 weeks 800 mg, orally, 6 weeks 400 mg, orally, 6 weeks 50 mg, orally, 6 weeks *Arm B patients received AMG 386 plus their prior anti-VEGF treatment (one other anti-VEGF treatment) 1.1.1 Endpoints in the Completed Trial Response: Evaluating the response (i.e., tumor shrinkage) was done by first identifying and measuring the longest diameter of all tumor lesions for which measurements were possible. This was done at baseline prior to the start of treatment; a CT scan or an MRI was performed on each patient and all tumor lesions were identified. For each patient, the baseline tumor burden was defined as the sum of the longest diameters of all lesions, up to 5 lesions on the day of measurement – selecting 5 lesions that were large and easily measured; the lesions included in the sum were labeled "target" lesions. The presence of all other (non-target) lesions was noted. Once the 5 target lesions were recorded and all other present lesions were noted at baseline, tumor burden was reassessed every 12 weeks (after every 2 cycles of therapy) using the same imaging (CT or MRI); the longest diameters of the target lesions were re-measured and the tumor burden at the follow-up time point was recalculated as the sum of these longest diameters. From this tumor burden, the overall response was calculated based on the criteria specified by RECIST version 1.1, a set of guidelines that measures patients’ tumor responses. Complete Response (CR) is when the tumor is completely gone, while partial 7 response (PR) is when the tumor burden measurement has decreased by 30% and there are no new lesions. Progressive disease (PD) is when the tumor burden measurement has increased more than 20% or there are new lesions. Stable disease (SD) is when there is no CR, PR or PD 4 . The overall best response was defined as a CR or PR. There was no requirement to confirm a CR or PR in this study. Progression-free survival (PFS): The next endpoint was PFS, which was calculated as the time when the patient started treatment until first documentation of progression of the disease or death; for patients who were still alive and had not progressed at the time of analysis, the PFS was censored at the time that the patient was last known to be alive and free of progression. This was measured in terms of months from start of treatment for both arms of the study. Toxicity: Toxicity, as measured by adverse events, measures how well the therapies have been tolerated. We will use toxicity throughout this thesis. The adverse events were evaluated on a scale of 0 to 5 as classified by the Common Terminology Criteria for Adverse Events (CTCAE) version 4.0 5 . The CTCAE v.4 has a list of over 200 toxicities grouped by system. For each toxicity, Grade 0 means no adverse events of the drug, 1 means mild but not requiring treatment, 2 is still mild and requiring treatment, 3 means adverse events are severe, 4 means that the events are life-threatening, and lastly grade 5 means fatal. 1.2 Purpose of this Project The purpose of this project was to analyze the data from the Phase II renal cancer clinical trial comparing AMG 386 alone to AMG 386 with a prior anti-VEGF therapy to see which arm was better at producing responses (CR and PR), as well as prolonging the progression free survival of RCC in these patients. The data analyzed included baseline characteristics, response, PFS, and toxicity. 2. Methods As stated above, the purpose of this thesis was to analyze the data from the clinical trial, through calculations on baseline characteristics, toxicity, response, PFS and logistic regression; although not done in the original trial, we also compared both arms. All the data management and analyses were done in SAS. 2.1 Baseline Characteristics We created a table that summarized patient baseline characteristics, which included demographics, performance status, and treatment. We used descriptive statistics to summarize these baseline patient characteristics. 8 ECOG Performance Status has five grades. Grade 0 means fully active, grade 1 means restricted physical activity, grade 2 means able to take care of basic needs but not able to work, grade 3 means restricted to the bed more the 50% of the day, grade 4 means completely disabled, and grade 5 means dead 6 . In this study, all patients had ECOG status between 0 and 1. Percentages were calculated for each arm and for both arms combined, for gender, ECOG Performance Status, race, and prior anti-VEGF; while for weight, and age, the median, range, 25 th and 75 th percentiles were calculated for each arm and for both arms combined (Table 2). We calculated the p-values for race, gender, ECOG performance status and prior anti-VEGF. We compared the two arms, in terms of these baseline characteristics, using Fisher’s Exact Test. This test was used since the total number of patients is small. For the continuous variables, age and weight, we used the Wilcoxon Rank Sum test to calculate p-values to compare the two arms in terms of these two continuous variables. 2.2 Toxicity The Common Terminology Criteria for Adverse Events (CTCAE) v4.0 was used to score all the toxicities experienced by patients in this trial 5 . There are over 200 types of toxicities that are graded from 0 (no toxicity) to 5 (resulting in death). These toxicities were grouped into 21 toxicity systems (Table 2). Table 2: Toxicity Systems Toxicity System (CTCAE v 4.0) Blood and lymphatic system disorders Cardiac disorders Congenital, familial and genetic disorders Ear and labyrinth disorders Endocrine disorders Eye disorders Gastrointestinal disorders General disorders and administration site conditions Infections and infestations Injury, poisoning and procedural complications Investigations Metabolism and nutrition disorders Musculoskeletal and connective tissue disorders Neoplasms benign, malignant and unspecified (incl cysts and polyps) Nervous system disorders Psychiatric disorders Renal and urinary disorders Reproductive system and breast disorders Respiratory, thoracic and mediastinal disorders Skin and subcutaneous tissue disorders Vascular disorders 9 To summarize the toxicities, we first calculated the maximum grade that each patient experienced for all the toxicities over the entire course of her/his treatment. We grouped grades 1 and 2 together but counted grades 3 and 4 separately. There were two patients who had grade 5 toxicities. For both patients, there was a difference in the data presented in the database and summary assessment regarding the classification (attributed to treatment or not) of the grade 5 toxicities. Due to the contradictory data, no patients who had a toxicity of grade 5 were included in this analysis. From this we created two tables: one that summarized the number of patients with each type of toxicity – with Arms A and B done separately, and a 2 nd table where only the toxicities (rows) with at least one patient having experienced a grade 3 or 4 or at least 2 patients have experienced that toxicity at any grade. Therefore, this second table was a reduction of the 1 st table, since toxicities in which only 1 patient who experienced grades 1 or 2 toxicity were omitted. To create a third table, we calculated the maximum grade that each patient experienced for each system (i.e., collapsing over all the toxicities within a system) over the entire course of her/his treatment – again, we grouped grades 1 and 2 together, but counted grades 3 and 4 separately (Tables 3 and 4). From this we created tables that summarized the number of patients with maximum toxicity of 1-2, 3, or 4 in each system for Arms A and B separately. 