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
/
Physiologically-based modeling to predict monoclonal antibody pharmacokinetics in humans from physiochemical properties
(USC Thesis Other)
Physiologically-based modeling to predict monoclonal antibody pharmacokinetics in humans from physiochemical properties
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Physiologically-Based Modeling to Predict Monoclonal Antibody
Pharmacokinetics in Humans from Physiochemical Properties
by
Shihao Hu
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
(BIOMEDICAL ENGINEERING)
December 2022
Copyright 2022 Shihao Hu
Acknowledgments
I want to express my gratitude to my advisor, Dr. David D’Argenio, for giving me a chance of
working with him and learning from him in the last five years. He has provided me with consistent
guidance and kind support with great enthusiasm for science. He has taught me by action to keep
thinking of big pictures, stay positive in face of challenges, focus the energy on looking for solutions
and be considerate and helpful to other people.
I want to thank the professors on my dissertation committee: Dr. Stacey Finley, Dr. Leslie
Khawli and Dr. Megan McCain, as well as Dr. Michael Khoo who served on my qualifying exam
committee, for their valuable suggestions and comments.
I also want to thank my colleagues and collaborators: Dr. Wenbo Chen, Dr. Giovanni Pacini,
Miss. YuzhiLu,Dr. AmitaDatta-MannanandDr. JianfengLufortheirkindhelpandsuggestions.
I also would like to express my gratitude to my friends and family for their continuous love and
support.
ii
Table of Contents
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
List of Figures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii
Chapter 1: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Monoclonal antibodies: clinical applications and structure . . . . . . . . . . . . . . . 1
1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Specific aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Chapter 2: Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1 Transport mechanisms of monoclonal antibodies . . . . . . . . . . . . . . . . . . . . 7
2.1.1 Blood and lymphatic exchange and transport . . . . . . . . . . . . . . . . . . 8
2.1.2 Paracellular exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.3 Endothelial cell transport and processing . . . . . . . . . . . . . . . . . . . . 9
2.1.4 Organ specific elimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.5 Uptake and transport following subcutaneous administration . . . . . . . . . 11
2.1.6 Role of target binding on biodistribution . . . . . . . . . . . . . . . . . . . . . 11
2.2 Properties of monoclonal antibodies and their effects on transport . . . . . . . . . . 13
2.2.1 Molecular weight and physical size . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.2 Charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
iii
2.2.3 Hydrophobicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.4 Glycosylation pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.5 Fcγ R binding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.6 Interaction with FcRn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.7 Immunogenicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2.8 Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 PBPK modeling of mAb disposition and SC absorption . . . . . . . . . . . . . . . . 17
2.4 Overall goal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Chapter 3: Physiologically Based Modeling to Predict Monoclonal Antibody
Pharmacokinetics in Humans from in vitro Physiochemical Properties . . . . 21
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.1 In vitro assay data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.2 Clinical study data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.3 Model structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.4 Model parameter values - fixed . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2.4.1 Individual organ/tissue values . . . . . . . . . . . . . . . . . . . . . 31
3.2.4.2 Endosomal processing and antibody-specific values . . . . . . . . . . 33
3.2.5 Model parameter values - estimated . . . . . . . . . . . . . . . . . . . . . . . 34
3.2.6 Biophysical determinants of antibody-specific model parameters . . . . . . . 34
3.2.7 Model-based prediction of antibody PK . . . . . . . . . . . . . . . . . . . . . 34
3.2.8 Internal validation of model parameter-assay associations . . . . . . . . . . . 35
3.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3.1 Antibody-FcRn dissociation constant at pH = 6.0 (K
6.0
d
) . . . . . . . . . . . 35
3.3.2 Antibody-specificscalefactorsforpinocytosis( S
pino
)anddiffusion-convection
(S
diff− conv
) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3.3 Biophysical determinants of antibody-specific model parameters . . . . . . . 39
3.3.4 Model-based prediction of antibody PK . . . . . . . . . . . . . . . . . . . . . 39
iv
3.3.5 Internal validation of model parameter-assay correlation . . . . . . . . . . . . 40
3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Chapter 4: Predicting monoclonal antibody pharmacokinetics following subcu-
taneous administration via whole-body
physiologically-based modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.2 Materials and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2.1 Clinical study data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2.2 Model structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2.2.1 Subcutaneous administration site . . . . . . . . . . . . . . . . . . . 51
4.2.2.2 Central lymphatic system model . . . . . . . . . . . . . . . . . . . . 60
4.2.2.3 Central venous plasma . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2.2.4 Peripheral sampling site . . . . . . . . . . . . . . . . . . . . . . . . . 61
4.2.2.5 Models for other organs/tissues. . . . . . . . . . . . . . . . . . . . . 61
4.2.3 Model parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.2.3.1 Individual organs and tissues . . . . . . . . . . . . . . . . . . . . . . 62
4.2.3.2 Central venous and lymphatic systems. . . . . . . . . . . . . . . . . 62
4.2.3.3 Subcutaneous tissue site. . . . . . . . . . . . . . . . . . . . . . . . . 63
4.2.3.4 Endosomal processing and two-pore transcapillary transport . . . . 64
4.2.3.5 Baseline values for endogenous IgG . . . . . . . . . . . . . . . . . . 66
4.2.4 Model parameter estimation: IV and SC studies . . . . . . . . . . . . . . . . 66
4.2.5 Model predicted bioavailability . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.2.6 Biophysical determinants of drug specific model parameters . . . . . . . . . . 67
4.2.7 Prediction of SC PK of other mAbs . . . . . . . . . . . . . . . . . . . . . . . 68
4.2.8 Simulations and sensitivity analyses . . . . . . . . . . . . . . . . . . . . . . . 68
4.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.3.1 Intravenous pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.3.2 Subcutaneous pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . 69
v
4.3.3 Biophysical determinants of mAb SC absorption . . . . . . . . . . . . . . . . 69
4.3.4 Model-based prediction of SC PK of other mAbs . . . . . . . . . . . . . . . . 74
4.3.5 Simulations and sensitivity analyses . . . . . . . . . . . . . . . . . . . . . . . 74
4.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Chapter 5: Monoclonal Antibody Pharmacokinetics in Cynomolgus Monkeys
Following Subcutaneous Administration: Physiologically Based Model Pre-
dictions from Physiochemical Properties . . . . . . . . . . . . . . . . . . . . . . . 86
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2.1 Pharmacokinetics in cynomolgus monkeys . . . . . . . . . . . . . . . . . . . . 89
5.2.2 Structure-based physiochemical properties . . . . . . . . . . . . . . . . . . . . 90
5.2.3 PBPK model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.2.4 Model parameter values – fixed . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.2.4.1 Individual organ/tissue values . . . . . . . . . . . . . . . . . . . . . 93
5.2.4.2 Transcapillary transport and antibody-FcRn binding affinities . . . 95
5.2.5 Model parameter values – estimated . . . . . . . . . . . . . . . . . . . . . . . 96
5.2.6 Biophysical determinants of antibody specific model parameters . . . . . . . 97
5.2.7 Model simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3 Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.1 Intravenous pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.2 Subcutaneous pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.3.3 Biophysical determinants of antibody specific model parameters . . . . . . . 101
5.3.4 Model simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
5.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Chapter 6: Summary and future work . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.1 Major findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
6.2 Limitations and future work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
vi
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
vii
List of Tables
2.1 List of transport mechanisms of mAbs and relevant mAb properties . . . . . . . . . 20
3.1 List of included mAbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2 Values of the fixed physiological parameters of individual organs/tissues . . . . . . . 32
3.3 Parameters values related to endosomal processing and two-pore transport of mAbs 33
3.4 Estimated antibody-specific parameters . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.1 mAbs with plasma PK following IV administration. Also shown are the estimated
drug-specific parameters ( S
pino
and S
diff− conv
) determined using IV PK data for
each mAb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
4.2 mAbswithplasmaPKfollowingSCadministration. Estimateddrug-specificparam-
eters (CL
SC
LymCap
and S
SC
LymUpt
) in SC tissue site model using SC PK data are also
shown for each mAb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.3 Definitions of parameters and variables used in model equations . . . . . . . . . . . . 56
4.4 Fixed physiological parameters of individual organs/tissues specified in the whole-
body PBPK model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.5 Fixed parameter values for central venous and lymphatic systems . . . . . . . . . . . 63
4.6 Fixed parameter values for endosomal processing and trans-capillary transport of
mAbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.1 Selected physiochemical properties of the antibodies evaluated in this study. See
supplemental material for complete list of values for all 26 properties . . . . . . . . . 90
5.2 Values of the fixed physiological parameters for individual organs/tissues . . . . . . . 94
5.3 Fixed parameter values for central venous and lymphatic systems (see Materials and
Methods) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
viii
5.4 Parameters values related to endosomal processing and two-pore transport of mAbs 96
5.5 Medianvalues(n=3)andstandarddeviationsacrossanimalsoftheantibodyspecific
parameters and model predicted bioavailability (F, %) for each antibody . . . . . . . 100
ix
List of Figures
1.1 Schematic representation of an IgG antibody structure, with locations for antigen
binding and Fc region binding labelled . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 Schematic showing antibody extravasation via paracellular exchange . . . . . . . . . 9
2.2 Schematic of the role of FcRn in IgG processing and transport within vascular en-
dothelial cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3 Subcutaneous initial lymphatics and deeper-lying collecting lymphatics . . . . . . . . 12
2.4 First PBPK model developed by Covell and others . . . . . . . . . . . . . . . . . . . 17
2.5 The model structure of whole-body SC PBPK model for therapeutic proteins . . . . 20
3.1 Schematic of the whole-body PBPK model for mAbs following IV administration . . 27
3.2 Observed and modeled relationship between FcRn K
6.0
d
and FcRn relative retention
time (hFcRn RT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.3 Model predicted and observed antibody plasma concentration versus time profiles
following intravenous delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.4 Association between the two estimated model parameters and antibody heparin rel-
ative retention time (Hep RT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.5 Model-predicted and observed plasma concentrations of four antibodies . . . . . . . 41
4.1 Schematicofthestructureofthedevelopedwhole-bodyPBPKmodelformAbsafter
IV or SC administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2 Schematic of the organ-level model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.3 Model predicted and observed mAb plasma concentration versus time profiles after
IV administration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
x
4.4 Model predicted and observed mAb plasma concentration versus time profiles after
SC administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.5 LinearcorrelationevaluationbetweenNCAcalculatedbioavailability(%)andmodel
predicted bioavailability (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.6 Correlation of observed SC PK characteristics versus estimated SC tissue model
parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.7 Correlation of estimated model parameters versus mAb PPC metric at CDR vicinity 75
4.8 Model-predicted and digitized concentration versus time profiles of Omalizumab for
each dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.9 Model-predicted and digitized concentration versus time profiles for Tildrakizumab
for each dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
4.10 Model-predicted and digitized concentration versus time profiles for Ixekizumab . . . 77
4.11 Model-predicted and digitized concentration versus time profiles for Lanadelumab
for each dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4.12 Simulation of absorption rate of different pathways after the SC dosing, using Goli-
mumab as an example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.13 Sensitivity analysis of SC site lymph flow, using Golimumab as an example . . . . . 81
5.1 PBPKmodelusedformAbsfollowingsubcutaneous(SC)administrationincynomol-
gus monkeys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.2 Model predicted (solid lines) and observed (blank dots) antibody concentration ver-
sus time profiles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.3 Differencesinantibodyspecificparameterscanbepartlyexplainedbytheirstructure-
based metrics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.4 Simulations illustrating the use of the model to predict bioavailability based on
physicochemical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
xi
Abstract
Therapeutic antibodies (mAbs) are effective in treating many diseases including cancers, auto-
immune diseases and others. Currently more than 100 mAbs have been approved for clinical use
in the United States and Europe with more than 600 in different stages of clinical development.
Given their complex in vivo disposition mechanisms, however, antibodies can display significant
differences in their pharmacokinetics (PK) behaviors, representing a challenge for antibody de-
sign, screening and development. Although traditionally antibodies have been administered via
intravenous (IV) infusion, there is a rapid acceleration in the number of antibodies formulated for
subcutaneous (SC) administration to improve convenience and adherence compared to IV dosing.
Subcutaneous administration, however, is associated with an additional level of variability in mAb
PK related to antibody differences in bioavailability and absorption kinetics. To date, no in vitro
or preclinical in vivo methodology has been reported for reliably predicting the bioavailability and
PK of SC administered mAbs. In this dissertation research, we developed and evaluated a compre-
hensive physiologically-based pharmacokinetic (PBPK) model platform for predicting disposition
and absorption of mAbs following IV and SC administration that incorporates physiochemical an-
tibody properties derived from in vitro assays and in silico structure-based metrics. The model
includesthefundamentalmechanismsoftransportandprocessingthatgovernantibodydisposition
and absorption, including pinocytosis uptake, endosomal transit and processing, and paracellular
transport. Relevant antibody specific properties such as molecular weight, non-specific interaction,
overall charge, local charge, hydrophobicity, FcRn binding and others, as measured by in vitro
assays and in silico methods, are integrated in the model. The PBPK model prediction framework
xii
was developed using available PK data from 22 clinical trials in humans involving IV administra-
tion and 12 trials with SC administered antibodies. In addition, results from IV and SC studies
in cynomolgus monkeys using nonclinical antibodies and their Fc variants were also used in model
development. Using these studies, we have demonstrated that antibody physiochemical proper-
ties derived from in vitro studies and in silico methods can inform underlying antibody transport
processes, which in turn can be used to make model-based predictions of antibody PK. We have
found that an in vitro assay of heparin chromatography relative retention time can be used to
predict an antibody’s pinocytosis uptake and paracellular transport processes in the model. In
addition, several in silico metrics of charge and hydrophobicity can be used to quantify antibody
pre-systemic degradation, pinocytosis and paracellular transport processes. The proposed PBPK
model framework also allows for predicting complete antibody concentration-time course in plasma
and in tissues of interest, in addition to the clinical pharmacology end points of drug clearance and
bioavailability. The framework presented in this dissertation can serve as a platform for incorpo-
rating other antibody specific information for more accurate PK predictions and contribute to the
design, engineering and development of new antibody-based therapies.
xiii
1 Introduction
1.1 Monoclonal antibodies: clinical applications and structure
Since the invention of hybridoma technology in 1975, monoclonal antibodies (mAbs) have evolved
from a “magic bullet” aspiration to a clinically pivotal class of therapeutics for a broad range
of disease areas, including cancers, autoimmune diseases, infectious diseases and cardiovascular
diseases[1,2]. Toalterabiologicalfunctionandachieveatherapeuticeffect, diversemechanismsof
actionofmAbshavebeenutilized,including: (1)ligandblockade(e.g.,cytokine,growthfactor),(2)
receptor blockade and regulation (e.g., Interleukin 6 receptor, α L integrin), (3) cellular depletion
(e.g., induce apoptosis of tumor cells through effector functions), (4) signal transduction (e.g.,
induce T cell receptor-mediated signals by CD3-specific antibody) and (5) delivery of conjugated
agents (e.g., antibody-drug conjugate) [3, 4]. By 2022, there are more than 100 mAb therapeutics
in clinical use with more than 600 in various stages of clinical development [5, 6]. Given their
clinical success, a substantial pipeline and a faster approval rate by both the U.S. Food and Drug
Administration (FDA) and the European Medicines Agency (EMA), the mAb market is expected
to continue its rapid growth [7].