2.2.1 Predictors of Toxicities: Logistic Regression Logistic regression was used to calculate odds ratios of hematologic and non- hematologic toxicities for the different explanatory variables (age, weight, gender, race/ethnicity, ECOG Performance Status, Arm, and anti-VEGF). These odds ratios were calculated and these associations were compared to a comparison group for each explanatory variable. For each patient, we calculated the maximum grade that each patient experienced over the entire course of her/his treatment for hematologic or non-hematologic toxicities separately. We then coded each patient as ever having experienced a maximum toxicity of 3 or 4 for hematologic and for non-hematologic toxicities (code = 1)– or not (code =0). We used a logistic regression model to assess whether there were baseline characteristics, which might be associated with developing grade 3+ hematologic, or non-hematologic toxicities. Since the number of patients eligible to be in this study was small, each explanatory variable was reduced to two or three categories. For the categorical variables: race/ethnicity was split into white versus non-white, ECOG performance status was split into value 0 versus value 1, and prior anti-VEGF was split into Bevacizumab versus the small molecule therapies. Since weight and age were continuous variables, we divided them into three groups and each group contained 11 to 12 patients. We analyzed these variables as categorical variables. Logistic Regression was conducted on these two types of toxicities for each explanatory variable separately. A stepwise function was performed with all the explanatory variables in order to create a prediction model for hematologic and non-hematologic toxicities. Since logistic regression was negative, we did not evaluate the validity of the model. 10 2.3 Tumor Response to Treatment Measurement of response i.e., tumor burden, as described by RECIST version 1.1, began by measuring the diameter of all the target lesions. For target lesions that were present in the lymph nodes, the length of the shortest axis was used; for all other target lesions, the length of the longest diameter was measured. The sum of the target lesions measured on the same day for each patient is the called the "LD". The minimum (earliest) exam date was found for each patient – which represented the date prior to the start of treatment. In reviewing the data set, there were several data entry errors. There was an error in the exam date for subject 10; all dates were supposed to be between December 23, 2013, and April 4, 2014. There were a couple of dates that were incorrect, and these have been corrected in the dataset used for this analysis. The LD that is corresponding to the minimum date for each patient is the baseline LD. For each date after the baseline, the percent change in the LD compared to the baseline LD was calculated by dividing the post-baseline LD by the baseline LD, subtracting 1.0 from the ratio and multiplying this difference by 100; this percent change in LD, is a measure of how much the tumor burden has changed since the baseline LD measurement, and reflects a response (or lack of response) to treatment. Patients who experienced a percent decrease of 30% or more were considered to have an objective response, i.e., CR or PR (as long as there were no new lesions). Next, we found the nadir LD value, which is the smallest LD measurement per patient over all of his/her dates that does not include the baseline LD. Using the nadir value progression was determined. Each post-baseline LD was compared to the nadir by dividing the post-baseline LD by nadir, subtracting 1.0 and multiplying by 100. Using all post nadir changes, the percent change measures how much the disease has increased from the smallest LD value. An increase of 20% or more was called disease progression. The appearance of new lesions at any time post baseline was also called disease progression. The Wilcoxon rank sum test was used to compare the best percent change in total LD for each arm. Since this is a secondary analysis, we will test this with alpha equal to 0.05. A waterfall plot was created to display the “best response” to treatment. This “best response” was calculated as the greatest percent change in LD (compared to baseline) for patients whose LD decreased at least once during treatment, or smallest percent increase in LD (compared to baseline) for patients whose LD never decreased compared to baseline. Subject 3 was not included in the plot since no data was collected after the study started. The best responses of all patients with post baseline LD’s were ordered from greatest to smallest and then plotted using a bar graph. 11 2.4 Progression-Free Survival (PFS) Progression Free Survival (PFS) is the time from the start of treatment to progression of disease or death. A table was created that included the patient ID, arm, date treatment started, date of progression, date of death and date of last follow-up. The dates for progression and death were checked to confirm they were before the date of last follow-up. First, the minimum (first) date that each patient started the treatment was found. Date of death and progression date were found for patients who died or progressed respectively. Lastly, the date of maximum (last) follow-up was found for each patient. If the patient died and had not previously recurred, PFS was calculated as date of death minus treatment start date. If the patient progressed, PFS was calculated as the date of disease progression minus treatment start date. Patients who were still alive and had not progressed at the time of their last follow-up were considered censored. The PFS for a patient who was censored was last follow-up date minus treatment start date. These PFS calculations showed the time in months it took for the patient to either die, progress or be censored during the study period. A Kaplan Meier plot and log rank test were used to compare the survival distributions of the two treatments. In SAS, the linear method was used to calculate the medians and the corresponding point-wise confidence intervals through the Greenwood’s formula. 3. Results This thesis summarizes patients’ baseline characteristics, toxicity, response, and PFS for each arm separately and often together. This trial enrolled the first patient on March 11, 2013. Patients were randomized (1:1) to Arm A and Arm B. After the 37 th patient was randomized, there were 18 patients assigned to Arm A and 19 patients assigned to Arm B. 17 patients were treated in Arm A and 15 were treated in Arm B. Arm A was closed after there were 0 responses in this arm. Four patients were enrolled in Arm B (without randomization) until Arm B was closed. One patient was retrospectively removed from Arm B. At the time Arm B was closed, only 1/15 had a response (partial response)-subsequently, a second patient on Arm B had a response. 3.1 Baseline Characteristics There were 41 patients enrolled in this study; 18 were randomized to Arm A, while 19 were randomized to Arm B. Once Arm A was closed for lack of sufficient activity, 4 additional patients were assigned to Arm B (for a total 23 on Arm B). These four patients were assigned to Arm B without randomization after two patients in Arm B experienced an objective response. Only 17 patients started treatment in Arm A and 19 patients started treatment in Arm B. One patient in Arm B was retrospectively removed. Hence, in this study there were 35 evaluable patients. 17 of these patients were in Arm A and 18 were in Arm B (Figure 2). In Arm A, most of the patients were male (76%), Caucasian (48%), and had a median age of 64 years. The median weight was 86 kg for those in Arm A. Also for those in Arm A, most patients had an ECOG performance status of 0 (70%) and were previously treated with Pazopanib (36%). In Arm B, most of the patients were male 12 (78%), Caucasian (66%), and had a median age of 60 years. The median weight was 82 kg for patients in Arm B. Also, for those in Arm B, most patients had an ECOG performance status of 0 (60%) and were previously treated with Bevacizumab (56%) (Table 3). As expected, due to randomization, the two arms did not differ due to baseline characteristics. Figure 2: Study Flow Diagram- Summary of number of patients in each arm for each step of the analyses 13 Using Fisher’s exact test, it was seen that the two arms did not statistically significantly differ in terms of the distribution of gender (p=1.00), race (p=0.53) and ECOG Performance status (p=0.72). Also using Fisher’s exact test, it was seen that the two arms did not statistically significantly differ in terms of either the distribution of each prior anti-VEGF therapy (p=0.29) or Bevacizumab compared to small molecules (p=0.18). Using the Wilcoxon Sign Rank test, it was seen that the two arms did not statistically significantly differ in terms of the distribution of age (p=0.23) and weight (p=0.55) (Table 3). 14 Table 3: Baseline Characteristics Arm A: AMG 386 Alone (17 Patients) Arm B: AMG 386+prior anti-VEGF therapy (18 patients) Arm A and Arm B combined (35 patients) Variable N Percentage (%) N Percentage (%) N Percentage (%) Gender p=1.00 + Male 13 76 14 78 27 78 Female 4 24 4 22 8 22 Race p=0.53 + Hispanic 5 30 4 22 9 26 Caucasian 8 48 12 66 20 58 Black 2 12 1 6 3 8 Pacific Islander 0 0 0 0 0 0 Asian 0 0 1 6 1 2 Native American 2 12 0 0 2 6 Unknown/Non- disclosed 0 0 0 0 0 0 Age (years) p=0.23 @ Median 64 - 60 - 60 - Range (min, max) 49,77 - 47,74 - 47,77 - (25th, 75th) (57,70) - (55,64) - (56,65) - Weight (kg) p=0.55 @ Median 86 - 82 - 85 - Range (min, max) 61, 155 - 54, 144 - 54,155 - (25th, 75th) (77,98) - (72,96) - (74,98) - ECOG Performance Status p=0.72 + 0 12 71 11 61 23 66 1 5 29 7 39 12 34 Prior Anti-VEGF p=0.29 +* p=0.18 +^ Bevacizumab 5 30 10 56 15 42 Pazopanib 6 36 5 28 11 32 Sorafenib 2 12 2 12 4 12 Sunitinib 4 24 1 6 5 14 • Out of 41 total patients, 35 included in analysis due to 5 patients not starting treatment, and 1 retrospectively found ineligible + p-value based on Fisher’s Exact test + * p-value based on Fisher’s Exact test, all prior anti-VEGF therapies +^ p-value based on Fisher’s Exact test, Bevacizumab vs. small molecules @ p-value based on Wilcoxon Rank Sum Test 15 3.2 Toxicity Out of the two patients who had a toxicity of grade 5, one patient suffered a myocardial infarction that was unrelated to the disease, and later died with attribution due to the disease. The other patient also suffered a myocardial infarction and later died. For both patients, there was a difference in the data presented in the database and summary assessment regarding the classification (attributed to treatment or not); ultimately both grade 5 events were not considered related to treatment. Tables and summaries include reports on all toxicities that were at least possibly attributed to one of the study drugs. The toxicities were summarized for both arms of this study separately. Most patients in both arms had a maximum toxicity grade of 1 or 2. 1 patient (2.8%) had grade 4 toxicity; this patient had Dyspnea and was in Arm A. Both arms had about an equal number of patients who had a toxicity of grade 3 (Appendix Table 1). When we collapsed the toxicity table based on toxicity systems, it was seen that there were 21 different systems ranging from blood and lymphatic system disorders to vascular disorders (Table 4.1). Gastrointestinal (10 patients (58%) in Arm A and 12 patients (67%) of Arm B) and general disorders (13 patients (76%) in Arm A and 15 patients (83%) in Arm B) had higher number of patients than any other toxicity systems. All the systems had about the same number of patients in Arm A and B. The third toxicity table showed only the toxicities where at least one patient had experienced a grade 3 or 4, or at least 2 patients had experienced that toxicity at any grade (Table 4.2). There were fourteen different toxicities for both arms. From this table, it was seen that there were more patients in both arms who had a grade 2 toxicity and more patients had fatigue (11 patients (65%) and 11 patients (61%) in both Arm A and Arm B respectively). For both arms, more patients who had a toxicity of grade three, had hypertension (3 patients (18%) in Arm A and 4 patients (22%) in Arm B). These results show that both arms had about the same number of patients for each grade and most patients had lower grade toxicity. 16 Table 4.1: Summary of Toxicities by System Toxicity System (CTCAE v 4.0) Number of patients who experienced this grade toxicity as the maximum grade for this system Arm A Arm B Grade 1 & 2 Grade 3 Grade 4 Grade 1& 2 Grade 3 Grade 4 Blood and lymphatic system disorders 5 (29%) 0 0 7 (39%) 0 0 Cardiac disorders 4 (24%) 1 0 7 (39%) 0 0 Congenital, familial and genetic disorders 0 0 0 1 0 0 Ear and labyrinth disorders 2 (12%) 0 0 0 0 0 Endocrine disorders 0 0 0 3 (17%) 0 0 Eye disorders 4 (24%) 0 0 3 (17%) 0 0 Gastrointestinal disorders 8 (47%) 2 (12%) 0 12 (67%) 0 0 General disorders and administration site conditions 11 (65%) 2 (12%) 0 13 (72%) 2 (11%) 0 Infections and infestations 2 (12%) 1 0 6 (33%) 1 0 Injury, poisoning and procedural complications 2 (12%) 0 0 1 0 0 Investigations 8 (47%) 1 0 10 (56%) 3 (17%) 0 Metabolism and nutrition disorders 9 (53%) 3 (18%) 0 13 (72%) 2 (11%) 0 Musculoskeletal and connective tissue disorders 9 (53%) 2 (18%) 0 9 (50%) 4 (22%) 0 Neoplasms benign, malignant and unspecified (incl cysts and polyps) 0 0 0 0 1 0 Nervous system disorders 9 (53%) 1 0 10 (56%) 0 0 Psychiatric disorders 4 (24%) 0 0 3 (17%) 0 0 Renal and urinary disorders 7 (41%) 0 0 9 (50%) 1 0 Reproductive system and breast disorders 0 0 0 1 0 0 Respiratory, thoracic and mediastinal disorders 11 (65%) 0 1 11 (61%) 1 0 Skin and subcutaneous tissue disorders 3 (18%) 0 0 6 (33%) 0 0 Vascular disorders 2 (12%) 3 0 10 (56%) 4 (22%) 0 17 Table 4.2: Summary of Toxicities by higher grade toxicities and higher number of