Immunoglobulins (Ig) are produced by B lymphocytes in vivo and can be divided into five
isotypes with different properties: IgM, IgD, IgG, IgA and IgE. Among them, IgG is most widely
usedfordrugdevelopmentandclinicaltherapy. IgGantibodiesaremadeoffourpeptidechainswith
a Y-shaped structure (Figure 1.1): two identical heavy chains (about 50 kDa) and two identical
lightchains(about25kDa). IgGantibodiesconsistoftwofunctionalparts: thefragmentofantigen
binding (Fab) and the fragment crystallizable (Fc) [8]. The Fab region includes a variable region
1
(Fv), in which each chain contains three regions of hypervariability called the complementarity
determining regions (CDRs). The CDRs determine antibody specificity and directly contact target
antigens in the binding process. The Fc region of antibodies can interact with the neonatal Fc
receptor (FcRn), the Fc-gamma receptor (Fcγ R) and complement component 1q (C1q) [9]. IgG
canbedividedintofoursubclasses: IgG1, IgG2, IgG3, IgG4, withIgG1asthemostlyusedsubclass
in mAbs. Each of the subclass can neutralize target antigens. Typically, IgG1 and IgG3 trigger
more potent effector mechanisms than IgG2 and IgG4 [10].
Because mAbs are produced in living cells like other biologics, their properties are affected not
only by their chemical structure but also by production processes. Initially, only murine mAbs
were produced with the hybridoma technology, e.g., tositumomab and ibritumomab tiuxetan [10].
To mitigate the immune responses, chimeric mAbs, with variable regions from murine sources and
constantregionsfromhumans,werecreated(e.g.,cetuximabandrituximab). Furtherhumanization
led to humanized mAbs, e.g., trastuzumab, with only CDR from murine sources. Finally, phage
display and transgenic mice technologies made the production of fully human mAbs possible [10].
Besides the design and engineering of antibody variable regions for different targets, modifications
of Fc regions can also alter effector functions [11] and increase half-life [12].
1.2 Motivation
One of the challenges in mAb development is their variable pharmacokinetics (PK). Half-lives
of different mAbs in vivo vary from hours to weeks [13]. In recent years, more mAbs have been
approvedforsubcutaneous(SC)administration, whichprovidesbetterconvenienceandcompliance
for antibody delivery compared to intravenous (IV) administration. However, SC administration
also leads to greater variability in mAb PK. SC bioavailability of different mAbs can vary from 50
to 90% and time to maximal concentration (T
max
) from 3 to 8 days [14]. The ability to predict
PK of mAbs in humans early in the drug development process would significantly contribute to the
efficientdevelopmentoftherapeuticantibodies,includingengineeringandscreeningdrugcandidates
and determining first-in-human doses.
2
Figure 1.1: Schematic representation of an IgG antibody structure, with locations for antigen
binding and Fc region binding labelled
3
Various factors affect antibody SC absorption and disposition in vivo, including molecular
weight, physical size, global charge and local charge, hydrophobicity and others. These antibody
specific physiochemical properties can be measured by in vitro assays and characterized by in silico
structure-based metrics and can be used to help predict mAb PK. However, current approaches
for mAb PK prediction have limitations. Previous studies for antibody screening have focused on
empirically using physiochemical properties to categorize antibody clearances (e.g., slow or fast).
They have not been used to predict the full mAb PK time course in plasma or tissues of interest.
Also, inter-species scaling of PK from preclinical animal models to humans, which has been applied
with high success for small molecule drugs, has been considerably less successful for protein drugs.
To date, no reliable in vitro or preclinical in vivo methodology has been reported for predicting the
bioavailability and PK of SC administered mAbs [15].
We propose to develop a comprehensive physiologically-based pharmacokinetics (PBPK) model
platformformAbPKprediction, whichintegratesinformationfromantibodyphysiochemicalprop-
erties, allowing us to more accurately predict IV and SC PK of different mAbs from in vitro assays
and in silico structure-based metrics. The underlying premise of our approach is that antibody
physiochemical properties are best related to the associated physiological mechanisms that are
responsible for their transport and processing. The proposed PBPK model framework can also
predict of the complete antibody concentration-time course in plasma and in tissues of interest.
Moreover, the PBPK model allows us to gain mechanistic insights into the factors that determine
mAb disposition and absorption and how pathophysiological factors influence antibody pharma-
cokinetics.
1.3 Specific aims
The overall goal of our work was to develop and evaluate a PBPK model platform for predicting
the disposition and absorption of mAbs following intravenous and subcutaneous administration in
humans, based on physiochemical antibody properties as determined from in vitro assays and in
silico structure-based metrics. The specific aims of this work are as follows:
4
1. Develop a mechanistic whole-body PBPK model of mAb disposition following IV administra-
tion in humans, which incorporates in vitro measures of antibody properties to predict antibody
disposition in humans (Chapter 3).
a. Includethefundamentalprocessesthatgovernantibodytransportandprocessing(e.g.,blood
and lymphatic exchange and transport, paracellular transport, endothelial cell transport and
processing)intoacomprehensivecirculatory-basedmodeldescribingthewhole-bodydisposition
of mAbs.
b. Usingantibodyplasmaconcentration-timecoursemeandatafromtheliteratureon22mAbs,
determine the model antibody-specific parameters related to underlying antibody transport
processes (paracellular convection and diffusion, non-specific binding, FcRn interaction).
c. Identify predictive relationships between estimated antibody-specific model parameters and
results from 14 in vitro assays designed to measure various antibody physiochemical properties,
includingnon-specificcell-surfaceinteractions,FcRnbinding,thermalstability,hydrophobicity,
and self-association.
d. Incorporate the identified predictive relationships into the PBPK model and conduct model-
based predictions of the plasma PK time course of each antibody.
2. Develop a PBPK model of mAbs absorption and disposition following SC antibody administra-
tioninhumansandexploretheabilityofchargerelatedantibodypropertiestopredictmodel-based
transport processes (Chapter 4).
a. Extend the PBPK model for IV administrated antibodies to include a mechanistic rep-
resentation of the pre-systemic degradation, trafficking and absorption of mAbs at the SC
administration site.
b. EstimatethemAb-specificmodelparametersusingthewholebodyPBPKmodelandhuman
IV plasma PK data from 20 mAbs and SC plasma PK data from 12 of these mAbs available in
the literature.
c. Evaluate the influence of antibody charged-related attributes (isoelectric point, local charge)
on mAb PK and combine the identified relations with the developed model to predict PK and
bioavailability of other mAbs.
5
d. Use the mechanistic model to provide a better quantitative understanding of the processes
determining the extent and rate of mAb following SC administration.
3. Develop an integrated PBPK modeling framework for predicting pharmacokinetics and bioavail-
ability of subcutaneously administered monoclonal antibodies in cynomolgus monkeys, based on
in silico structure-derived metrics characterizing antibody size, overall charge, local charge, and
hydrophobicity (Chapter 5).
a. Modify and extend the models developed in aims 1 and 2, to develop a PBPK model to
describe mAbs PK in cynomolgus monkeys following IV and SC administration.
b. Using individual animal IV and SC PK data available to us from wild type IgG4 antibodies
and their variants with Fc mutation in cynomolgus monkeys, estimate the mAb-specific model
parameters using the PBPK model.
c. Identify predictive relationships between estimated antibody specific model parameters and
antibody structure-based physiochemical metrics related to their overall and local charge, hy-
drophobicity, and others.
d. Predict bioavailability and pharmacokinetics of the different mAbs based on identified
structure-based physiochemical properties, and apply the model to explore the relative con-
tributions of mechanisms that contribute to pre-systemic antibody degradation.
6
2 Background
2.1 Transport mechanisms of monoclonal antibodies
The mechanisms of absorption, distribution, metabolism and excretion (pharmacokinetics, PK) of
mAbs differ significantly from small molecule drugs. Typically, small molecule drugs (molecular
weight less than 1 kDa) are administrated orally and eliminated mainly by liver metabolism and
renal clearance [13]. In contrast, antibody therapeutics suffer from poor oral bioavailability due
to gastrointestinal degradation and low permeation ability across the intestinal membrane. Thus,
antibodies are delivered via intravenous (IV), subcutaneous (SC) and intramuscular (IM) routes
[16]. Recently, due to the better convenience and compliance provided by the SC route compared
to IV administration, the number of mAb approvals via SC formulation has an increasing trend
[14]. Also, the mechanisms of disposition of mAbs are distinctly different, resulting in a wide
range of half-lives from hours to weeks [13]. The rate and extent of mAb distribution to peripheral
tissues are different from small molecules due to their reduced ability to cross vascular endothelium
and altered binding affinity to proteins or targets within plasma and tissues [16]. Within the
tissue,thedistributionofmAbsisdeterminedbylymphaticdrainage,endosomalcatabolism,FcRn-
mediated processing, non-specific binding and target-mediated drug disposition (TMDD). Thus,
mAbstypicallydisplayheterogeneousintra-tissuedistribution,comparedtowell-mixeddistribution
of small molecules in tissues [13]. The elimination of mAbs in vivo occurs largely by non-specific
catabolism in vascular endothelial cells, as well as in hematopoietic and potentially other cell
types, all of which involve the FcRn-mediated salvage pathway that prevents degradation of IgG
by lysosomal proteolytic pathway [16]. Compared to small molecules, specific elimination of mAbs
7
caused by TMDD and the subsequent non-linear PK is more prevalent and varies considerably
dependingontargetexpression,targetturnoveranddrug-targetinteractions. Thefollowingsections
summarize the transport mechanisms of mAbs, including: (1) blood and lymphatic transport via
convection, (2) paracellular exchange between the vascular and interstitial spaces of tissues via
diffusionandconvection, (3)extravasationthroughcapillaryendothelialcellsinvolvingpinocytosis,
processing by FcRn and lysosomal degradation, (4) organ specific elimination, (5) uptake and
transport following SC administration and (6) role of target binding.
2.1.1 Blood and lymphatic exchange and transport
Antibodies(andIgGs)distributetoandfromvascularspacesofdifferentorgansviaconvectionwith
blood. Also, antibodies within the organ interstitial space are removed by entering lymphatic cap-
illaries via lymph flow. Exchange between interstitial space and lymphatics is also via convection.
Becauselymphaticcapillarieshavelargerporesbetweencells, resistanceformAbentryismuchless
compared to that of vascular capillaries [17], leading to smaller reflection coefficient for lymphatic
capillaries. As a result, free mAb concentration within interstitial space is typically lower than in
the vascular space (ratio 0.1 to 0.5) [10]. The degradation of endogenous IgG and exogenous mAbs
in the interstitial space is assumed to be negligible given that it represents a small fraction of the
overall degradation of IgG from all organs and tissues.
2.1.2 Paracellular exchange
The movement of mAbs through paracellular gaps (pores) of endothelial cells is mediated by dif-
fusion and convection (Figure 2.1, [18]). Paracellular convection is determined by the vascular to
tissue hydrostatic gradient and the sieving effect, with the latter dependent on pore physiology,
as well as the size, shape, and charge shape of antibodies [10]. Endothelia with smaller gaps in
between(suchasbrain, muscle, skin)willgivelargerresistancetothemovementoflargemolecules.
In contract, organs with discontinuous endothelium with large fenestrations (such as liver, spleen)
will be more permissive for antibody transport, having greater exposure to mAbs. Convection is
driven by differences in blood-tissue hydrostatic and oncotic gradient from vascular space to tissue,
8
Figure 2.1: Schematic showing antibody extravasation via paracellular exchange [18]
while diffusion between the vascular and interstitial spaces depends on the molecules concentration
gradient and the surface area for exchange [10]. Based on a modeling analysis, for mAbs with
molecular weights of about 150 kDa, convection through pores between endothelial cells account
for about 80 % of total paracellular transport [19].
2.1.3 Endothelial cell transport and processing
Monoclonalantibodiescanenterendothelialcellsvianonspecificfluid-phasepinocytosisintubulovesic-
ular transport carriers, followed by endosomal fusion, endosomal processing and lysosomal degra-
dation or exocytotic recycling [20] (Figure 2.2). CH2-CH3 domain of the constant region (Fc) of
IgG can interact with several cell surface receptors, including FcRn, Fcγ receptors and components
ofthecomplementsystem[10]. Amongthem, FcRnisaMHCclassI-relatedreceptorthatinteracts
with IgG isotype antibodies and albumin within the body [20]. FcRn is a heterodimer composed of
an chain and β 2-microglobulin expressed in parenchymal (endothelial, epithelial) and hematopoi-
etic cells [20]. FcRn expression levels have been measured and can vary across different organs
and species [21]. Several residues (e.g., IIe253, His310, His435) on the CH2-CH3 domain of the Fc
9
Figure 2.2: Schematic of the role of FcRn in IgG processing and transport within vascular endothe-
lial cells [20]
region of IgG play a pivotal role in binding to FcRn. Among them, the His residues of IgG interact
with acidic residues of FcRn, and the protonation/deprotonation cycle of imidazole side chain of
His leads to pH-dependent binding. As a result, IgG and FcRn display stronger affinity at acidic
pHandweaktonegligiblebindingatneutralpH.TheweakaffinityofFcRnforIgGatnear-neutral
pH supports the pinocytosis as the primary endocytosis pathway, especially for IgGs without Fc
region engineering [20]. During the transit within endosomes, IgGs that bind to FcRn are recycled
and released by exocytosis when endosomes fuse with the cell membrane. In contrast, IgGs that
remain unbound to FcRn in sorting endosomes will be sorted to lysosomes for degradation.
2.1.4 Organ specific elimination
As mentioned, antibodies (and IgGs) are eliminated mostly through intracellular catabolism via
lysosomal degradation [10]. Renal elimination of antibodies is generally considered to be negligible,
10
due to their high molecular weights (about 150 kDa) above the glomerular filtration threshold
(about 55 kDa), except in pathologic conditions [5, 22].
2.1.5 Uptake and transport following subcutaneous administration
After SC administration, antibodies reach the hypodermis, which is composed of various cells (i.e.,
adipocytes, fibroblasts) and a network of extracellular matrix (ECM) (such as glycosaminoglycans,
collagen) with negative charges [17]. Antibodies transverse ECM through convection and diffusion
to reach blood capillaries and lymphatic capillaries (see Figure 2.3) [17]. Convection is believed to
bethemajormechanismofantibodies’migrationwithintheECMcomparedtodiffusion,asaresult
oftheirlowtransportvelocitythroughtheECM.SCsiteinterstitialfluidflowstakemacromolecules
into local lymphatic capillaries, which allow the fluid to flow in but not out [17]. Antibodies within
the lymphatic system will then drain into the systemic circulation followed by disposition processes
mentioned previously. Because the transport within lymphatic system is much faster than the
migration of antibodies within ECM, interstitial convection determines the rate of antibody SC
absorption [23]. Antibodies can also diffuse into local blood capillaries through the gaps between
endothelial cells [17]. Presystemic degradation of mAbs can occur at the SC site interstitial space,
in local lymphatic capillaries and lymphatic vessels via proteolytic enzymes, hematopoietic cells,
soluble and cell-membrane targets [23, 24]. Different factors collectively determine the degree of
SC absorption of mAbs, including SC site physiology (such as subject characteristics, target bind-
ing, local protease activities), drug formulation and antibody physiochemical properties (discussed
below) [23, 25]. Thus, SC bioavailability can vary considerably within the range from 20 to 95%
[26].