patients per toxicity Toxicity System (CTCAE v 4.0) Number of patients where at least one patient had experienced a grade 3 or 4, or at least 2 patients had experienced that toxicity at any grade Arm A Arm B Toxicity System Toxicity Grade 1&2 Grade 3 Gra de 4 Grade 1&2 Grade 3 Grade 4 Gastrointestinal disorders pain 3 (18%) 2 (12%) 0 2 (11%) 0 0 General disorders and administration site conditions Fatigue 10 (59%) 1 0 10 (56%) 1 0 General disorders and administration site conditions pain 3 (18%) 1 0 3 (17%) 1 0 Investigations Lymphocyte count decreased 1 1 0 2 (11%) 2 (11%) 0 Investigations Weight gain 2 (12%) 0 0 5 (28%) 1 0 Metabolism and nutrition disorders Hypercalcemia 2 (12%) 0 0 1 1 0 Metabolism and nutrition disorders Hyperglycemia 1 2 (12%) 0 1 0 0 Metabolism and nutrition disorders Hypokalemia 0 1 0 2 (11%) 0 0 Metabolism and nutrition disorders Hyponatremia 6 (35%) 0 0 6 (33%) 1 0 Musculoskeletal and connective tissue disorders Arthralgia 2 (12%) 0 0 3 (17%) 1 0 Musculoskeletal and connective tissue disorders Pain 7 (41%) 2 (12%) 0 8 (44%) 4 0 Respiratory, thoracic and mediastinal disorders Dyspnea 6 (35%) 0 1 2 (11%) 1 0 Respiratory, thoracic and mediastinal disorders Pleural effusion 1 1 0 2 (11%) 0 0 Vascular disorders Hypertension 2 (12%) 3 3(18%) 0 8(44%) 4 (22%) 0 18 3.2.1 Predictors of Toxicities: Logistic Regression The two arms did not differ in terms of baseline characteristics (Table 5). Logistic regression was used to determine the risk of developing a hematologic and non- hematologic toxicity associated with each explanatory variable (demographics, ECOG Performance Status, prior anti-VEGF and Arm) (Table 5). There were 6/35 (17.14%) patients who had a hematologic toxicity of grade 3+ and 12/35 (34.29%) patients who had a non-hematologic toxicity of grade3+. 4/35 (11.43%) patients had both types of toxicities and 13/35 (37.14%) did not have a toxicity of grade 3+. The odds of experiencing a hematologic toxicity for those who weighed 92 kg to 155 kg were 1.49 times greater than the odds of those who weighed less than or equal to 77.8 kg; this was not statistically significantly different from 1.0 (p=0.89). The odds of experiencing a non- hematologic toxicity for those who were 92 kg to 155 kg was 4.20 times greater than those who were less than or equal to 77.8 kg (p=0.19). The odds of experiencing a hematologic toxicity for those who weighed 77.8 kg to 89.1 kg were 1.33 times greater than the odds of those who weighed less than or equal to 77.8 kg; this was not statistically significantly different from 1.0 (p=0.89). The odds of experiencing a non-hematologic toxicity for those who were 77.8 kg to 89.1 kg was 3.60 times greater than those who were less than or equal to 77.8 kg (p=0.19). For those who were older than 65 years of age the odds of experiencing a hematologic toxicity were 2.50 times greater than the odds for those who were 59 years or younger in age (p=0.5); while for those who were older than 65 years, the odds of experiencing a non-hematologic toxicity were 2.00 times greater than those who were 59 years or younger (p=0.05). For those who were between the ages 59 and 65 years, the odds of experiencing a hematologic toxicity were 2.86 times greater than the odds for those who were 59 years or younger in age (p=0.5); while for those who were between the ages 59 and 65 years, the odds of experiencing a non- hematologic toxicity were 0.11 times greater than those who were 59 years or younger (p=0.05). The odds of males experiencing a hematologic toxicity were 1.72 times greater than the odds for females (p=0.53); while the odds of males experiencing a non- hematologic toxicity were 1.55 times greater than the odds for females (p=0.59). For those who were not Caucasian the odds of experiencing a hematologic toxicity were 4.35 times greater than the odds for those who were Caucasian (p=0.08); while for those who were not Caucasian, the odds of developing a non-hematologic toxicity were 1.07 times greater than those who were Caucasian (p=0.92). For those who those who had a ECOG status of 1 the odds of experiencing a hematologic toxicity were 1.42 times greater than those who had an ECOG status of 0 (p=0.65); while for those who had an ECOG status of 1, the odds of experiencing a non-hematologic toxicity were 1.30 greater than those who had an ECOG status of 0 (p=0.71). The odds of experiencing a hematologic toxicity for those who used a prior small molecule anti-VEGF (Sunitinib, Sorafenib, Pazopanib) were 2.15 times greater than the odds of those who used the antibody –drug (Bevacizumab) (p=0.32); while the odds of experiencing a non-hematologic toxicity for those who used a prior small molecule anti-VEGF were 7.14 times greater than the odds of those who had been administered the antibody drug (p=0.14). The odds of experiencing a hematologic toxicity for those who were in Arm B was 1.09 times greater than the odds of those who were in Arm A (p=0.91); while the odds of experiencing a non-hematologic toxicity for those who were in Arm B was 2.29 times greater than the 19 odds of those who were in Arm A (p=0.23). For both hematologic and non-hematologic toxicities, the p-values were not significant (p value>0.05) for all the explanatory variables. This means that the explanatory variables do not have a statistically significant association with the risk of developing either type of toxicity. Lastly for both hematologic and non-hematologic toxicities, a stepwise approach was performed to see which explanatory variables would be in the model to predict whether someone would get either type of toxicity. For hematologic toxicities, only race was present in the model. For non-hematologic toxicities, only prior anti-VEGF was present in the final model (Table 5). Table 5 Odds Ratio for Hematologic and Non-Hematologic Toxicities Hematologic Toxicity Non-Hematologic Toxicity Explanatory Variable Odds Ratio (95% CI) P value Odds Ratio (95% CI) P value Weight 0.89 0.19 weight<=77.8 1.00 1.00 77.8<weight<=89.1 0.75 (0.13,4.49) 3.60 (0.62,21.03) 92.6<=weight=<155.0 0.67 (0.11,3.93) 4.20 (0.74, 23.91) Age 0.50 0.05 Age=<59.0 1.00 1.00 59.0<Age<=65.0 2.86 (0.41,20.14) 0.11 (0.02, 0.78) 65.0<Age 2.50 (0.36,17.32) 0.50 (0.09, 2.60) Gender 0.53 0.59 Female 1.00 1.00 Male 0.58 (0.11, 3.09) 1.55 (0.31, 7.81) Race 0.08 0.92 Caucasian 1.00 Other 0.23 (0.04, 1.31) 1.07 (0.28, 4.09) ECOG Performance Status 0.65 0.71 0 1.00 1.00 1 1.42 (0.31, 6.47) 1.30 (0.32,5.27) Prior Anti-VEGF 0.32 0.14 Bevacizumab Small Molecule Therapies 2.15 (0.45, 10.29) 0.36 (0.09-1.43) Arm 1.08 (0.25, 4.69) 0.91 0.23 Arm A 1.00 1.00 Arm B 0.92 (0.21,4.00) 2.29 (0.59,8.94) 20 3.3 Response We looked at response to treatment in two ways. One way was using the RECIST criteria and the other way was using a waterfall plot. Appendix table 2 contains all the data that these analyses are based on. Using the RECIST criteria, the tumor response was measured by the LD and nadir values. Progressive disease (PD) is when there is a new lesion or a 20 percent or more increase at the first time after baseline. If there was no PD at first assessment after baseline, then it was considered as stable disease or PR depending on the nadir value at any time. If the greatest decrease was 30% or more in nadir LD compared to baseline, this was classified as a partial response (PR); if the entire tumor disappeared, this was classified as a complete response (CR). In this study, only two patients in Arm B, experienced a response, a PR. Seven (20%) patients had an increase in LD of 20% or more at the time of their first evaluation, a PD; while 13 (37.14%) patients did not have disease progression at their first assessment after baseline, a SD. In addition, there were 3 (8.57%) patients who did not have any assessment after baseline and 10 (28.57%) patients who had clinical progression (Table 8). Initially, 7 patients had progressive disease, and 10 patients had clinical progression. At the time of this analysis, four patients are still alive and have not progressed. Eventually, most patients who had SD as their best response progressed as well as one patient with PR. Table 8: Summary of Response Type by Arm Arm A (17 patients) Arm B (18 patients) Both Arms (35 patients) Response Type Number of Subjects CR 0 0 0 PR 0 2 2 SD 5 8 13 PD 3 4 7 Not Assessed (NA) 1 2 3 Clinical Progression 8 2 10 Patients who experienced a percent decrease of 30% or more were considered to have a partial response (as long as there were no new lesions). For calculating this percent change, the nadir value (if nadir is equal to baseline LD) was not included. Subjects 25 and 39, who were in Arm B, were the only patients who had a partial response. Since only two (6%) patients had a partial response, this showed that there were not enough responses to proceed to stage 2 of the study design, and hence these two treatments were not very effective; a waterfall plot was also created to show the best percent change for each patient (Figure 3). The best response was defined as the greatest percent decrease or the smallest percent increase in LD. The red bars represent patients who were in Arm A and green bars represent patients in Arm B. The waterfall plot was created using the nadir value that did not include the baseline LD. There were only 24 patients represented on this plot. Since no percent change could be calculated for eleven 21 patients, they could not be included in the waterfall plot. The plot showed that for patients in either Arm A or Arm B, the best response was about the same due to the similar number of patients at around the same percent change. Hence, these results from the two tables and waterfall plot showed that there was not much difference in LD response for patients in either arm. The Wilcoxon rank sum test was used to compare the two arms in terms of the best percent change in total LD. Formally, the Wilcoxon Rank sum test compares the median best percent change between the two arms. If we let M A equal the median best percent change for Arm A and M B equal the median best percent change for Arm B, then we tested the null hypothesis of H o : M A =M B versus the alternative hypothesis that H A : M A ≠ M B . In this case, the p-value based on the Wilcoxon test was equal to p=0.08, indicating that the median best percent change for both arms were not statistically significantly different from each other. Figure 3: Waterfall Plot Summary of the greatest decrease (or the smallest increase)-best change for 24 patients. 11 patients had no follow-up disease assessment. Red represent patients in Arm A and green represents patients in Arm B -80 -60 -40 -20 0 20 40 Best Percent Change Compared to Baseline (%) Waterfall Plot of Greatest Tumor Shrinkage For Each Patient 22 3.4 PFS PFS was displayed using the Kaplan Meier plot; PFS for the two arms were compared using the log rank test. The Kaplan Meier plot portrays the progression in months for both Arm A (blue) and Arm B (red); the shaded areas correspond to the confidence limits. The values on the bottom of the graph indicated how many patients were at risk for each arm. The vertical drop in the graph showed indicates that at least one patient experienced an event at that time, and therefore results in a decrease in the survival probability for the patients. The tick marks on the graph represent patients who were censored (Figure 4). Overall, the log rank test showed that the null hypothesis failed to be rejected since the p-value was 0.19, indicating no statistically significant difference in survival distributions between the treatments at any time point. Out of the 35 patients, 4 were censored, 27 progressed and 4 died prior to progression. For those in Arm A, the median survival time was 8.5 months (CI: 6.8 months to 10.4 months), while the median survival time was 14.8 months (CI: 7.3 months to 28.0 months) for those in Arm B. This Kaplan Meier plot showed that both treatment groups had similar PFS probabilities. The log rank test and the Kaplan Meier plot both indicated that both treatments were similar in terms of survival. Figure 4: Kaplan Meier Plot -Summary of progression free survival in months of patients in both arms. Point-wise confidence intervals and number of patients at risk are displayed on the plot as well. Blue represent patients in Arm A and red represents patients in Arm B. 23 4. Discussion All the patients in this trial were previously taking an anti-VEGF treatment. These treatments were well established in treating RCC by the VEGF pathway. The FDA approved Sunitinib, Bevacizumab and Pazopanib, and Sorafenib in 2006, 2009, 2013, and 2013 respectively. AMG 386 is a new drug that is being tested for treating RCC by the angiopoietin- Tie pathway. This thesis analyzes the data from the Phase II clinical trial on RCC evaluating AMG 386 alone and AMG 386 in combination with a prior anti-VEGF therapy through measuring the baseline characteristics, toxicity, response, survival and logistic regression. The researchers hypothesized that AMG 386 alone and in combination with the prior anti-VEGF therapy would have anti-tumor activity in patients with metastatic RCC, by inhibiting the angiopoietin-Tie pathway. Toxicity analyses show that there were several patients who had a grade 2 toxicity and about an equal number of patients with grade 3. There was only one (2.8%) patient who had a grade 4 toxicity. Since there were many lower grade toxicities compared to higher-grade toxicities, had AMG 386 proven effective in these types of patients, clinicians would probably have been willing to administer this treatment. We wanted to look at whether variables were associated with hematologic and non-hematologic toxicities through logistic regression. Using univariate analyses, we observed that none of the explanatory variables were statistically significantly associated with the likelihood of developing Grade 3+ hematologic and non-hematologic toxicities. Using a stepwise selection method for developing a logistic model to predict grade 3+ hematologic toxicities, only race remained in the final model; using a stepwise selection method for developing a logistic model to predict Grade 3+ non-hematologic toxicity, only prior anti-VEGF remained in the final model. The p-values for both models were both greater than 0.05, indicating that these variables were not statistically significant. When