2.1.6 Role of target binding on biodistribution
Target binding of antibodies can affect their ADME behaviors, depending on expression levels and
locationsoftargets, turnoverrate, bindingaffinityandadministrateddose. Antibodiescaninteract
with receptors with either the Fc region or one of the Fab binding regions [10]. Soluble targets
within the circulation can interact with antibodies and limit their extravasation processes [10]. In
11
Figure2.3: Subcutaneousinitiallymphatics(lightershading)anddeeper-lyingcollectinglymphatics
(darker shading)in the leg of a 130-mm, 4.3-month human fetus. Arrows indicate flow direction
where valves are present. The long edge is approximately 3 mm. Abbreviation: SV, saphenous
vein [27].
12
contrast, interaction with cell surface targets can trigger endocytotic internalization of antibodies,
leading to more tissue distribution (greater volume of distribution) and non-linear target-mediated
drug disposition (TMDD) [28].
2.2 Propertiesofmonoclonalantibodiesandtheireffectsontransport
Understanding the relationship between antibody properties and their underlying transport pro-
cesses can aide in screening, designing, and developing new antibodies. Relevant properties include
molecular weight, physical size (such as hydrodynamic radius), global charge and local charge, hy-
drophobicity, glycosylation pattern, Fcγ R binding, interaction with FcRn, formulation and others.
The role of each property on mAb transport is summarized below, along with their expected effects
on the overall antibody ADME processes.
2.2.1 Molecular weight and physical size
Molecularweight(orapparenthydrodynamicsize)ofanantibodyinfluencesparacellulartransport.
The size and hydrophobic content of mAbs hinder passive diffusion across vascular endothelial cells
[5]. Tissue distribution depends on tissue capillary endothelium (discussed above) and molecular
sizes, with larger antibodies possessing smaller tissue-to-plasma ratios [25]. Based on a modeling
analysis, proteins with molecular weights of less than 30 kDa have diffusion as the major trans-
capillary transport pathway [19]. In contrast, for mAbs with molecular weights of about 150 kDa,
convection through large pores between endothelial cells account for about 80 % of total trans-
capillary transport [19]. As for elimination, increased molecular weights reduce the rate of renal
clearance of macromolecules (discussed above). The relationship between the pinocytosis rate and
molecular weight has not been reported yet [19]. The SC absorption fraction through lymphatic
capillariesishigherformacromoleculeswithhighermolecularweightssuchasmAbs,althoughsome
deviations exist probably due to species differences and injection sites [25]. However, no signifi-
cant correlation between bioavailability and molecular weight has been found [29]. Quantitative
information of SC local degradation in different locations (SC injection site, lymphatic capillaries)
13
is lacking [29]. The physical size of an antibody depends on its shape, as well as its molecular
weight. For example, molecules with polyethylene glycol (PEG) are less structured and can have
significant different ADME characteristics compared to native antibodies of a similar size, which
usually possess a globular shape [25].
2.2.2 Charge
The global charge of an antibody is determined by its amino acid sequence and the pH of its sur-
roundings. Charge-mediated interactions of the molecule can affect paracellular transport, pinocy-
tosisuptake,SCabsorptionandTMDD,thusaffectingPK.Charged-mediateddifferencesinmacro-
molecule paracellular transport has been reported in many studies, but inconsistency in specific
effects still exists [30]. In some studies, negatively charged molecules showed increased rention,
while others showed limited transport of positively charged solutes or faster transport of negative
solutes [30]. More positive charge of antibodies can also result in stronger interactions with cell
membranes (mainly negatively charged) and lead to more endocytosis, tissue retention and faster
clearance[10]. BecauseSCECMcomponentshaveanetnegativecharge,mAbswithhigherisoelec-
tric points (higher overall positive charge) display stronger interactions with SC tissues and lower
bioavailability [31, 32]. Antibodies with charge tend to have nonspecific binding with proteins or
targets within plasma or tissues, followed by internalization and degradation. It has been reported,
however, that differences in isoelectric points of more than one unit are required to produce an
appreciable change in an antibody’s systemic clearance and tissue distribution [33].
More recently, local charges have been related to ADME of antibody therapeutics. Charge
balance in an antibody’s CDR or Fv region was found to influence the PK, through direct charge-
based interactions with cell membranes [34] or excessive FcRn-IgG binding [35]. Increased local
positive charge in antibody’s complementarity-determining regions (CDR) negatively affects the
extent and rate of SC absorption in rats and cynomolgus monkeys [36].
14
2.2.3 Hydrophobicity
Higher hydrophobicity will cause higher propensity of antibodies to have hydrophobic interactions
andaggregation[37], whichcanaltertheirsizeandshapeandhindertheirtrans-capillarytransport
[36]. Also, stronger hydrophobic interactions decrease the extent and rate of SC absorption in rats
and cynomolgus monkeys [36].
2.2.4 Glycosylation pattern
Glycosylation is defined as the addition of carbohydrates (glycans) to molecules, which can hap-
pen during the expression, purification and in vivo exposure of mAbs. The reported effects of
glycosylation on antibody PK have been mixed [5].
2.2.5 Fcγ R binding
Fcγ R are cell-surface glycoproteins expressed by most immune cells, such as macrophages, mono-
cytes and dendritic cells [38]. To date, no clear effect of Fc γ R on antibody PK has been demon-
strated [39], which may be due to the low contribution of Fcγ R-mediated clearance to the total
clearance [26].
2.2.6 Interaction with FcRn
Both FcRn expression level and FcRn-IgG binding affinity affect antibody disposition in vivo. As
mentioned previously, FcRn-IgG interactions show pH dependency. During IgG’s processing and
transit within the endosomal spaces, stronger affinity at acidic pH can provide more protection
of antibodies from lysosomal degradation, thus improving half-life [40]. However, this effect can
be offset by excessive binding at neutral pH, which will limit antibody’s efficient release during
exocytosis [41]. The role of FcRn binding on SC absorption is less clear, with studies in mice
showing marked differences in bioavailability in animals with or without FcRn binding [24, 42],
while reports using cynomolgus monkeys found no change in bioavailability after Fc engineering
[43, 44].
15
2.2.7 Immunogenicity
Administration of mAbs can cause the formation of antidrug antibodies (ADAs) as an immune
response. The occurrence of ADA depends on a variety of factors, including nonhuman sequence
of the antibody, route of administration, dose and duration of therapy and patient variability.
ADA-mAb complex formation can impact ADME of affected mAbs as another clearance pathway
[10].
2.2.8 Formulation
Due to the differences in mAbs’ amino acid sequences and resulting physiochemical properties,
no universal formulation excipient exists. Individual formulation development studies need to be
carried out for mAb candidates. In general, the solubility and viscosity of mAbs in the formulation
should ensure their shelf life and stability [17]. Upon administration, the formulation is exposed to
the physiologic environment of blood or SC tissue. This change should also be considered during
the formulation development.
For SC administered antibodies, the limited volume resulting from the ECM leads to higher
concentration of mAbs (> 50 mg/ml), which increases the likelihood of protein aggregation and
presents high viscosity. Formulation pH, salt and organic solvents can be utilized to increase drug
stability[17]. Formulationexcipientslikealbuminandhypertonicsolutionhavebeenusedtomanip-
ulateoncoticpressureforlymphaticuptake,thusenhancingthemovementoffluidintheinterstitial
space to help SC absorption of proteins [12]. Co-formulation of mAbs with recombinant human
hyaluronidase hasbeenshowntoassist antibodytransportafterSC administration, bytemporarily
degrading hyaluronan within SC ECM [45]. Several studies have demonstrated that hyaluronidase
can increase the rate of SC absorption [46, 47], leading to smaller T
max
. However, the effects
of hyaluronidase on SC bioavailability have shown mixed results, with increased bioavailability in
studies with lower doses [47, 48] but no improvement with a higher dose [46].
16
Figure 2.4: First PBPK model developed by Covell and others. Left panel: Compartmental model
for IgG pharmacokinetics. Antibody is distributed to each organ according to the arterial plasma
flow to that organ ( Q). Right panel: Organ model for exchange of antibody between capillary
plasma, interstial space and cell-associated compartment. L is lymph flow, and C and V are the
antibody concentrations and distribution volumes for different compartments [53]
2.3 PBPK modeling of mAb disposition and SC absorption
For over 30 years, the use of physiologically based pharmacokinetic (PBPK) modeling has con-
tributed to advancing our understanding of the disposition of IV administered mAbs and other
therapeutic proteins in humans and in other species [49–52]. In 1986, Covell and others first ap-
plied a PBPK model based on major physiological and pharmacological features of the system,
to study the PK of an IgG1 and antibody fragments in mice (Figure 2.4) [53]. Then Baxter and
others scaled physiological parameters of mice to humans to predict clinical PK profiles of a mAb
[49]. The influence of FcRn on mAb disposition was first incorporated into a PBPK model by Ferl
and others [18] and by Garg and Balthasar [50]. Later, Chen and Balthasar proposed a catenary
PBPK model to describe antibody transit within endosomal processing, which contains different
sub-compartments to represent different pH at which antibodies interact with FcRn [54].
Themethodofmodelingantibodyextravasationpathwayshasevolved. Inthelate1980s, Rippe
and Haraldsson proposed the ”two pore hypothesis” to quantitatively describe the trans-capillary
transport of macromolecules based on previous evidence of bimodal size selectivity (Figure 2.1)
17
[55, 56]. In heteroporous membranes, fluid and macromolecules can filter through large pores
drivenbylocalhydrostaticpressureevenwhennetfiltrationislow. Thisisduetothelowreflection
coefficientoftheselargepores. Largeporefiltrationoffluidandmacromoleculesiscounter-balanced
by an absorption of protein-free fluid through small pores, forming an isogravimetric circulation
[56]. Thus, this ”volume circulation” leads to a net flux of macromolecules across capillary walls
when net filtration is zero [56]. The general forms of solute transport across ”two-pore” membrane
can be expressed as:
CL
L
=PS
L
· (1− C
IS
C
V
)· Pe
L
e
Pe
L
− 1
+J
L
· (1− σ L
) (2.1)
CL
S
=PS
S
· (1− C
IS
C
V
)· Pe
S
e
Pe
S
− 1
+J
S
· (1− σ S
) (2.2)
The Eqs. 2.1 and 2.2 show solute transport from vascular space to interstitial spaces through large
pores and small pores, respectively (more details in Chapter 4). In each equation, the former term
representsdiffusionandthelatterrepresentsconvection. Theisogravimetricflow( J
ISO
)isincluded
in the lymph flows ( J
L
and J
S
).
However, the two-pore theory model was not widely used, because many parameter values were
hard to obtain, such as permeability surface area product (PS) and isogravimetric flow ( J
ISO
).
Researchers had to estimate these parameters or used the values reported in other studies. In 2015,
Sepp and others further derived the two-pore theory related equations, and found out that two
important parameters, permeability-surface area product (PS) and isogravimetric flow ( J
ISO
), are
directly correlated with tissue lymph flow. They applied this two-pore theory model to a domain
antibody’s pharmacokinetics in mice [57]. In 2019, Li and Shah generalized this model to proteins
of different sizes and derived de novo parameters for plasma PK prediction in mice [19].
PBPKmodelshavebeenfurtherappliedtodescribetheabsorptionandPKfollowingSCadmin-
istrationofbiotherapeutics,includingmAbs. In2013,Zhaoandcoworkerscoupleda2-compartment
disposition model with a physiologically based absorption model to quantify absorption process of
mAbs after SC or IM delivery [58]. Two absorption pathways of lymphatic drainage and blood
18
perfusion were incorporated in the model, as well FcRn-mediated rescue mechanism of IgG [58].
This hybrid PBPK model provides critical quantitative understanding of mAb SC absorption and
encourage further model improvement in this direction. Nevertheless, this model cannot be em-
ployed to predict antibody concentration in the site of interest as whole-body PBPK models. In
2015, a SC depot compartment was added to a PBPK model to describe the SC PK of a pegy-
lated peptide in monkeys and scaled to humans to predict plasma PK of this peptide in a clinical
trial [59]. A limitation is that FcRn-mediated IgG processing, as a pivotal mechanism determining
antibody disposition, is not incorporated in the model [59]. A physiological absorption site model
was included in a PBPK model and used to describe the PK of different therapeutic proteins in
humans (Figure 2.5) [60]. However, the model does not account for FcRn-mediated recycling pro-
cesses in the organ/tissues. Moreover, the application of this model is limited in that it requires an
estimation of bioavailability as a prior input [60]. A PBPK model with a SC tissue space was used
to explore the effects of protein size on bioavailability and pre-systemic degradation in mice [61].
In summary, previous SC PBPK models did not account for some of the physiological processes
involved with mAb transport. In addition, there is no report of PBPK models used to predict
bioavailability and PK of SC administered mAbs that incorporate information on antibody specific
physiochemical properties.
2.4 Overall goal
The overall goal of this research was to develop a model-based framework for predicting the phar-
macokinetics of mAbs in humans following IV and SC antibody administration, based on phys-
iochemical antibody properties as determined from in vitro assays and in silico structure-based
metrics. Toward this end, a comprehensive physiologically-based pharmacokinetics model was
developed that incorporates the antibody transport processes outline in section 2.1 above and sum-
marized in column one of Table 2.1. The antibody specific physiochemical antibody properties
reviewed in section 2.2 were then assessed for their ability to explain antibody-specific differences
in model-based transport process parameters (see column two of Table 2.1). As described in detail
in the following chapters, antibody physiochemical properties can be measured by in vitro assays
19
Figure 2.5: The model structure of whole-body SC PBPK model for therapeutic proteins [60]
or characterized by in silico structure-based metrics. For example, FcRn-IgG interaction has been
characterized by surface plasmon resonance, immunoassays, cell-based approaches, and solution-
basedmethodsamongothers[62,63]. NonspecificinteractionsofmAbscanbemeasuredbybinding
poly-specificity reagent (PSR), baculovirus particles, heparin, HEK293 cells, chaperone proteins,
and cross-interaction chromatography [64]. Also, computational platforms such as Therapeutic
Antibody Profiler and Molecular Operating Environment have been developed to characterize an-
tibody size, overall charge, local charge and hydrophobicity based on mAb sequences [65, 66]. The
identified relationships between specific transport processes and physiochemical properties were
then integrated in the PBPK model, so that the resulting model platform can be used to predict
the PK of mAbs following IV or SC administration.