inspecting table 5, the odds ratios were substantially different from 1, but were not statistically significantly different from 1 at a 0.05 significance level. This could be due to the sample size of this trial; since with a larger sample size, some of these effects would become statistically significant. There were two (5.7%) patients who experienced an objective response – a partial response (tumor shrinkage ≥ 30% - where tumor burden was measured by LD). Oncologists think that a partial response or complete response is evidence that the treatment is effective in a patient. At the 1 st tumor evaluation (after start of treatment), 7 (20%) patients had radiologic evidence of progressive disease (either a new lesion or increase in tumor burden >20%), and 13 (37.14%) patients had terminated treatment because of toxicity, clinical progression, or death. This shows that these treatments are not very effective at reducing tumor size. To display the quantitative tumor burden changes after treatment, a waterfall plot was generated to show the best percent decrease in tumor burden for each patient. RECIST is a standard way of measuring tumor size and classifying patients into responders and non-responders, while waterfall plots provide a more quantitative display of the changes in tumor burden. 24 This study was designed as two parallel trials but we compared both treatments for the purpose of this thesis. When we compared the two treatment arms in terms of progression-free survival, the log rank test was not statistically significant and the Kaplan Meier plot showed similarities in PFS probabilities between the two treatments. This shows that the two treatment arms had similar overall outcomes when used for treating patients with advanced RCC. In this trial, there were five patients out of the forty-one total patients enrolled and randomized who were not included in the study. Since these patients never started treatment, there were no implausible or missing values that affected the summaries of the data. Due to the small sample size, there was an increased chance of a higher margin of error during analysis – i.e., estimates had slightly larger standard errors. 5. Conclusion The analysis showed that AMG 386 alone or with a prior anti-VEGF therapy was not effective in treating RCC through the angiopoietin-Tie pathway. Due to the results of this study, AMG 386 is no longer being pursued for this disease in this population of patients. Further research is needed for developing therapies to treat RCC through the angiopoietin-Tie pathway in order to prevent anti-VEGF therapy resistance, produce better responses (CR and PR), and prolong the progression free survival of patients with RCC. 25 6. References 1. “Kidney Cancer: Introduction.” Cancer.Net, 27 Dec. 2017, www.cancer.net/cancer-types/kidney-cancer/introduction. 2. Semrad TJ. PhII-122: A Randomized Phase 2 Study of AMG 386 with or without Continued Anti-Vascular Endothelial Growth Factor (VEGF) Therapy in Patients with Renal Cell Carcinoma Who Have Progressed on Bevacizumab, Pazopanib, Sorafenib, or Sunitinib. (NCI Protocol # 9048) – Private Document but registered in ClinicalTrials.gov as NCT01664182) 3. MORAIS, C. Sunitinib resistance in renal cell carcinoma. Journal of Kidney Cancer and VHL, [S.l.], v. 1, n. 1, p. 1-11, 2014. Available at: <https://jkcvhl.com/index.php/jkcvhl/article/view/7/36>. Date accessed: 21 Oct. 2018. doi:http://dx.doi.org/10.15586/jkcvhl.2014.7. 4. “RECIST 1.1 – Update and Clarification: From the RECIST Committee.” NCBI, 14 May 2016, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5737828/ 5. Common Terminology Criteria for Adverse Events (CTCAE) Version 4.0. May 2009, evs.mci.nih.gov/fp1/CTCAE/CTCAE_4.03/Archive/CTCAE_4.0_2009-05- 29_QuickReference_8.5x11.pdf. 6. Oken, M.M., Creech, R.H., Tormey, D.C., Horton, J., Davis, T.E., McFadden, E.T., Carbone, P.P.: Toxicity And Response Criteria Of The Eastern Cooperative Oncology Group. Am J Clin Oncol 5:649-655, 1982. 26 7. Appendix A1 Complete Listing of Toxicities Table 1: Complete Toxicity Table Toxicity System (CTCAE v 4.0) Arm A: AMG 386 Alone Arm B: AMG 386+prior anti-VEGF therapy Toxicity System Toxicity Grade 2 Grade 3 Grade 4 Grade 2 Grade 3 Grade 4 Blood and lymphatic system disorders Anemia 5 0 0 6 0 0 Blood and lymphatic system disorders Decreased RBC count 0 0 0 1 0 0 Blood and lymphatic system disorders Leukocytosis 1 0 0 0 0 0 Cardiac disorders Myocardial infarction 0 1 0 0 0 0 Cardiac disorders Sinus bradycardia 3 0 0 5 0 0 Cardiac disorders Sinus tachycardia 1 0 0 3 0 0 Congenital, familial and genetic disorders Leukocytes in Urine 0 0 0 1 0 0 Ear and labyrinth disorders Hearing impaired 2 0 0 0 0 0 Endocrine disorders Hyperthyroidism 0 0 0 1 0 0 Endocrine disorders Hypothyroidism 0 0 0 2 0 0 Eye disorders Blurred vision 4 0 0 0 0 0 Eye disorders Dry eye 0 0 0 1 0 0 Eye disorders Glaucoma 0 0 0 1 0 0 Eye disorders Watering eyes 0 0 0 2 0 0 Eye disorders redness 0 0 0 1 0 0 Gastrointestinal disorders Ascites 2 0 0 0 0 0 Gastrointestinal disorders Colitis 0 0 0 1 0 0 Gastrointestinal disorders Constipation 3 0 0 4 0 0 Gastrointestinal disorders Diarrhea 4 0 0 5 0 0 Gastrointestinal disorders Dry heaves 1 0 0 0 0 0 Gastrointestinal disorders Dry mouth 0 0 0 1 0 0 Gastrointestinal disorders Dyspepsia 2 0 0 1 0 0 Gastrointestinal disorders Dysphagia 0 0 0 1 0 0 Gastrointestinal disorders Flatulence 0 0 0 1 0 0 Gastrointestinal disorders Gastritis 1 0 0 0 0 0 Gastrointestinal disorders Gastroesophageal reflux disease 0 0 0 1 0 0 Gastrointestinal disorders Hemorrhoids 0 0 0 1 0 0 Gastrointestinal disorders Mucositis oral 1 0 0 2 0 0 Gastrointestinal disorders Nausea 7 0 0 7 0 0 Gastrointestinal disorders Oral hemorrhage 1 0 0 0 0 0 Gastrointestinal disorders Solitary Diverticulm in Caecum 0 0 0 1 0 0 Gastrointestinal disorders Vomiting 2 0 0 2 0 0 Gastrointestinal disorders belching 1 0 0 0 0 0 27 Gastrointestinal disorders pain 3 2 0 2 0 0 General disorders and administration site conditions Actinic Keratosis 0 0 0 1 0 0 General disorders and administration site conditions Atrophic Dermatitis 0 0 0 1 0 0 General disorders and administration site conditions Chills 1 0 0 2 0 0 General disorders and administration site conditions Claudication 1 0 0 0 0 0 General disorders and administration site conditions Decreased Monocytes 0 0 0 1 0 0 General disorders and administration site conditions Edema face 1 0 0 0 0 0 General disorders and administration site conditions Edema limbs 3 0 0 8 0 0 General disorders and administration site conditions Fatigue 10 1 0 10 1 0 General disorders and administration site conditions Fever 1 0 0 0 0 0 General disorders and administration site conditions Gait disturbance 1 0 0 0 0 0 General disorders and administration site conditions General edema 0 0 0 2 0 0 General disorders and administration site conditions Hypoproteinemia 0 0 0 1 0 0 General disorders and administration site conditions Injection site reaction 1 0 0 1 0 0 General disorders and administration site conditions Irritability 1 0 0 0 0 0 General disorders and administration site conditions Leukocytes in Urine 1 0 0 1 0 0 General disorders and administration site conditions Localized edema 2 0 0 1 0 0 General disorders and administration site conditions Malaise 0 0 0 1 0 0 General disorders and Right Trochanter 0 0 0 1 0 0 28 administration site conditions Bursitis General disorders and administration site conditions Tendinitis at left knee joint 0 0 0 1 0 0 General disorders and administration site conditions pain 3 1 0 3 1 0 General disorders and administration site conditions sweaty 0 0 0 1 0 0 Infections and infestations Bronchial infection 0 0 0 1 0 0 Infections and infestations Catheter related infection 0 0 0 1 0 0 Infections and infestations Epididymidis 1 0 0 0 0 0 Infections and infestations Otitis media 0 0 0 1 0 0 Infections and infestations Rash 0 0 0 1 0 0 Infections and infestations Skin infection 0 0 0 1 0 0 Infections and infestations Suspected pneumonia 0 0 0 0 1 0 Infections and infestations Upper respiratory infection 1 0 0 2 0 0 Infections and infestations Urinary tract infection 0 1 0 1 0 0 Infections and infestations Urosepsis 0 1 0 0 0 0 Injury, poisoning and procedural complications Bruising 1 0 0 1 0 0 Injury, poisoning and procedural complications Fall 2 0 0 0 0 0 Injury, poisoning and procedural complications Fracture 0 0 0 1 0 0 Investigations Alanine aminotransferase increased 2 0 0 1 0 0 Investigations Alkaline phosphatase increased 3 0 0 1 0 0 Investigations Aspartate aminotransferase increased 2 0 0 1 0 0 Investigations Blood bilirubin increased 0 0 0 2 0 0 Investigations Cholesterol high 0 0 0 1 0 0 Investigations Creatinine increased 6 0 0 10 0 0 Investigations Hemoglobin increased 0 0 0 1 0 0 Investigations Lymphocyte count 1 1 0 2 2 0 29 decreased Investigations Neutrophil count decreased 1 0 0 1 0 0 Investigations Platelet count decreased 4 0 0 3 0 0 Investigations Weight gain 2 0 0 5 1 0 Investigations Weight loss 0 0 0 2 0 0 Investigations White blood cell decreased 1 0 0 4 0 0 Metabolism and nutrition disorders Anorexia 4 0 0 3 0 0 Metabolism and nutrition disorders Dehydration 3 0 0 2 0 0 Metabolism and nutrition disorders Hypercalcemia 2 0 0 1 1 0 Metabolism and nutrition disorders Hyperglycemia 1 2 0 1 0 0 Metabolism and nutrition disorders Hyperkalemia 4 0 0 5 0 0 Metabolism and nutrition disorders Hypertriglyceridemia 2 0 0 1 0 0 Metabolism and nutrition disorders Hypoalbuminemia 4 0 0 6 0 0 Metabolism and nutrition disorders Hypocalcemia 2 0 0 2 0 0 Metabolism and nutrition disorders Hypoglycemia 1 0 0 1 0 0 Metabolism and nutrition disorders Hypokalemia 0 1 0 2 0 0 Metabolism and nutrition disorders Hyponatremia 6 0 0 6 1 0 Metabolism and nutrition disorders Hypophosphatemia 3 0 0 0 0 0 Metabolism and nutrition disorders Obesity 0 0 0 1 0 0 Musculoskeletal and connective tissue disorders Arthralgia 2 0 0 3 1 0 Musculoskeletal and connective tissue disorders Arthritis 1 0 0 0 0 0 Musculoskeletal and connective tissue disorders Gout flare 0 0 0 1 0 0 Musculoskeletal and connective tissue disorders Myalgia 0 0 0 1 0 0 Musculoskeletal and connective tissue disorders Neurogenic claudication 1 0 0 0 0 0 Musculoskeletal and connective tissue left rotator cuff tendinitis 0 0 0 1 0 0 30 disorders Musculoskeletal and connective tissue disorders muscle weakness 5 0 0 2 0 0 Musculoskeletal and connective tissue disorders neurogenic claudication 1 0 0 0 0 0 Musculoskeletal and connective tissue disorders pain 7 2 0 8 4 0 Neoplasms benign, malignant and unspecified (incl cysts and polyps) Tumor pain 0 0 0 0 1 0 Nervous system disorders Cognitive disturbance 0 0 0 1 0 0 Nervous system disorders Dizziness 5 0 0 3 0 0 Nervous system disorders Dysarthria 1 0 0 0 0 0 Nervous system disorders Dysgeusia 0 0 0 1 0 0 Nervous system disorders Headache 5 0 0 2 0 0 Nervous system disorders Memory impairment 0 0 0 1 0 0 Nervous system disorders Paresthesia 2 0 0 2 0 0 Nervous system disorders Peripheral motor neuropathy 1 1 0 0 0 0 Nervous system disorders Peripheral sensory neuropathy 2 0 0 1 0 0 Nervous system disorders Somnolence 1 0 0 1 0 0 Nervous system disorders Tremor 1 0 0 1 0 0 Nervous system disorders jerking sensation 1 0 0 0 0 0 Nervous system disorders orthostasis 0 0 0 1 0 0 Psychiatric disorders Anxiety 1 0 0 2 0 0 Psychiatric disorders Confusion 1 0 0 1 0 0 Psychiatric disorders Depression 2 0 0 2 0 0 Psychiatric disorders Insomnia 3 0 0 2 0 0 Renal and urinary disorders Acute kidney injury 0 0 0 1 0 0 Renal and urinary disorders Chronic kidney disease 1 0 0 0 1 0 Renal and urinary disorders Hematuria 1 0 0 1 0 0 Renal and urinary disorders Nocturia 0 0 0 1 0 0 Renal and urinary disorders Proteinuria 4 0 0 8 0 0 Renal and urinary disorders Urinary Hesitancy 0 0 0 1 0 0 Renal and urinary disorders Urinary frequency 1 0 0 0 0 0 Renal and urinary disorders Urinary incontinence 2 0 0 0 0 0 31 Renal and urinary disorders Urinary retention 1 0 0 0 0 0 Renal and urinary disorders Urine discoloration 0 0 0 1 0 0 Reproductive system and breast disorders Gynecomastia 0 0 0 1 0 0 Respiratory, thoracic and mediastinal disorders Cough 5 0 0 4 0 0 Respiratory, thoracic and mediastinal disorders Dyspnea 6 0 1 2 1 0 Respiratory, thoracic and mediastinal disorders Hiccups 1 0 0 0 0 0 Respiratory, thoracic and mediastinal disorders Hoarseness 1 0 0 1 0 0 Respiratory, thoracic and mediastinal disorders Nasal congestion 0 0 0 2 0 0 Respiratory, thoracic and mediastinal disorders Pleural effusion 1 1 0 2 0 0 Respiratory, thoracic and mediastinal disorders Postnasal drip 0 0 0 1 0 0 Respiratory, thoracic and mediastinal disorders Productive cough 3 0 0 2 0 0 Respiratory, thoracic and mediastinal disorders Sleep apnea 0 0 0 1 0 0 Respiratory, thoracic and mediastinal disorders Sore throat 0 0 0 1 0 0 Respiratory, thoracic and mediastinal disorders shortness of breath 0 0 0 1 0 0 Skin and subcutaneous tissue disorders Dry skin 0 0 0 1 0 0 Skin and subcutaneous tissue disorders Erythema multiforme 0 0 0 1 0 0 Skin and subcutaneous tissue disorders Nail loss 0 0 0 1 0 0 Skin and subcutaneous tissue disorders Periorbital edema 1 0 0 0 0 0 Skin and subcutaneous tissue disorders Pruritus 1 0 0 0 0 0 Skin and subcutaneous tissue disorders Rash 2 0 0 6 0 0 Skin and subcutaneous tissue disorders Telangiectasia 0 0 0 1 0 0 Vascular disorders Flushing 0 0 0 1 0 0 Vascular disorders Hot flashes 1 0 0 0 0 0 Vascular disorders Hypertension 2 3 0 8 4 0 Vascular disorders Hypotension 2 0 0 2 0 0 Vascular disorders Lymphedema 0 0 0 1 0 0 32 A2 Table 2: Summary of Patients’ Responses Subject Name Arm Exam Date Total LD New Lesion Nadir Progression Difference Nadir1 Percent Change1 1 ArmA 6-Mar-13 27.8 0 27.8 0 0 31.3 12.58992806 1 ArmA 31-May-13 31.3 0 27.8 12.58992806 3.5 31.3 1 ArmA 22-Aug-13 32.1 0 27.8 15.4676259 4.3 31.3 1 ArmA 1-Nov-13 33.2 1 27.8 19.42446043 5.4 31.3 2 ArmA 8-Jul-13 4.4 0 3.8 15.78947368 0.6 3.8 -13.63636364 2 ArmA 11-Oct-13 3.8 0 3.8 0 0 3.8 3 ArmB 0 12.4 3 ArmB 30-Jul-13 12.4 0 12.4 0 0 5 ArmB 23-Sep-13 4.5 0 3.5 28.57142857 1 3.5 -22.22222222 5 ArmB 30-Dec-13 3.5 0 3.5 0 0 3.5 6 ArmA 23-Sep-13 10.9 0 10.9 0 0 8 ArmA 28-Oct-13 5.5 0 4.8 14.58333333 0.7 4.8 -12.72727273 8 ArmA 29-Jan-14 4.8 0 4.8 0 0 4.8 9 ArmB 3-Dec-13 9 0 8.4 7.142857143 0.6 8.4 -6.666666667 9 ArmB 21-Jan-14 8.5 0 8.4 1.190476191 0.1 8.4 33 9 