Table 2.1: List of transport mechanisms of mAbs and relevant mAb properties
Transport process Relevant drug properties
Blood and lymphatic exchange and transport Molecular weight and physical size
Paracellular exchange Molecular weight and physical size; Charge; Hydrophobicity
Endothelial cell transport and processing Charge; FcRn-IgG binding affinity
Organ specific elimination Molecular weight and physical size
Uptake and transport following SC administration Charge; Hydrophobicity; Formulation
Target binding Charge; Immunogenicity
20
3 Physiologically Based Modeling to
PredictMonoclonalAntibodyPhar-
macokineticsinHumansfrominvitro
Physiochemical Properties
S. Hu, A. Datta-mannan and D. Z. D’Argenio, “Physiologically Based Modeling to Predict Mon-
oclonal Antibody Pharmacokinetics in Humans from in vitro Physiochemical Properties,” mAbs,
vol. 14, no. 1, 2022.
21
Abstract
Amodel-basedframeworkispresentedtopredictmonoclonalantibody(mAb)pharmacokinetics
(PK) in humans based on in vitro measures of antibody physiochemical properties. A physi-
ologically based pharmacokinetic (PBPK) model is used to explore the predictive potential of
14 in vitro assays designed to measure various antibody physiochemical properties, including
nonspecific cell-surface interactions, FcRn binding, thermal stability, hydrophobicity, and self-
association. Based on the mean plasma PK time course data of 22 mAbs from humans reported
in the literature, we found a significant positive correlation (R = 0.64, p = .0013) between the
model parameter representing antibody-specific vascular to endothelial clearance and heparin
relativeretentiontime,anin vitro measureofnonspecificbinding. Wealsofoundthatantibody-
specific differences in paracellular transport due to convection and diffusion could be partially
explained by antibody heparin relative retention time (R = 0.52, p = .012). Other physio-
chemical properties, including antibody thermal stability, hydrophobicity, cross-interaction and
self-association, in and of themselves were not predictive of model-based transport parameters.
In contrast to other studies that have reported empirically derived expressions relating in vitro
measures of antibody physiochemical properties directly to antibody clearance, the proposed
PBPK model-based approach for predicting mAb PK incorporates fundamental mechanisms
governing antibody transport and processing, informed by in vitro measures of antibody phys-
iochemical properties, and can be expanded to include more descriptive representations of each
of the antibody processing subsystems, as well as other antibody-specific information.
Keywords FcRn interaction; nonspecific binding; antibody convective transport; two-pore the-
ory; in vitro-in vivo prediction; PBPK model; antibody physiochemical properties; antibody
biophysical characterization
22
3.1 Introduction
The ability to predict the pharmacokinetics (PK) of monoclonal antibodies (mAbs) in humans
during the early screening of drug candidates would contribute significantly to the efficient devel-
opment of therapeutic antibodies. The various factors that determine the disposition of mAbs in
humans continue to be the subject of extensive investigation [2, 12]. Recent studies have reported
empirically derived relationships between antibody clearance and experimentally determined mAb
physiochemical properties [62, 67–69]. In this report, we propose a model-based approach for pre-
dicting mAb PK that incorporates the fundamental mechanisms that govern antibody transport
and processing, coupled with in vitro assay measures of antibody physiochemical properties.
The in vivo disposition of antibodies is governed largely by their nonspecific off-target bind-
ing, affinity for neonatal Fc receptor (FcRn-IgG interaction), and target-mediated drug disposition
(TMDD) [70]. Among these antibody-specific factors, FcRn-IgG interaction has been character-
izedbysurfaceplasmonresonance(SPR),immunoassays,cell-basedapproaches,andsolution-based
methods among others [62, 63]. Given that antibodies with common crystallizable fragment (Fc)
regions and no TMDD display different PK behaviors [71], a number of approaches have been
developed to screen antibodies for their nonspecific interactions. The approaches have included
binding poly-specificity reagent (PSR), baculovirus particles (BVP), heparin, HEK293 cells, chap-
erone proteins, and crossinteraction chromatography [2, 64].
Several groups have proposed empirically derived relationships relating antibody clearance to
the aforementioned in vitro assays of physiochemical properties [62, 67, 68]. Avery and others [68]
found the combination of biophysical assays assessing nonspecific interactions, self-association, and
FcRn binding can be used to differentiate antibodies with lower and higher clearances. Goulet
and others [67] proposed that FcRn binding together with thermal stability could predict in vivo
half-life and clearance based on an analysis of eight antibodies. The study by Kraft and others
[62] further delineated the role of nonspecific cell-surface interaction on in vivo clearance of IgG,
using heparin chromatography as an in vitro surrogate. Grinshpun and others [72] investigated 12
in vitro assays, of which several in combination were found to be able to categorize the clearance
of 64 antibodies as slow or fast. Although the results reported in these studies have demonstrated
23
predictive ability to characterize the clearance or half-life of mAbs as slow or fast, they cannot
predict the full mAb PK time course in plasma or tissues of interest.
The ability of physiologically based pharmacokinetic (PBPK) models to describe the disposi-
tion of mAbs in humans and other species is now well established [51, 52, 73]. PBPK models of
therapeutic antibodies that include subsystems representing the fundamental mechanisms respon-
sible for antibody transport such as paracellular exchange, nonspecific binding, FcRn interaction
and transcytosis, can be informed by in vitro assays designed to characterize these interactions
and processes. The resulting models can be used to predict the plasma and tissue disposition of
antibodies. For example, Jones and colleagues [73] associated an affinity-capture self-interaction
nanoparticle spectroscopy (AC-SINS) assay with a model parameter that represents nonspecific
interaction affinity between mAbs and the cell membrane. In our previous study, an in silico-based
metricrepresentingthepositivechargeinantibodycomplementarity-determiningregion(PPC)was
incorporated into a PBPK model for predicting antibody PK in humans [74].
In the work reported herein, a PBPK modeling framework is used to explore the predictive
potential of 14 in vitro assays designed to measure various antibody physiochemical properties, in-
cluding nonspecific cell-surface interactions, FcRn binding, thermal stability, hydrophobicity, and
self-association. The model is developed using mean plasma PK time course data of 22 mAbs from
humans reported in the literature. For each antibody, model parameters related to paracellular
transport, nonspecific binding, and FcRn interaction are then estimated. Potential relations be-
tween estimated model parameters and the different in vitro assay results for the 22 antibodies
are explored. The established relations between antibody-specific parameters and in vitro assays
are then used to conduct PBPK model-based predictions of the plasma PK time course of each
antibody. The results demonstrate the utility of the proposed model-based framework that inte-
grates physiochemical characteristics of antibodies to predict PK profiles in humans, with the goal
of facilitating antibody screening and engineering in early development stages.
24
3.2 Materials and methods
3.2.1 In vitro assay data
ResultsfromapreviouslyreportedstudybyKraftandothers[62]wereusedtocharacterizeantibody
non-specific cell-surface interaction and FcRn interaction. As described, the variable regions of 131
antibodies approved or in clinical development were grafted onto the identical IgG1 Fc domain
and their relative retention times in heparin (Hep RT) and on human FcRn (hFcRn RT) columns
weremeasured. Inaddition, resultsfrom12physiochemicalpropertyassaysdesignedtoassessnon-
specific interaction, thermal stability, hydrophobicity and antibody self-association were obtained
from the report of Jain and others [64]. The results from these 14 assays from Kraft and others
[62] and Jain and others [64] for the 22 antibodies used in this work (see below) are provided in
the Supplemental Material.
Todetermineantibody FcRndissociationconstantatpH6.0(K
6.0
d
)oftheevaluatedantibodies,
measurements of antibody FcRn affinities (SPR assay) and human FcRn column retention time
from 37 antibodies [68] were used to construct an empirical relationship between and hFcRn RT
(data in Supplemental Material), from which the antibody FcRn dissociation constants at pH 6.0
for the 22 antibodies in this study were calculated.
3.2.2 Clinical study data
For each of the 131 antibodies reported in Kraft and others [62] and the 137 antibodies in Jain and
others [64], a literature and database search was conducted to identify those mAbs with available
plasmaconcentration-timeprofilesfollowingintravenous(IV)administration. ELISAswereusedin
these previous studies to measure the concentration of antibodies. We digitized the reported mean
plasma concentration-time course plots for those mAbs with linear kinetics (no obvious TMDD)
and with frequent sampling. The 22 antibodies identified, of which 14 were approved, are listed in
Table 3.1 along with their dosing, reported clearance, Hep RT, hFcRn RT and molecular weight.
The PK data from 16 of thesemAbs included two or more dose levels. Additional information (IgG
25
subtype, mechanism of action, global status, indications) about these 22 mAbs are provided in the
Supplemental Material.
Table 3.1: List of included mAbs
mAb (abbreviation) Dosing Reported clearance (ml/day) Hep RT hFcRn RT Molecular weight (g/mol x 10
5
) Reference
Adalimumab (Ada) 0.25–5 mg/kg 210 0.79 1.04 1.44 [75]
Atezolizumab (Ate) 1-20 mg/kg; 1200 mg 239 0.53 1.16 1.45 [76]
Bapineuzumab (Bap) 5 mg/kg 216 0.86 -0.02 1.49 [77]
Benralizumab (Ben) 0.3 mg/kg 261 0.71 0.99 1.46 [78]
Bevacizumab (Bev) 5 mg/kg 201 0.55 0.69 1.49 [79]
Canakinumab (Can) 1-3 mg/kg; 600 mg 161 0.55 0.32 1.45 [80]
Daclizumab (Dac) 200, 400 mg 286 0.84 0.02 1.43 [81]
Emibetuzumab (Emi) 700, 1400 mg 250 0.84 0.86 1.44 [82]
Enokizumab (Eno) 0.3-9 mg/kg 137 0.54 0.41 1.48 [83]
Farletuzumab (Far) 200, 400 mg/m
2
188 0.49 0.17 1.45 [84]
Figitumumab (Fig) 10, 20 mg/kg 164 0.77 1.39 1.46 [85]
Fulranumab (Ful) 3-30 mg 233 0.9 1.09 1.45 [86]
Gevokizumab (Gev) 0.01-3 mg/kg 175 0.63 0.4 1.45 [87]
Guselkumab (Gus) 0.03-10 mg/kg 338 0.87 1.24 1.44 [88]
Mepolizumab (Mep) 250 mg 157 0.6 -0.03 1.49 [89]
Olokizumab (Olo) 3 mg/kg; 10 mg/kg 116 0.21 0.14 1.5 [90]
Ozanezumab (Oza) 0.1-15 mg/kg 221 0.64 0.69 1.45 [91]
Pertuzumab (Per) 2-15 mg/kg 258 0.59 0.37 1.48 [92]
Tildrakizumab (Til) 0.1-10 mg/kg 154 0.55 0.23 1.47 [93]
Tralokinumab (Tral) 150 mg 168 0.68 0.63 1.44 [94]
Trastuzumab (Tras) 6 mg/kg 225 0.66 0.46 1.48 [95]
Ustekinumab (Ust) 0.1-5 mg/kg 148 0.68 0.13 1.49 [96]
3.2.3 Model structure
The PBPK model used in this study (Figure 3.1) is based on that reported in our previous work
[74], which itself extended and modified a mAb PBPK model reported by Glassman and Balthasar
[51]. Figure 3.1a depicts the whole-body circulatory system structure of the model, with antibody
(and IgG) distribution to and from the indicated organs via convection shown by black lines and
returned from the organs to the central venous plasma pool through convection via the lymphatic
flow (grey lines).
Figure 3.1b shows the model used to represent administered mAb and endogenous IgG dispo-
sition in each organ/tissue, and includes vascular, endothelial, interstitial and cellular spaces, with
transcapillary exchange via paracellular transport and pinocytosis. For generality, and to allow
for future extension, a two-pore model was used to describe the paracellular transport of endoge-
nous IgG and administrated mAbs as detailed in our previous report [74] (following that presented
26
Figure 3.1: Schematic of the whole-body PBPK model for mAbs following IV administration.
(a) Overall circulatory model. (b) Organ-level structure of a typical tissue, including vascular
space, endothelial space, interstitial space and cell space. Paracellular transport via convection and
diffusion through pores is depicted along with endothelial transport and process. (c) Endosomal
sub-compartments, showing the competition between endogenous IgG and mAb for FcRn at a
specific pH value
27
by Li and Shah [19]). The following mass balance equations describe the organ vascular space
concentrations of endogenous IgG (CIgG
Organ
vasc
) and mAb (CmAb
Organ
vasc
):
V
Organ
vasc
· dCIgG
Organ
vasc
dt
=Q
Organ
· CIgG
Arterial
vasc
− (Q
Organ
− L
Organ
)· CIgG
Organ
vasc
− CL
Organ
pino
· CIgG
Organ
vasc
+fIgG
recyc
· CL
Organ
pino
· CIgFc
Organ
recyc
− PS
Organ
L
· (CIgG
Organ
vasc
− CIgG
Organ
inter
)· Pe
L
e
PeL
− 1
− PS
Organ
S
· (CIgG
Organ
vasc
− CIgG
Organ
inter
)· Pe
S
e
PeS
− 1
− J
Organ
L
· (1− σ L
)· CIgG
Organ
vasc
− J
Organ
S
· (1− σ S
)· CIgG
Organ
vasc
(3.1)
V
Organ
vasc
· dCmAb
Organ
vasc
dt
=Q
Organ
· CmAb
Arterial
vasc
− (Q
Organ
− L
Organ
)· CmAb
Organ
vasc
− Spino · CL
Organ
pino
· CmAb
Organ
vasc
+Spino · fIgGrecyc · CL
Organ
pino
· CIgFc
Organ
recyc
− S
diff− conv
· PS
Organ
L
· (CmAb
Organ
vasc
− CmAb
Organ
inter
)· PeL
e
PeL
− 1
− S
diff− conv
· PS
Organ
S
· (CmAb
Organ
vasc
− CmAb
Organ
inter
)· PeS
e
PeS
− 1
− S
diff− conv
· J
Organ
L
· (1− σ L)· CmAb
Organ
vasc
− S
diff− conv
· J
Organ
S
· (1− σ S)· CmAb
Organ
vasc
(3.2)
The symbols used to represent concentration variables and parameters involved in modeling the
transport processes are provided in the Supplemental Material. The first line of Eq. 3.1 repre-
sents IgG transport via blood perfusion, the second line accounts for IgG exchange via pinocytosis
and exocytosis (discussed below), while the last two lines represent IgG’s paracellular transport
via diffusion and convection based on the two-pore model. A similar equation representing mAb
exchange (Eq. 3.2) includes scaling factors to represent antibody-specific differences in transport
relative to IgG, including a scaling term for pinocytosis (S
pino
) and a single term for diffusion and
convection (S
diff− conv
), following the approach reported by Glassman and Balthasar [51] and used
in our previous work [74]. These two antibody-specific scaling factors were estimated as described
below.
The corresponding mass balance for the concentrations of mAb and endogenous IgG in the
interstitial space (CIgG
Organ
inter
and CmAb
Organ
inter
) yields:
V
Organ
inter
dCIgG
Organ
inter
dt
=(1− fIgG
recyc
)· CL
Organ
pino
· CIgFc
Organ
recyc
− CL
Organ
pino
· CIgG
Organ
inter
− L
Organ
· CIgG
Organ
inter
+PS
Organ
L
· (CIgG
Organ
vasc
− CIgG
Organ
inter
)· Pe
L
e
PeL
− 1
+PS
Organ
S
· (CIgG
Organ
vasc
− CIgG
Organ
inter
)· Pe
S
e
PeS
− 1
+J
Organ
L
· (1− σ L
)· CIgG
Organ
vasc
+J
Organ
S
· (1− σ S
)· CIgG
Organ
vasc
(3.3)
V
Organ
inter
dCmAb
Organ
inter
dt
=Spino · (1− fIgGrecyc)· CL
Organ
pino
· CmAbFc
Organ
recyc
− Spino · CL
Organ
pino
· CmAb
Organ
inter
− L
Organ
· CmAb
Organ
inter
+S
diff− conv
· PS
Organ
L
· (CmAb
Organ
vasc
− CmAb
Organ
inter
)· PeL
e
PeL
− 1
+S
diff− conv
· PS
Organ
S
· (CmAb
Organ
vasc
− CmAb
Organ
inter
)· PeS
e
PeS
− 1
+S
diff− conv
· J
Organ
L
· (1− σ L)· CmAb
Organ
vasc
+S
diff− conv
· J
Organ
S
· (1− σ S)· CmAb
Organ
vasc
(3.4)
28
See Supplemental Material for definition of all symbols.
Themodelusedforendosomaltransitandprocessing(Figure3.1band1c)isanupdatedversion
ofthatusedinourpreviousworkandisbasedonfurtherinsightintothephysiologicalprocessesgov-
erning IgG-FcRn interaction as reported by Ward and Ober [20]. Three endosomal compartments
were used to represent separate stages of IgG endosomal transit (Figure 3.1b): early endosome (pH
7.4a), late endosome (pH 6.0) and cell membrane (pH 7.4b). Within each endosomal compartment,
endogenous IgG or exogenous mAbs can interact with FcRn in a pH-dependent manner [74], via
mass action association and dissociation rate constants, k
pHx
on
and k
pHx
off
, as depicted in Figure 3.1c.
Free IgG in the late endosome (pH 6.0) will be sorted to the lysosome for degradation (represented
by “CL” term in Figure 3.1b), while FcRn-bound IgG is designated to the cell membrane compart-
ment (pH 7.4b) for recycling or exocytosis. In the cell membrane compartment, a fraction of free
IgG (fIgG
recyc
) is recycled to the vascular space, while that bound to FcRn (FcRn-IgG complex)
remains in the endosome (early endosome compartment). Free FcRn is assumed to move from
the early endosome to subsequent compartments, with internalization from the last cell membrane
compartment (pH 7.4b) to the first early endosome compartment (pH 7.4a). Based on these as-
sumptions, the following equations describe the concentrations of the species in the early endosome
compartment (CIgG
Organ
7.4a
, CIgFc
Organ
7.4a
, CmAb
Organ
7.4a
, CmAbFc
Organ
7.4a
and CFcRn
Organ
pH7.4a
):
dCIgG
Organ
7.4a
dt
=(CL
Organ
pino
· CIgG
Organ
vasc
+CL
Organ
pino
· CIgG
Organ
inter
)/V
Organ
endo,sub
+k
7.4
off
· CIgFc
Organ
7.4a
− k
7.4
on
· CIgG
Organ
7.4a
· CFcRn
Organ
7.4a
− 1
τ · CIgG
Organ
7.4a
(3.5)
dCIgFc
Organ
7.4a
dt
=k
7.4
on
· CIgG
Organ
7.4a
· CFcRn
Organ
7.4a
− k
7.4
off
· CIgFc
Organ
7.4a
+
1
τ · (CIgFc
Organ
7.4b
− CIgFc
Organ
7.4a
)
(3.6)
dCmAb
Organ
7.4a
dt
=(S
pino
· CL
Organ
pino
· CmAb
Organ
vasc
+S
pino
· CL
Organ
pino
· CmAb
Organ
inter
)/V
Organ
endo,sub
+k
7.4
off
· CmAbFc
Organ
7.4a
− k
7.4
on
· CmAb
Organ
7.4a
· CFcRn
Organ
7.4a
− 1
τ · CmAb
Organ
7.4a
(3.7)
dCmAbFc
Organ
7.4a
dt
=k
7.4
on
· CmAb
Organ
7.4a
· CFcRn
Organ
7.4a
− k
7.4
off
· CmAbFc
Organ
7.4a
+
1
τ · (CmAbFc
Organ
7.4b
− CmAbFc
Organ
7.4a
)
(3.8)
29
dCFcRn
Organ
pH7.4a
dt
=k
7.4
off
· CIgFc
Organ
7.4a
− k
7.4
on
· CIgG
Organ
7.4a
· CFcRn
Organ
7.4a
+k
7.4
off
· CmAbFc
Organ
7.4a
− k
7.4
on
· CmAb
Organ
7.4a
· CFcRn
Organ
7.4a
+
1
τ · (CFcRn
Organ
pH7.4b
− CFcRn
Organ
pH7.4a
)
(3.9)
The following equations describe the concentrations of the different species in the late endosome
compartment (pH 6.0):
dCIgG
Organ
6.0
dt
=k
6.0
off
· CIgFc
Organ
6.0
− k
6.0
on
· CIgG
Organ
6.0
· CFcRn
Organ
6.0
+
1
τ · (CIgG
Organ
7.4a
− CIgG
Organ
6.0
)
(3.10)
dCIgFc
Organ
6.0
dt
=k
6.0
on
· CIgG
Organ
6.0
· CFcRn
Organ
6.0
− k
6.0
off
· CIgFc
Organ
6.0
+
1
τ · (CIgFc
Organ
7.4a
− CIgFc
Organ
6.0
)
(3.11)
dCmAb
Organ
6.0
dt
=k
6.0
off
· CmAbFc
Organ
6.0
− k
6.0
on
· CmAb
Organ
6.0
· CFcRn
Organ
6.0
+
1
τ · (CmAb
Organ
7.4a
− CmAb
Organ
6.0
)
(3.12)
dCmAbFc
Organ
6.0
dt
=k
6.0
on
· CmAb
Organ
6.0
· CFcRn
Organ
6.0
− k
6.0
off
· CmAbFc
Orgna
6.0
+
1
τ · (CmAbFc
Organ
7.4a
− CmAbFc
Organ
6.0
)
(3.13)
dCFcRn
Organ
pH6.0
dt
=k
6.0
off
· CIgFc
Organ
6.0
− k
6.0
on
· CIgG
Organ
6.0
· CFcRn
Organ
6.0
+k
6.0
off
· CmAbFc
Organ
6.0
− k
6.0
on
· CmAb
Organ
6.0
· CFcRn
Organ
6.0
+
1
τ · (CFcRn
Organ
pH7.4a
− CFcRn
Organ
pH6.0
)
(3.14)
Similarly, the following equations apply to the cell membrane compartment (pH 7.4b):
dCIgG
Organ
7.4b
dt
=k
7.4
off
· CIgFc
Organ
7.4b
− k
7.4
on
· CIgG
Organ
7.4b
· CFcRn
Organ
7.4b
− CL
Organ
pino
· CIgG
Organ
7.4a
/V
Organ
endo,sub
(3.15)
dCIgFc
Organ
7.4b
dt
=k
7.4
on
· CIgG
Organ
7.4b
· CFcRn
Organ
7.4b
− k
7.4
off
· CIgFc
Organ
7.4b
+
1
τ · (CIgFc
Organ
6.0
− CIgFc
Organ
7.4b
)
(3.16)
dCmAb
Organ
7.4b
dt
=k
7.4
off
· CmAbFc
Organ
7.4b
− k
7.4
on
· CmAb
Organ
7.4b
· CFcRn
Organ
7.4b
− S
pino
· CL
Organ
pino
· CmAb
Organ
7.4a
/V
Organ
endo,sub
(3.17)
dCmAbFc
Organ
7.4b
dt
=k
7.4
on
· CmAb
Organ
7.4b
· CFcRn
Organ
7.4b
− k
7.4
off
· CmAbFc
Organ
7.4b
+
1
τ · (CmAbFc
Organ
6.0
− CmAbFc
Organ
7.4b
)
(3.18)
30
dCFcRn
Organ
pH7.4b
dt
=k
7.4
off
· CIgFc
Organ
7.4b
− k
7.4
on
· CIgG
Organ
7.4b
· CFcRn
Organ
7.4b
+k
7.4
off
· CmAbFc
Organ
7.4b
− k
7.4
on
· CmAb
Organ
7.4b
· CFcRn
Organ
7.4b
+
1
τ · (CFcRn
Organ
pH6.0
− CFcRn
Organ
pH7.4b
)
(3.19)
The lymph node compartment in Figure 3.1, representing all the body’s lymph nodes, collects
the lymph drainage from all organs. The post-nodal lymph flow travels to a lumped lymphatic
vessel (representing lymphatic trunk, thoracic duct and cysterna chyli) before entering the central
venous pool. The mass balance equations for the amounts and concentrations of endogenous IgG
and exogenous mAbs in the lymph node and lymphatic vessel compartments are the same as our
previous report [74].
As reported in our previous modeling work [74], a sampling site compartment was included
to better reflect the skin and muscle contributions to antibody concentrations at the peripheral
sampling site, where CmAb
Peri
=0.7· CmAb
Skin
vasc
+0.3· CmAb
Muscle
vasc
.
The complete set of equations and definition of all symbols used in the model, including all
organs, is provided in the Supplemental Material.
3.2.4 Model parameter values - fixed
3.2.4.1 Individual organ/tissue values
Valuesfororgan/tissuevolumes(vascular,interstitial,cell)wereobtainedbycombininginformation
from various sourses [49, 52, 97–99]. The organ weights (without blood) for a representative 70-kg
human were taken from the report by Sowby [99], while reported [97] tissue densities were used
to calculate organ volumes. Previously reported interstitial volume fractions [52] were adopted
to calculate the interstitial volumes (V
Organ
inter
) and cell volumes (V
Organ
cell
) given in Table 3.2. Total
plasmavolumeof2906mlwasobtainedassumingabloodmassof5.6kg,blooddensityof1.06g/ml
and hematocrit of 45%. Based on reported blood distribution in the vascular system [99], total
plasma volume was partitioned as follows: lung capillaries (2%), systemic organ/tissue capillaries
(5%), central venous space (67%, including pulmonary artery, heart right chamber, large veins
31
and organ veins) and arterial space (26%, including pulmonary vein, heart left chamber, large
arteries and organ arteries). The individual organ vascular volumes (V
Organ
vasc
) listed in Table 3.2
were calculated by partitioning the total capillary volume using the blood volume distribution in
adults reported by Sowby and others [99].
The total cardiac output of plasma (Q
Total
) was assumed to be 5.148 x 10
6
ml/day, consis-
tent with the values reported in different sources [97–99]. Based on reported organ blood flow
distribution values [99], the organ-specific plasma flow rates ( Q
Organ
) listed in Table 3.2 were cal-
culated. The lymph flow of each organ ( L
Organ
) given in Table 3.2 was assumed to be 0.078% of
the corresponding plasma flow, to yield a total pre-nodal lymph flow of approximately 8 L/day
[27].
Assuming central venous and arterial plasma occupy 67% and 26% of the total plasma volume
yields central venous and arterial volumes of 1947 and 755 ml, respectively. The values for the
lymphatic system parameters, including volumes and flows were determined as described in our
previous work [74]. The Supplemental Material provides a complete list of all model parameters
and their values.
Table 3.2: Values of the fixed physiological parameters of individual organs/tissues
Organ/tissue Interstitial volume (ml)
a
V
Organ
inter
Cell volume (ml)
a
V
Organ
cell
Vascular volume (ml)
a
V
Organ
vasc Plasma flow (ml/day) Q
Organ
Lung 151.2 325 58.11 5.148 x 10
6
Heart 47.67 272.7 2.571 2.059 x 10
5
Liver 382.5 1365 25.71 3.346 x 10
5
Spleen 32.59 110.3 3.6 1.544 x 10
5
GI Tract 335.1 1636 18 7.722 x 10
5
Kidney 46.85 248.4 5.143 9.781 x 10
5
Muscle 3702 24155 36 8.752 x 10
5
Skin 1099 2105 7.714 2.574 x 10
5
Other 4677 21479 46.54 1.570 x 10
6
Organ/tissue Lymph flow
b
L
Organ
FcRn (nM)
c
CFcRn Vascular IgG clearance via pinocytosis (ml/day)
d
CL
Organ
pino
Endosomal volume
d
V
Organ
endo
Lung 4.015 x 10
3
33000 1.083 8.120 x 10
− 3
Heart 1.606 x 10
2
20200 0.7286 5.460 x 10
− 3
Liver 2.610 x 10
2
33000 3.975 2.980 x 10
− 2
Spleen 1.205 x 10
2
33000 0.3249 2.430 x 10
− 3
GI Tract 6.023 x 10
2
4180 4.484 3.360 x 10
− 2
Kidney 7.629 x 10
2
33000 0.6714 5.036 x 10
− 3
Muscle 6.826 x 10
2
33000 63.36 4.751 x 10
− 1
Skin 2.008 x 10
2
33000 7.286 5.460 x 10
− 2
Other 1.225 x 10
3
33000 59.49 4.461 x 10
− 1
a
Based on previous literature, as detailed in text
b
Fixed to be 0.078% of organ plasma flow, with a total prenodal lymph flow
˜
8000 ml/day as
reported.
c
Fixed based on the FcRn concentration measured by Li and others [21] or from Glassman and
others [51]
d
Calculated, as detailed in text
32
3.2.4.2 Endosomal processing and antibody-specific values
The basal values of FcRn concentration for each organ, listed in Table 3.2, were based on measure-
ments reported in Li and others [21] or taken from Glassman and others [51], as detailed in our
previous work [74]. Parameter values related to the two-pore transport model depend on molecular
weights, or organ lymph flow, or both. Values for these parameters were calculated as described
previously [74] and are listed in Table 3.3, along with other two-pore transport model parameter
values.
ThebindingaffinitytoFcRnatpH6.0wasassumedtobeantibody-specificandwasdetermined
for each of the 22 antibodies in our study based on a relation derived between mAb equilibrium
dissociation constant K
6.0
d
and hFcRn RT. Using the K
6.0
d
and hFcRn RT data from 37 mAbs as
noted above, anexponential (decaying) relation was determined using the ID application(weighted
least squares estimation) in the ADAPT software (version 5) [100]. The resulting relationship was
used to determine K
6.0
d
for each of the antibodies used in this study given their known measured
hFcRn RT. Assuming a value of k
6.0
off
of 573.1 day
− 1
for all antibodies [51], k
6.0
on
of each antibody
was estimated as: k
6.0
on
=k
6.0
off
/K
6.0
d
.
Table 3.3: Parameters values related to endosomal processing and two-pore transport of mAbs
Parameter (unit) Value Description Ref.