ArmB 4-Mar-14 8.4 0 8.4 0 0 8.4 10 ArmA 23-Dec-13 18.6 0 18.6 0 0 20.2 8.602150538 10 ArmA 4-Apr-14 20.2 0 18.6 8.602150538 1.6 20.2 11 ArmA 6-Jan-14 8.4 0 8.4 0 0 12 ArmB 8-Jan-14 7.2 0 7.2 1.23E-14 8.88E-16 7.2 -1.23E-14 12 ArmB 16-Apr-14 7.2 0 7.2 0 0 7.2 14 ArmA 11-Mar-14 2.6 0 2.6 0 0 15 ArmB 19-Mar-14 2.5 0 2.2 13.63636364 0.3 2.2 -12 15 ArmB 1-Jul-14 2.2 0 2.2 0 0 2.2 15 ArmB 22-Sep-14 2.2 0 2.2 0 0 2.2 15 ArmB 16-Dec-14 2.7 0 2.2 22.72727273 0.5 2.2 15 ArmB 9-Mar-15 3.4 0 2.2 54.54545455 1.2 2.2 16 ArmA 14-Mar-14 9.3 0 9.3 0 0 18 ArmA 16-May-14 20.4 0 20.4 0 0 19 ArmA 28-May-14 15.6 0 15.6 0 0 20.6 32.05128205 19 ArmA 2-Sep-14 20.6 0 15.6 32.05128205 5 20.6 20 ArmB 4-Aug-14 5 0 5 0 0 21 ArmB 26-Aug-14 19.6 0 19.6 0 0 22 ArmA 26-Aug-14 7.6 0 7.6 0 0 24 ArmA 6-Nov-14 11.1 0 8.2 35.36585366 2.9 8.2 -26.12612613 24 ArmA 6-Feb-15 8.2 1 8.2 0 0 8.2 25 ArmB 18-Nov-14 7.2 0 2.9 148.2758621 4.3 2.9 -59.72222222 25 ArmB 20-Feb-15 4.6 0 2.9 58.62068966 1.7 2.9 25 ArmB 6-Apr-15 4 0 2.9 37.93103448 1.1 2.9 25 ArmB 29-Jun-15 3.4 0 2.9 17.24137931 0.5 2.9 25 ArmB 21-Sep-15 3.5 0 2.9 20.68965517 0.6 2.9 34 25 ArmB 2-Nov-15 2.9 1 2.9 0 0 2.9 26 ArmB 7-Nov-14 10.5 0 10.5 0 0 11.7 11.42857143 26 ArmB 23-Feb-15 11.7 1 10.5 11.42857143 1.2 11.7 27 ArmB 11-Nov-14 1.9 0 1.9 0 0 1.9 0 27 ArmB 6-Mar-15 1.9 1 1.9 0 0 1.9 28 ArmA 31-Dec-14 4.1 0 2 105 2.1 2 -51.2195122 28 ArmA 3-Apr-15 4.6 0 2 130 2.6 2 28 ArmA 26-Jun-15 4.7 0 2 135 2.7 2 28 ArmA 18-Sep-15 4.1 0 2 105 2.1 2 28 ArmA 11-Dec-15 4.2 0 2 110 2.2 2 28 ArmA 4-Mar-16 5.2 0 2 160 3.2 2 28 ArmA 27-May-16 2.2 0 2 10 0.2 2 28 ArmA 12-Aug-16 2 0 2 0 0 2 28 ArmA 19-Aug-16 3 0 2 50 1 2 28 ArmA 11-Nov-16 2 0 2 0 0 2 29 ArmB 28-Jan-15 16.6 0 16.6 0 0 20.1 21.08433735 29 ArmB 27-Apr-15 20.1 0 16.6 21.08433735 3.5 20.1 30 ArmA 11-Feb-15 28.7 0 28.7 0 0 31 ArmB 11-Feb-15 1.7 0 1.7 0 0 1.8 5.882352941 31 ArmB 28-May-15 1.8 0 1.7 5.882352941 0.1 1.8 31 ArmB 29-May-15 1 1.8 32 ArmA 23-Feb-15 7.1 0 7.1 1.25E-14 8.88E-16 7.1 -1.25E-14 32 ArmA 20-May-15 7.1 0 7.1 0 0 7.1 33 ArmA 10-Mar-15 4.7 0 4.3 9.302325581 0.4 4.3 -8.510638298 33 ArmA 26-May-15 23.5 0 4.3 446.5116279 19.2 4.3 33 ArmA 6-Jul-15 4.3 0 4.3 0 0 4.3 35 33 ArmA 14-Aug-15 4.4 0 4.3 2.325581395 0.1 4.3 34 ArmB 2-Apr-15 10.9 0 9.9 10.1010101 1 9.9 -9.174311927 34 ArmB 29-Jun-15 9.9 0 9.9 0 0 9.9 34 ArmB 3-Aug-15 11 0 9.9 11.11111111 1.1 9.9 35 ArmB 22-Apr-15 13.2 0 13.2 0 0 14.6 10.60606061 35 ArmB 7-Aug-15 14.6 1 13.2 10.60606061 1.4 14.6 37 ArmA 12-Jun-15 10.8 0 10.8 0 0 38 ArmB 25-Jun-15 13.6 0 13.6 0 0 15.4 13.23529412 38 ArmB 6-Oct-15 15.4 0 13.6 13.23529412 1.8 15.4 38 ArmB 28-Dec-15 16.2 0 13.6 19.11764706 2.6 15.4 38 ArmB 29-Mar-16 19.5 0 13.6 43.38235294 5.9 15.4 39 ArmB 7-Aug-15 3.8 0 1.7 123.5294118 2.1 1.7 -55.26315789 39 ArmB 11-Nov-15 2.8 0 1.7 64.70588235 1.1 1.7 39 ArmB 28-Jan-16 2.3 0 1.7 35.29411765 0.6 1.7 39 ArmB 22-Apr-16 2.3 0 1.7 35.29411765 0.6 1.7 39 ArmB 22-Jul-16 2.4 0 1.7 41.17647059 0.7 1.7 39 ArmB 14-Oct-16 2.3 0 1.7 35.29411765 0.6 1.7 39 ArmB 6-Jan-17 2.4 0 1.7 41.17647059 0.7 1.7 39 ArmB 31-Mar-17 1.7 0 1.7 0 0 1.7 40 ArmB 12-Aug-15 15 0 15 0 0 15.3 2 40 ArmB 11-Nov-15 15.3 0 15 2 0.3 15.3 41 ArmB 30-Oct-15 4.4 0 3.4 29.41176471 1 3.4 -22.72727273 41 ArmB 10-Feb-16 3.9 0 3.4 14.70588235 0.5 3.4 41 ArmB 21-Apr-16 3.4 0 3.4 0 0 3.4
Abstract (if available)
Abstract
The VEGF pathway is associated with angiogenesis and is an important target for molecular therapies for treating Renal Cell Carcinoma. Existing anti-VEGF therapies, including Bevacizumab, Pazopanib, Sorafenib, and Sunitinib, are available to treat this cancer. However, the presence of alternative pathways, such as the angiopoietin-Tie pathway, is hypothesized to lead to anti-VEGF therapy resistance. AMG 386 was developed to treat renal cell carcinoma through the angiopoietin-Tie pathway. In a Phase II randomized clinical trial, the overall response rate of AMG 386 alone and in combination with an anti-VEGF therapy was assessed. We analyzed the data from this trial to evaluate the efficacy of AMG 386. For this Phase II randomized trial, AMG 386 alone and in combination with anti-VEGF therapy were evaluated separately. The primary endpoint was overall response. For this thesis, we compared the two arms, although this was not done for the original trial. ❧ A total of 35 evaluable patients were available in this study, where we evaluated baseline characteristics, toxicity, response, and progression-free survival. Toxicity was summarized by grade and toxicity system
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The effect of renal function on toxicity of E7389 (eribulin) among patients with bladder cancer
PDF
Survival of children and adolescents with low-risk non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated with surgery only: an analysis of 234 patients from the Children’s Oncology Group stud...
PDF
A novel risk-based treatment strategy evaluated in pediatric head and neck non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) patients: a survival analysis from the Children's Oncology Group study...
PDF
Cryopreserved umbilical cord mesenchymal stem cells therapy for the treatment of knee osteoarthritis: in-vitro evaluation and phase I clinical trial protocol
PDF
A pilot study of a global approach to assessing air pollution exposure in port communities: passive air monitoring of nitrogen dioxide concentrations
PDF
Origins of the gender disparity in bladder cancer risk: a SEER analysis
PDF
The effect of inhibiting the NFkB pathway on myeloid derived suppressor cell cytokine production
PDF
Clinical outcomes of allogeneic hematopoietic stem cell transplant in acute lymphoblastic leukemia patients: a quality improvement project and systematic review meta-analysis
PDF
Capture and analysis of circulating tumor cells in patients with hepatocellular carcinoma: analysis of a pilot study
PDF
A phase I study of vincristine, escalating doses of irinotecan, temozolomide and bevacizumab (VIT-B) in pediatric and adolescent patients with recurrent or refractory solid tumors of non-hematopo...
Asset Metadata
Creator
Raghunathan, Harini
(author)
Core Title
Statistical analysis of a Phase II study of AMG 386 versus AMG 386 combined with anti-VEGF therapy in patients with advanced renal cell carcinoma
School
Keck School of Medicine
Degree
Master of Science
Degree Program
Applied Biostatistics and Epidemiology
Publication Date
04/28/2019
Defense Date
03/25/2019
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
AMG 386,anti-VEGF,clinical trial,OAI-PMH Harvest,phase II,renal cell carcinoma
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Groshen, Susan (
committee chair
), Alonzo, Todd (
committee member
), Patino-Sutton, Cecilia (
committee member
)
Creator Email
hraghuna@usc.edu,rharini1008@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-152325
Unique identifier
UC11660645
Identifier
etd-Raghunatha-7269.pdf (filename),usctheses-c89-152325 (legacy record id)
Legacy Identifier
etd-Raghunatha-7269.pdf
Dmrecord
152325
Document Type
Thesis
Format
application/pdf (imt)
Rights
Raghunathan, Harini
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
AMG 386
anti-VEGF
clinical trial
phase II
renal cell carcinoma