All organs, tissues and mAbs
τ (day) 0.00375 Transit time of IgG between endosomal sub-compartments [51]
fIgGrecyc 0.715 IgG recycling fraction of IgG-FcRn complex back to vascular space [50]
k
6.0
on
(nM
− 1
day
− 1
) antibody-specific Association rate constant between IgG and FcRn at pH = 6.0 [51]
k
6.0
off
(day
− 1
) 573.1 Dissociation rate constant between IgG and FcRn at pH = 6.0 [51]
k
7.4
on
(nM
− 1
day
− 1
) 0.06336 Association rate constant between IgG and FcRn at pH = 7.4 [51]
k
7.4
off
(day
− 1
) 573.1 Dissociation rate constant between IgG and FcRn at pH = 7.4 [51]
rS (nm) 4.44 Small pore radius [19]
rL (nm) 22.85 Large pore radius [19]
α S 0.958 Fractional hydraulic conductance of small pores [19]
α L 0.042 Fractional hydraulic conductance of large pores [19]
Adjusted for mAb MW (150 kDa IgG as an example); All organs and tissues
ae (nm) 4.81 Stokes Einstein radius [19]
σ S 0.998 Small pore vascular reflection coefficient [19]
σ L 0.18 Large pore vascular reflection coefficient [19]
A/A0S 9.280 x 10
− 7
Fractional accessible pore size of small pore [19]
A/A0L 0.349 Fractional accessible pore size of large pore [19]
PeS 9.82 Peclet number of small pores [19]
PeL 4.48 Peclet number of large pores [19]
Adjusted for mAb MW and individual organs and tissues (150 kDa IgG in liver as an example)
PS
Liver
S
(ml/day) 0.03229 Permeability-surface area product of small pores [19]
PS
Liver
L
(ml/day) 20.1 Permeability-surface area product of large pores [19]
J
Liver
iso
(ml/day) 99.18 Circular isogravimetric flow [19]
J
Liver
S
(ml/day) 150.8 Lymph flow through small pores [19]
J
Liver
L
(ml/day) 110.1 Lymph flow through large pores [19]
33
3.2.5 Model parameter values - estimated
As indicated above, model equations describing endogenous IgG and its interactions with FcRn
were included into the model. Assuming an endogenous IgG production rate (IgG0) of 1.540
x 10
4
nmol/day as reported in the ref. [101], the whole-body pinocytosis rate (CL
Total
pino
) was
then determined to achieve a steady state concentration of 12.1 mg/ml [101]. Individual organ
pinocytosis rates were assumed to be in proportion to their individual total organ volumes and
calculated as follows: CL
Organ
pino
=
V
Organ
Total
/V
Total
· CL
Total
pino
(Table 3.2). The endosomal volume
ofeachorganwasthencalculatedusingV
Organ
endo
=2· CL
Organ
pino
· τ [102](Table3.2). Foreachantibody
listed in Table 3.1, the two antibody-specific scaling factors, S
pino
and S
diff− conv
, were estimated
using on the plasma concentration time data described above (Table 3.4). The ADAPT software
(version 5) was used for parameter estimation, with maximum likelihood estimation (additive plus
proportional error variance model) [100]. Model simulation was first conducted to reach the steady
stateofendogenousIgGwithinthesystembeforeantibodyIVadministration. Thena¨ ıvepooldata
(NPD) application was used for antibodies that were administer at more than one dose, while the
ID application was used for antibodies given at one dose amount.
3.2.6 Biophysical determinants of antibody-specific model parameters
Potential associations between the 14 assay readouts and each of estimated values of the two model
parameters,S
pino
andS
diff− conv
, wereexploredvialinearregressionanalysisusingthe“lm”(linear
model) function in the “stats” package as part of R [103]. The selection of significant biophysical
assays was based on p values from the regression analysis.
3.2.7 Model-based prediction of antibody PK
As an illustration of the application of the developed PBPK model prediction framework, we
incorporated the identified relationships between the two antibody-specific scale factors ( S
pino
and
S
diff− conv
) and the physiochemical properties identified as significant into the PBPK model, to
conduct model-based predictions of plasma PK for each of the 22 antibodies. For each antibody,
34
a population simulation was also conducted (n = 1000) using the SIM application in ADAPT)
[100], assuming a log-normal distribution of S
pino
and S
diff− conv
. Mean values for the parameter
distributions were determined as the predicted values of S
pino
and S
diff− conv
obtained from the
identified relationship between these parameters and those physiochemical properties identified as
significant. Thestandarderrorsofthelinearregressionmodelpredictionswereusedasthestandard
deviation of S
pino
and S
diff− conv
in the population simulations.
3.2.8 Internal validation of model parameter-assay associations
Asaninternalvalidationoftheidentifiedmodelparameter-assayassociations,werandomlyselected
18mAbs(approximately80%ofthetotal)andusedtheseantibodiestodeterminenewrelationships
between the model parameter scaling factors and the physiochemical properties determined to be
significant as described above. This random selection was repeated (n = 100) and the resulting
and p values were summarized.
3.3 Results
ThePBPKmodelusedinthisstudyisdepictedinFigure3.1anddescribedindetailintheMaterials
and Methods section.
3.3.1 Antibody-FcRn dissociation constant at pH = 6.0 (K
6.0
d
)
The following exponential relation between K
6.0
d
and hFcRn RT was determined based on the data
from37mAbs[68]: K
6.0
d
=K0
6.0
d
· e
− λ ·hFcRn RT
, whereK0
6.0
d
=1136,λ =0.52withR=0.79(Figure
3.2). Based on this relationship, the values for K
6.0
d
for the 22 mAbs that are subject of this study
were calculated and are listed in Table 3.4.
35
Figure 3.2: Observed and modeled relationship between FcRn K
6.0
d
and FcRn relative retention
time(hFcRn RT).Thedotsrepresentavailableassayresultsof37antibodies[68]. Thelineindicates
the nonlinear regression relationship: K
6.0
d
=1136· e
− 0.52·hFcRn RT
, R = 0.79
36
Table 3.4: Estimated antibody-specific parameters
Abbreviation K
6.0
d
(nM) (RSE%) S
pino
(RSE%) S
diff− conv
(RSE%)
Ada 663 (49) 1.17 (3.5) 1.08 (4.5)
Ate 623 (52) 0.93 (7.5) 0.81 (6.4)
Bap 1148 (28) 0.99 (2.2) 1.16 (8.0)
Ben 680 (48) 1.31 (4.2) 0.98 (7.1)
Bev 795 (41) 1.14 (2.2) 0.68 (8.7)
Can 963 (34) 0.90 (1.1) 0.88 (4.1)
Dac 1125 (29) 1.26 (1.4) 1.06 (5.7)
Emi 727 (45) 1.09 (17.9) 0.87 (9.9)
Eno 919 (35) 0.86 (2.4) 0.76 (4.0)
Far 1041 (31) 1.14 (8.1) 0.48 (14)
Fig 552 (59) 0.94 (1.9) 0.92 (6.6)
Ful 646 (50) 1.15 (2.4) 1.24 (4.5)
Gev 924 (35) 0.84 (2.6) 0.69 (3.7)
Gus 597 (54) 1.45 (1.2) 1.13 (2.9)
Mep 1154 (28) 0.95 (2.5) 0.48 (8.5)
Olo 1057 (31) 0.61 (2.4) 0.88 (4.1)
Oza 795 (41) 1.10 (1.3) 0.93 (5.0)
Per 938 (35) 1.09 (7.7) 0.90 (6.8)
Til 1009 (32) 0.81 (1.0) 0.75 (4.0)
Tral 820 (40) 0.94 (3.6) 0.55 (9.4)
Tras 895 (36) 1.02 (11.2) 0.61 (12)
Ust 1062 (31) 0.84 (1.6) 0.51 (5.3)
K
6.0
d
: FcRn-mAb dissociation rate constant at pH = 6.0
S
pino
: scale factors for antibody-specific endothelial pinocytosis uptake rate
S
diff− conv
: scale factors for antibody-specific paracellular transport rate
3.3.2 Antibody-specificscalefactorsforpinocytosis( S
pino
)anddiffusion-convection
(S
diff− conv
)
Using the plasma concentration-time data for each of the 22 mAbs resulted in estimates of the two
antibody-specific model parameters S
pino
and S
diff− conv
listed in Table 3.4 for each antibody. The
estimatesofS
pino
rangedfrom0.614to1.454andthoseofS
diff− conv
between0.475to1.24. Thetwo
parameters were estimated with good precision (RSE% < 20%) as indicated in Table 3.4. Figure
3.3 provides plots of the fitted concentration-time profiles, demonstrating good agreement between
observed plasma concentrations and model predictions (R > 0.94 for all antibodies), indicating
that the PBPK model can describe the plasma concentration time courses observed following the
administration of these antibodies.
37
Figure 3.3: Model predicted and observed antibody plasma concentration versus time profiles fol-
lowing intravenous delivery. The solid lines are model predictions. The symbols represent digitized
data at different dose levels from literature (Table 3.1).
38
3.3.3 Biophysical determinants of antibody-specific model parameters
The results of the linear regression analyses relating the estimated values of S
pino
for each antibody
to each of the in vitro physiochemical properties individually found that a significant relationship
wasobtainedwithonlyheparinrelativeretentiontime(Hep RT)asshowninFigure3.4a(R=0.64,
p = 0.0013). The p values for the other in vitro assays evaluated individually ranged from 0.093
to 0.93 (see Supplemental Material). Similarly, S
diff− conv
was found to be significantly associated
onlywithheparinrelativeretentiontime(Figure3.4b,R=0.52,p=0.012),butnotwithanyofthe
otherphysiochemicalproperties(individualprangesfrom0.13to0.82, seeSupplementalMaterial).
Multiple regression relations were explored via step-wise regression, but no significant associations
were identified with the addition of other physiochemical properties. For example, when “Slope
for Accelerated Stability” was added to Hep RT to describe the values of S
pino
obtained from the
22 antibodies, its regression coefficient was 0.99, with a standard error of 0.59 (p = 0.11). The
addition of “HEK titer” in the regression to predict S
diff− conv
yielded its regression coefficient as
-0.00067, with standard error of 0.00074 (p = 0.37).
WenotethatHep RTalsoshowsasignificantbutweakerpositiveassociationwithseveralassays
for non-specific interaction (e.g., PSR, R = 0.47, p < 10
− 3
; enzyme-linked immunosorbent assay
(ELISA),R=0.40, p< 10
− 3
), aswell asforself-association(AC-SINS,R=0.37, p < 10
− 3
,) based
on the values reported [64], indicating the concurrence between these in vitro assays.
3.3.4 Model-based prediction of antibody PK
Using the identified regression relations between S
pino
, S
diff− conv
and Hep RT (Figure 3.4), each
antibody’s Hep RT, hFcRn RT and molecular weight (MW) were used with the PBPK model
to perform a population simulation of the plasma concentration for each of the 22 antibodies.
Figure 3.5 shows the predictions for four antibodies with the smallest (Olokizumab) and largest
(Fulranumab)Hep RTvalues, aswellastwowithvaluesofHep RTnearthenearthemeanheparin
relative retention times (Ozanezumab, Trastuzumab) of the 22 antibodies studied. The mean
(±standard deviation) of S
pino
of these four antibodies (Fulranumab, Olokizumab, Ozanezumab,
Trastuzumab) are 1.21 (±0.16), 0.69 (±0.18), 1.01 (±0.15), 1.03 (±0.15), respectively. Also, the
39
Figure3.4: Associationbetweenthetwoestimatedmodelparametersandantibodyheparinrelative
retention time (Hep RT). Estimated regression lines: (a) S
pino
= 0.75· Hep RT +0.53 (R = 0.64,
p = 0.0013, residual standard error = 0.148). (b) S
diff− conv
= 0.73· Hep RT +0.35 (R = 0.52, p
= 0.012, residual standard error = 0.197). Shaded areas represent 95% confidence region for the
line of means
corresponding mean (±standard deviation) of S
diff− conv
of these four antibodies are 1.01 (±0.20),
0.50 (±0.20), 0.82 (±0.20) and 0.84 (±0.20), respectively. As indicted in the figure, the 5th-95th
percentile ranges of the model-based PK predictions include the observed plasma concentrations of
these antibodies.
3.3.5 Internal validation of model parameter-assay correlation
The results of an internal validation using the heparin relative retention times of randomly selected
subsets of 18 mAbs to construct new S
pino
and S
diff− conv
regression models are as follows. Of
the 100 regression relations, 94 resulted in a significant association (p < 0.05) between S
pino
and
Hep RT, while 74 were found to yield a significant association between S
diff− conv
and Hep RT.
These results suggests that our finding of a significant association between S
pino
and Hep RT and
between S
diff− conv
and Hep RT is likely independent of the specific antibodies within the subset
of the 22 antibodies tested.
40
Figure 3.5: Model-predicted and observed plasma concentrations of four antibodies: (a) Ful-
ranumab (Hep RT = 0.90), (b) Olokizumab (Hep RT = 0.21), (c) Ozanezumab (Hep RT = 0.64)
and (d) Trastuzumab (Hep RT = 0.66) for the indicated doses. Dots are mean concentrations from
literature(Table3.1). Solidlinesindicatethepredictedmedianconcentrationsfromthepopulation
simulations. Dash lines display the 5th and 95th percentiles of the predicted plasma concentrations
from the population simulations
41
3.4 Discussion
Previous studies have established that stronger non-specific antibody binding leads to increased
pinocytotic uptake, resulting in greater lysosomal degradation of mAbs [34, 62]. Consistent with
these findings, we found a significant positive correlation (R = 0.64, p = 0.0013) between the
model parameter representing antibody-specific vascular to endothelial clearance ( S
pino
) and hep-
arin relative retention time (Hep RT), an in vitro measure of the non-specific binding. We also
found that antibody-specific differences in paracellular transport due to convection and diffusion
(S
diff− conv
)couldbepartiallyexplainedbyHep RT(R=0.52, p=0.012). Paracellularconvection
is determined by the vascular to tissue hydrostatic gradient and the sieving effect, with the latter
dependent on pore physiology, as well as the size, shape and charge shape of antibodies [10]. This
observed positive correlation, could be a consequence of the longer heparin retention time that
results from the stronger binding of the more positively charged antibodies to the negative charges
of the heparin column.
The results from 12 physiochemical property assays for non-specific interaction, thermal sta-
bility, hydrophobicity and antibody self-association reported by Jain and others [64] for each of
the 22 antibodies examined in our study were considered (see Supplemental Material), both indi-
vidually and in combination, for their ability to predict the model-based estimates of S
pino
and
S
diff− conv
. These included assays related to cross-interaction (e.g., PSR, ELISA, BVP, CIC) and
self-association (AC-SINS), which were reported [64] to be predictive of clearance. None of these
12 assays, however, was found to be a significant predictor of either S
pino
or S
diff− conv
for the 22
antibodies we investigated. For example, although PSR had the strongest association with S
pino
,
among the 12 assays, it was not significant (R = 0.37, p = 0.095). Similarly, AC-SINS was not
found to be associated with S
pino
(R = 0.101, p = 0.63) or with S
diff− conv
(R = 0.19, p = 0.40).
Interestingly, several of these assays were positively correlated with Hep RT (e.g., PSR, R = 0.47,
p < 10
− 3
; ELISA, R = 0.40, p < 10
− 3
; AC-SINS, R = 0.37, p < 10
− 3
), but unlike Hep RT their
association with the antibody-specific model parameters did not rise to the level of significance.
ELISA and other assays measure direct interactions for heterogenous surrogates, focusing on var-
ious non-covalent interactions with different degrees of involvement of charge and hydrophobicity
42
interactions that do not appear to have the sensitivity to adequately describe cellular events. On
the other hand, heparin chromatography isolates heparin as a homogenous surrogate for relevant
cell surface-based component interaction, thus facilitating increased sensitivity to be connected to
parameters such as S
pino
, which are also focused on cellular events. In another study, Avery and
others used DNA- and insulin-binding assays to characterize non-specific interactions and also con-
sidered FcRn interaction using SPR and column chromatography assays [68]. However, since only
four of the antibodies in our study were investigated by Avery and others, a predictive statistical
analysis is not warranted.
Structure-based in silico metrics have previously been considered as predictors of PK during
antibodyscreening. Accordingly,wealsoassessed23in silico metricsfromtheMolecularOperating
Environment (MOE) [66] and five from the Therapeutic Antibody Profiling (TAP) platform [65]
for their potential association with the estimates of the model-based pinocytosis and diffusion-
convection parameters for the 22 antibodies in our study (see Supplemental Material). Only one
in silico measure related to hydrophobicity was found to be a significant predictor of S
pino
(R =
0.74, p = 0.0053) for 12 of the antibodies in common between the two sets. In a further analysis of
those antibodies reported with known Hep RT values and in silico measures from MOE and TAP
platforms (n = 59), we found that Hep RT was positively associated with several charge-related
and hydrophobicity-related metrics. In particular, Hep RT was found to be associated with the
charge-related metric pI 3D in combination with the hydrophobicity descriptor asa hph (R = 0.82,
p < 10
− 13
). This suggests that Hep RT may represent the combined effects of electrostatic and
hydrophobic factors, as fundamental determinants of antibody-specificity [104]. We note that a
comprehensive investigation of various in vitro and in silico properties of 64 antibodies identified
charge and hydrophobicity as important predictors of non-specific antibody clearance based on
statistical and machine learning analyses [72].
Physiologically based pharmacokinetics models have been used to predict disposition of mAbs
in humans based on antibody physiochemical properties assessed via in vitro assays and in sil-
ico methods. In a study of 12 antibodies, an in vitro self-association score (AC-SINS) was used
to predict antibody-specific binding affinity with cell membrane within a PBPK model [73]. In
43
our modeling analysis, however, we did not find a significant relationship between the values of
AC-SINS reported by Jain and others [64] and the estimated pinocytosis scaling parameter (S
pino
)
values for the 22 antibodies considered (R = 0.109, p = 0.63). Since self-association interactions
aremechanisticallymoredirectlyrelatedtothesolubility, viscosityandopalescenceoftheantibody
formulation than to non-specific antibody interactions that occur in vivo, it may be more difficult
to identify any potential associations between AC-SINS and PBPK model-based parameters char-
acterizing non-specific interactions [105]. Further PBPK studies using additional antibodies would
be needed to explore AC-SINS as a model-based predictor of antibody PK in humans.
In our previous PBPK modeling work, a sequence-based in silico charge metric (patches of
positive charge in the complementarity-determining region, PPC) was found to be positively asso-
ciated with S
pino
in 16 mAbs (R = 0.73, p = 0.0013) [74]. The positive relation between S
pino
and
the PPC metric is theoretically reasonable and consistent with other in vitro and in vivo studies
[34, 106]. However, we found no significant relationship between S
pino
and the PPC metric in the
larger set of 22 mAbs evaluated in this work (R = 0.24, p > 0.2). Pooling the distinct antibodies
from our current and previous work (n = 28), a borderline but statistically significant association
was found between S
pino
and PPC (R = 0.44, p = 0.02). Taken together, these results suggest
local charge in the complementarity-determining region may inform the model-based estimate of
antibody pinocytotic transport, but other antibody properties not incorporated in the modeling
analysis may make it difficult to detect any contribution of this charged-based property [106].
Although we found statistically significant associations between both the PBPK model param-
eters representing antibody non-specific binding and that representing paracellular transport with
an in vitro heparin chromatography assay, the resulting regression relationships involved notable
variability as indicated by the regression line 95% confidence regions in Figure 3.4. Given that the
modeling analysis was based on mean data reported for each of the 22 antibodies considered, we
could not include subject specific factors that could account for some of the unexplained variability
reflected in the regression relationships. Moreover, differences in subject disease states and FcRn
expression, as well as antibody-specific differences in physical sizes and glycosylation patterns were
not investigated as additional explanatory predictors of S
pino
andS
diff− conv
[5]. It should be noted
44
that the antibodies selected have been approved or were evaluated in Phase 2 or Phase 3 clinical
trials, and have thus have been selected for their desirable PK properties. This de facto selection
bias yields a set of antibodies with smaller differences in physiochemical properties, as measured
by the in vitro assays, making it more challenging to quantify any relationships between model-
based antibody transport processes and in vitro readouts. An additional limitation of the work is
the relatively small number of antibodies considered, and further investigation that includes more
antibodies with available plasma and tissue concentration time course data is required to assess
the suitability of the proposed model-based approach for predicting antibody plasma and tissue
PK in humans. Also, the Hep RT values of the 22 mAbs were adopted from Kraft and others
[62], in which the variable domains of different antibodies where grafted onto a common IgG1 Fc.
AlthoughsixoftheantibodiesincludedinthisstudywhereIgG2andIgG4isotypes, thecorrelation
betweenandHep RTandandHep RTremainedsignificantevenafterexcludingthesesixnon-IgG1
mAbs (results not shown). Finally, the model represents antibody-specific differences in transport
as linear scale factors, which may not represent these antibody-dependent processes with sufficient
fidelity to reflect the full extent of any underlying differences in antibody disposition.
Incontrasttootherstudiesthathaveusedempiricallyderivedexpressionstorelate invitro mea-
sures of antibody physiochemical properties directly to antibody clearance, we propose a PBPK
model-based approach for predicting mAb PK that incorporates the mechanisms that govern anti-
body transport and processing, which are in turn informed by in vitro measures of antibody phys-
iochemical properties. The underlying premise of this work is that the fundamental mechanisms
responsible for antibody transport (paracellular exchange, non-specific binding, FcRn interaction,
transcytosis) are more directly relatable to the in vitro assays designed to characterize these inter-
actions and processes than is overall systemic clearance. Given its physiological basis, the proposed
PBPK model can be expanded to include more descriptive representations of each of the antibody
processing subsystems such as TMDD in extracellular spaces, as well as other antibody-specific
information.
The supplementary material of this chapter is available on:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9067474/
45
4 Predictingmonoclonalantibodyphar-
macokineticsfollowingsubcutaneous
administration via whole-body
physiologically-based modeling
S. Hu and D. Z. D’Argenio, “Predicting monoclonal antibody pharmacokinetics following subcuta-
neousadministrationviawhole-bodyphysiologically-basedmodeling,”JournalofPharmacokinetics
and Pharmacodynamics, vol. 47, no. 5, pp. 385–409, 2020.
46
Abstract
Use of the subcutaneous (SC) route for administering monoclonal antibodies (mAbs) to treat
chronic conditions has been hindered because of an incomplete understanding of fundamental
mechanisms controlling mAb absorption from the SC site, and due to the limited translata-
bility of preclinical studies. In this paper, we report on the development and evaluation of
a whole-body physiologically-based model to predict mAb pharmacokinetics following SC ad-
ministration. The circulatory model is based on the physiological processes governing mAb
transport and includes two mAb-specific parameters representing differences in pinocytosis rate
and the diffusive/convective transport rates among mAbs. At the SC administration site, two
additional parameters are used to represent mAb differences in lymphatic capillary uptake and
in pre-systemic clearance. Model development employed clinical intravenous (IV) plasma PK
data from 20 mAbs and SC plasma PK data from 12 of these mAbs, as obtained from the litera-
ture. TheresultingmodelreliablydescribedboththeIVandSCmeasuredplasmaconcentration
data. In addition, a metric based on the positive charge across the mAb’s complementarity de-
termining region vicinity was found to positively correlate with the model-based estimates of
the mAb-specific parameter governing organ/tissue pinocytosis transport and with estimates of
the mAb’s SC lymphatic capillary clearance. These two relationships were incorporated into
the model and accurately predicted the SC PK profiles of three out of four separate mAbs not
included in model development. The whole-body physiologically-based model reported herein,
provides a platform to characterize and predict the plasma disposition of monoclonal antibodies
following SC administration in humans.
Keywords Monoclonal antibody; Clinical pharmacokinetics; Physiologically-based pharma-
cokinetics (PBPK); Subcutaneous; Bioavailability
47
4.1 Introduction
Monoclonal antibodies (mAbs) are used in treating cancers, auto-immune, inflammatory and other
diseases [107]. Between 2008 and 2019, the U.S. Food and Drug Administration (FDA) approved
approximately 60 mAb-based treatments, 25 of which have involved subcutaneous (SC) delivery
[17, 108]. The European Medicines Agency has approved 30 mAb products (including biosimilars)
for SC administration since 2010 [109]. While SC delivery of mAbs for treating chronic diseases
has significant advantages compared to intravenous (IV) administration [17, 110], it has been chal-
lenging to predict the variable absorption and delivery of mAbs following SC administration from
preclinical experimental or modeling studies [16, 29].
For over 30 years, physiologically-based pharmacokinetic (PBPK) models have been applied to
explore the disposition of mAbs for different applications, including inter-species scaling, guiding
antibody engineering and predicting drug-drug interactions [13, 111]. In 1986, Covell et al. first
applied a PBPK model to study the pharmacokinetics (PK) of an immunoglobulin G1 (IgG1) and
antibody fragments in mice [53]. Baxter et al. scaled physiological parameters of mice to humans
to predict clinical PK profiles of a mAb [49]. The influence of neonatal Fc receptor (FcRn) on
mAb disposition was incorporated in a PBPK model by Ferl et al. [18] and by Garg and Balthasar
[50]. In 2012, Shah and Betts reported a PBPK model platform for different species [52]. More
recently, Glassman and Balthasar reported a comprehensive whole-body PBPK model to make
priori predictions of the clinical PK of mAbs exhibiting target-mediated disposition [51], and Li
andBalthasarappliedthisPBPKmodelframeworktoevaluateanti-FcRntherapyeffectsinhumans
[102].
Whole-body PBPK models and hybrid compartmental-PBPK models have been developed to
describe and predict the absorption and SC PK of therapeutic proteins, including mAbs. In 2013,
Zhao andcoworkers coupled a compartment disposition model witha physiologically-based absorp-
tion model to quantify absorption process of mAbs after SC or intramuscular delivery [58]. Gill
et al. developed a whole-body PBPK model for various protein drugs in humans [60], which rep-
resents SC tissues using model compartments that are linked to the circulatory system via blood
48
flow and lymph flow anatomically. This model, however, does not account for FcRn-mediated recy-
cling processes in the organ compartments, and requires previously determined values of the mAbs’
bioavailability. Offmanetal. incorporatedthelymphaticuptakepathwayintoawhole-bodyPBPK
modelforapeptidegivensubcutaneously[112]. Morerecently,Varkhedeetal. developedaminimal
PBPK model, composed of a two-compartment PK model and an expanded physiologically-based
model of lymphatic system [113]. They also proposed that the isoelectric point (pI) of the mAb
may be a predictor of its clearance and bioavailability following SC administration.
The above-mentioned models for SC absorption of mAbs have several limitations. The hybrid
models([58]and[113]),becauseoftheircompartmentalcomponents,arelesssuitableforpreclinical
to clinical translation, nor do they allow predictions of drug concentrations at sites of action. The
previously published whole-body PBPK models ([60] and [112]) do not account for some important
physiological processes involved with mAb transport, do not predict antibody SC bioavailability,
and do not incorporate biophysical properties of the mAb to predict SC absorption and plasma
PK.
In this study, we modified and extended the previous PBPK for IV administrated mAbs re-
ported [51] to include a mechanistic representation of the pre-systemic degradation, trafficking and
absorption of mAbs at the SC injection site. The model also incorporates a metric based on the
local charge of the complementarity-determining regions of the mAb to explain differences among
mAbs in organ transport processes, as well as degradation at the SC site. For each organ, the
model represents transcapillary exchange using a two-pore transport theory and includes the se-
quential transit and processing of the mAb in the endosomal space of endothelial cells. The model
for the central lymphatic system includes lumped lymph node and lymphatic vessel compartments,
where the latter incorporates endothelial cell mAb processing. A peripheral sampling site is used
to better reflect measured concentration in clinical studies. For the SC site model, both lymphatic
capillary uptake and degradation processes are included. The mAb-specific model parameters were
estimated using the whole body PBPK model and clinical intravenous (IV) plasma PK data from
20 mAbs and SC plasma PK data from 12 of these mAbs, as obtained from the literature. The
49
model is used to predict bioavailability following SC administration, investigate the relative path-
way specific contributions to systemic delivery from SC administration, and to examine, through
sensitivity analysis, the role of key properties involved in SC absorption.
4.2 Materials and methods
4.2.1 Clinical study data
Mean plasma concentrations following IV administration of the 20 mAbs listed in Table 4.1 and
SC administration of 12 of these mAbs listed in Table 4.2 were digitized from figures reported in
the cited publications. The molecular weights of these mAbs range from 1.42 x 10
5
to 1.50 x 10
5
g/mol. Data from most of the mAbs included results at multiple dose levels (see Tables 1 and 2).
Dose normalization of PK data indicated no obvious non-linearity caused by target-mediated drug
disposition. Mean plasma concentrations following the SC administration of another four mAbs
(Omalizumab, Tildrakizumab, Ixekizumab, Lanadelumab) were also digitized from publications
and used for further model evaluation as described below.
Table 4.1: mAbs with plasma PK following IV administration. Also shown are the estimated
drug-specific parameters ( S
pino
and S
diff− conv
) determined using IV PK data for each mAb
mAb (abbreviation) Molecular weight (g/mol x 10
5
) Dosing Ref. S
pino
(RSE%) S
diff− conv
(RSE%)
Actoxumab (Act) 1.459 1-20 mg/kg [114] 0.627 (5.2) 1.04 (5.7)
Adalimumab (Ada) 1.442 0.25-5 mg/kg [75, 115] 1.19 (3.4) 1.24 (5.3)
Belimumab (Bel) 1.47 240mg [116] 1.21 (2.6) 1.00 (10)
Canakinumab (Can) 1.452 1-3mg/kg; 600mg [80] 0.868 (1.2) 1.15 (4.9)
Daclizumab (Dac) 1.426 200 mg; 400 mg [81] 1.20 (1.3) 1.46 (6.6)
Enokizumab (Eno) 1.484 0.3-9 mg/kg [83] 0.839 (2.7) 1.05 (4.7)
Figitumumab (Fig) 1.46 10 mg/kg; 20mg/kg [85] 0.897 (2.1) 1.16 (7.6)
Gevokizumab (Gev) 1.452 0.01-3 mg/kg [87] 0.833 (2.8) 0.823 (4.9)
GNbAC1 (GNb) 1.47 2 mg/kg; 6mg/kg [117] 0.849 (11) 1.27 (9.6)
Golimumab (Gol) 1.469 100mg [118] 2.14 (2.0) 1.49 (7.6)
Guselkumab (Gus) 1.436 0.03-10 mg/kg [88] 1.43 (1.2) 1.89 (3.3)
Infliximab (Inf) 1.491 5 mg/kg [119] 1.41 (3.4) 1.34 (9.7)
Mepolizumab (Mep) 1.492 250 mg [89] 0.963 (2.4) 0.452 (13)
PAmAb (Pam) 1.5 1-40 mg/kg [120] 1.13 (1.5) 0.640 (6.5)
Risankizumab (Ris) 1.456 200, 600, 1200mg [121] 0.783 (3.2) 1.41 (5.9)
Siltuximab (Sil) 1.45 1-12 mg/kg [122] 0.872 (11) 2.21 (5.6)
Tefibazumab (Tef) 1.476 10-20 mg/kg [123] 1.23 (2.9) 1.62 (5.9)
Tralokinumab (Tral) 1.439 150 mg [94] 0.936 (4.0) 0.630 (13)
Trastuzumab (Tras) 1.48 6mg/kg [95] 1.16 (11) 0.822 (15)
Urtoxazumab (Urt) 1.446 0.1-3 mg/kg [124] 0.742 (4.0) 0.981 (6.4)
50
Table4.2: mAbswithplasmaPKfollowingSCadministration. Estimateddrug-specificparameters
(CL
SC
LymCap
and S
SC
LymUpt
) in SC tissue site model using SC PK data are also shown for each mAb
mAb (abbreviation) Dosing Ref. CL
SC
LymCap
(ml/day) (RSE%) S
SC
LymUpt
(RSE%)
Adalimumab (Ada) 40 mg [125] 0.957 (9.5) 0.166 (4.6)
Belimumab (Bel) 200, 240 mg [116] 0.501 (15) 0.204 (3.9)
Canakinumab (Can) 150, 300 mg [80] 0.798 (11) 0.220 (5.8)
Daclizumab (Dac) 50, 150, 300 mg [81] 0.469 (17) 0.0904 (3.7)
Enokizumab (Eno) 1, 3, 9 mg/kg [83] 0.441 (20) 0.301 (5.0)
Gevokizumab (Gev) 0.03, 0.3 mg/kg [87] 0.732 (22) 0.130 (7.5)
Golimumab (Gol) 100 mg [118] 2.72 (8.1) 0.193 (7.5)
Guselkumab (Gus) 10, 30, 100, 300 mg [88] 3.66 (5.0) 0.283 (5.1)
Infliximab (Inf) 120 mg [126] 0.950 (15) 0.262 (8.6)
Mepolizumab (Mep) 250 mg [89] 0.700 (14) 0.272 (5.9)
Risankizumab (Ris) 18, 90, 300mg [121] 1.38 (9.0) 0.217 (5.6)
Tralokinumab (Tral) 150, 300 mg [94] 1.69 (7.4) 0.186 (5.2)
4.2.2 Model structure
The model builds upon that reported by Glassman and Balthasar to describe the PK of mAbs
after IV administration [51]. A subcutaneous administration site and a central lymphatic system
model were added to the model reported in [51], as depicted in Figure 4.1. In addition, the model
incorporates a two-pore formalism to describe vascular to interstitial exchange in each tissue as
proposed by Li and Shah [19], along with a distinct peripheral blood sampling site.
4.2.2.1 Subcutaneous administration site
The model for the subcutaneous administration site is linked to the whole-body circulatory system
through the SC administration site plasma flow ( Q
SC
) and lymph flow ( L
SC
). The subcutaneous
tissue model is composed of the following spaces: plasma, endothelial, interstitial, lymphatic, and
cell (see Figure 4.2a). Following SC administration, the mAb distributes within the extracellular
matrix (ECM) of SC tissue and exchanges with the blood capillaries and is transported to the local
lymphaticcapillariesbeforedrainingintothesystemicvenousplasmaandcentrallymphaticsystem
[127].
The predominant processes governing transport of IgG between the vascular and interstitial
spaces are convection, diffusion and pinocytosis [10]. The work reported in [19] concluded that the
transcapillarytransportoftherapeuticproteins(13to150kDa)viaconvectionanddiffusioncanbe
51
Figure 4.1: Schematic of the structure of the developed whole-body PBPK model for mAbs after
IV or SC administration
52
Figure 4.2: Schematic of the organ-level model. (a) of SC tissue, including vascular space, endothe-
lial space (with endosomal sub-compartments), interstitial space and cells space. The SC dose was
administrated into the interstitial space, from where mAbs can be drained into local lymphatic
capillaries to enter the lymphatic system or be absorbed via blood perfusion. (b) of endosomal
sub-compartments, representing the interaction of FcRn and IgG at a certain pH value (x). (c) of
the lymphatic vessel, which is divided into lymph flow and lymphatic endothelial space
53
described using a two-pore model. Based on the model and results presented in [19], the transport
of 142-150 kDa mAbs could be described using a single (large) pore model. For generality and to
allow for future extension, however, we have incorporated the two-pore model in this work (see
Figure 4.2a).
ExchangeofIgGbetweenthevascularandinterstitialspacesalsooccursviavascularendothelial
pinocytosis followed by pH-dependent, FcRn-mediated endosomal processing [23]. To incorporate
thesemechanisms, wefollowedtheapproachin[51], whichincludedseparate, sequentialpH-specific
endosomal sub-compartments to represent this continuous endosomal processing. Based on the
report in [128], we included four endosomal sub-compartments to represent early endosomal pro-
cessing and sorting, each with a distinct pH value ranging from 7.4 to 6.0 (Figure 4.2a). From the
fourth sub-compartment, IgG is sorted into either a recycling endosome pool from which IgG is
transportedtothevascularandinterstitialspaces,orintolysosomepoolforproteolyticdegradation
(Figure 4.2a).
Given these assumptions, and following Figs. 2a, the following equations can be written to
describe the SC vascular space concentrations of endogenous IgG (CIgG
SC
vasc
) and administrated
mAb (CmAb
SC
vasc
):
V
SC
vasc
· dCIgG
SC
vasc
dt
=Q
SC
· CIgG
Lung
vasc
− (Q
SC
− L
SC
)· CIgG
SC
vasc
− CL
SC
pino
· CIgG
SC
vasc
+fIgG
recyc
· CL
SC
pino
· CIgFc
SC
recyc
− PS
SC
L
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
L
e
Pe
L
− 1
− PS
SC
S
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
S
e
Pe
S
− 1
− J
SC
L
· (1− σ L
)· CIgG
SC
vasc
− J
SC
S
· (1− σ S
)· CIgG
SC
vasc
(4.1)
V
SC
vasc
· dCmAb
SC
vasc
dt
=Q
SC
· CmAb
Lung
vasc
− (Q
SC
− L
SC
)· CmAb
SC
vasc
− S
pino
· CL
SC
pino
· CmAb
SC
vasc
+S
pino
· fIgG
recyc
· CL
SC
pino
· CmAbFc
SC
recyc
− S
diff− conv
· PS
SC
L
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
L
e
PeL
− 1
− S
diff− conv
· PS
SC
S
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
S
e
PeS
− 1
− S
diff− conv
· J
SC
L
· (1− σ L
)· CmAb
SC
vasc
− S
diff− conv
· J
SC
S
· (1− σ S
)· CmAb
SC
vasc
(4.2)
The symbols for the model parameters and other concentration variables are defined in Table 4.3.
The first lines of Eqs. 4.1 and 4.2 represent vascular perfusion of SC space, while the second
lines reflect the uptake via pinocytosis and recycling pathway through vascular endothelial cells.
The third lines in each equation represent the transport of IgG and mAb via diffusion through
54
large and small pores driven by concentration differences CIgG
SC
vasc
− CIgG
SC
inter
and CmAb
SC
vasc
− CmAb
SC
inter
, while the fourth lines describe convection through large and small pores. The values of
the permeability surface area products (PS
SC
L
and PS
SC
S
), Peclet numbers (Pe
L
and Pe
S
), lymph
flow ( J
SC
L
and J
SC
S
) and reflection coefficients ( σ L
and σ S
) depend on mAb molecular weight as
detailed below. Additional differences in exchange of mAbs relative to IgG were represented in Eq.
4.2 by scaling factors applied to pinocytosis (S
pino
) and to the large and small pore diffusion and
convention processes (S
diff− conv
), following the approach in the IV PBPK model of [51]. These
two drug-specific scaling factors were estimated for each mAb as described below.
In the endothelial space within each sub-compartment, the pH-dependent mass action between
IgGandFcRnismodeledassumingpHspecificassociationanddissociationrateconstants( k
pHx
on
and
k
pHx
off
) as introduced in [51] (see Figure 4.2b). In the endothelial sub-compartment for pH=7.4, the
concentrations of CIgG
SC
7.4
, CIgFc
SC
7.4
, CmAb
SC
7.4
and CmAbFc
SC
7.4
can be described by the following
equations:
dCIgG
SC
7.4
dt
=(CL
SC
pino
· CIgG
SC
vasc
+CL
SC
pino
· CIgG
SC
inter
)/V
SC
endo,sub
+k
7.4
off
· CIgFc
SC
7.4
− k
7.4
on
· CIgG
SC
7.4
· CFcRn
SC
7.4
− 1
τ · CIgG
SC
7.4
(4.3)
dCIgFc
SC
7.4
dt
=k
7.4
on
· CIgG
SC
7.4
· CFcRn
SC
7.4
− k
7.4
off
· CIgFc
SC
7.4
− 1
τ · CIgFc
SC
7.4
(4.4)
dCmAb
SC
7.4
dt
=(S
pino
· CL
SC
pino
· CmAb
SC
vasc
+S
pino
· CL
SC
pino
· CmAb
SC
inter
)/V
SC
endo,sub
+k
7.4
off
· CmAbFc
SC
7.4
− k
7.4
on
· CmAb
SC
7.4
· CFcRn
SC
7.4
− 1
τ · CmAb
SC
7.4
(4.5)
dCmAbFc
SC
7.4
dt
=k
7.4
on
· CmAb
SC
7.4
· CFcRn
SC
7.4
− k
7.4
off
· CmAbFc
SC
7.4
− 1
τ · CmAbFc
SC
7.4
(4.6)
The symbols for model parameters and concentration variables are defined in Table 4.3. The con-
centrationoftheFcRnreceptorinthisfirstendosomalsub-compartment( CFcRn
SC
pH7.4
)isassumed
tobeconstantat3.3x10
4
nM[51]. ThefirstlinesofEqs. 4.3and4.5representuptakeviapinocyto-
sis of free IgG or free mAb from the SC site vascular and interstitial spaces, and the first two terms
of the second lines of these equations reflect the interaction between FcRn and IgG. The second
55
Table 4.3: Definitions of parameters and variables used in model equations
Parameter/variable Unit Definition
t day time
V
Plasma
ml Volume of central plasma compartment
CIgG
Plasma
nM Concentration of endogenous IgG in central plasma compartment
CmAb
Plasma
nM Concentration of administrated mAb in central plasma compartment
IgG0 nmol/day Production rate of endogenous IgG
a
V
Organ
vasc ml Volume of vascular space
V
Organ
inter
ml Volume of interstitial space
V
Organ
endo
ml Volume of endosomal space
V
Organ
endo,sub
ml Volume of endosomal sub-compartment
Q
Organ
ml/day Plasma flow
L
Organ
ml/day Lymph flow
CIgG
Organ
vasc nM Concentration of endogenous IgG in vascular space
CmAb
Organ
vasc nM Concentration of administrated mAb in vascular space
CIgG
Organ
inter
nM Concentration of endogenous IgG in interstitial space
CmAb
Organ
inter
nM Concentration of administrated mAb in interstitial space
CIgG
Organ
pHx
nM Concentration of free endogenous IgG in endosomal sub-compartment with pH = x of vascular endothelium
CmAb
Organ
pHx
nM Concentration of free administrated mAb in endosomal sub-compartment with pH = x of vascular endothelium
CIgFc
Organ
pHx
nM Concentration of endogenous IgG-FcRn complex in endosomal sub-compartment with pH = x of vascular endothelium
CmAbFc
Organ
pHx
nM Concentration of administrated mAb-FcRn complex in endosomal sub-compartment with pH = x of vascular endothelium
CIgG
Organ
recyc nM Concentration of endogenous free IgG in recycling endosome of vascular endothelium
CmAb
Organ
recyc nM Concentration of endogenous free administrated mAb in recycling endosome of vascular endothelium
CIgFc
Organ
recyc nM Concentration of endogenous IgG-FcRn complex in recycling endosome of vascular endothelium
CmAbFc
Organ
recyc nM Concentration of administrated mAb-FcRn complex in recycling endosome of vascular endothelium
CFcRn
Organ
pHx
nM Concentration of FcRn in endosomal sub-compartment with pH = x of vascular endothelium
fIgGrecyc - IgG recycling fraction of IgG-FcRn complex back to vascular space
τ day IgG transit time through endosomal sub-compartments
k
pHx
on nM
− 1
day
− 1
Association rate constant between IgG and FcRn at pH = x
k
pHx
off
day
− 1
Dissociation rate constant between IgG and FcRn at pH = x
V
LymCap
ml Volume of local lymphatic capillaries
k
LymNode
day
− 1
Rate constant of mAb leaving lymph node
V
LymVes
ml Volume of lymphatic vessel compartment
Q
LymVes
ml/day Lymph flow rate from lymphatic vessel compartment
AIgG
LymNode
pmol Amount of endogenous IgG in lymph node compartment
AmAb
LymNode
pmol Amount of administrated mAb in lymph node compartment
CIgG
LymVes
nM Concentration of endogenous IgG in lymphatic vessel compartment
CmAb
LymVes
nM Concentration of administrated mAb in lymphatic vessel compartment
CIgG
LymCap
nM Concentration of endogenous IgG in local lymphatic capillary compartment
CmAb
LymCap
nM Concentration of administrated mAb in local lymphatic capillary compartment
σ - Lymph reflection coefficient
CL
Organ
pino
ml/day Pinocytosis rate of vascular endothelium
IVmAb(t) pmol/day Intravenous infusion rate of monoclonal antibodies
Dose pmol Subcutaneous dose of monoclonal antibodies
Spino - mAb specific pinocytosis transport rate scale factor
S
diff− conv
- mAb specific diffusion and convective transport scale factor
CL
SC
LymCap
ml/day SC site lymphatic capillary mAb clearance
S
SC
LymUpt
- SC site interstitial to lymphatic capillary mAb uptake scale factor
σ S
- Small pore vascular reflection coefficient
σ L
- Large pore vascular reflection coefficient
Pe
S
- Peclet number of small pores
Pe
L
- Peclet number of large pores
PS
Organ
S
ml/day Permeability-surface area product of small pore of an organ
PS
Organ
L
ml/day Permeability-surface area product of large of an organ
J
Organ
iso
ml/day Circular isogravimetric flow of an organ
J
Organ
S
ml/day Lymph flow through small pores of an organ
J
Organ
L
ml/day Lymph flow through large pores of an organ
a
The organs are listed in Table 4.4
56
lines of Eqs. 4.3 and 4.5 represent the transit of IgG into the next sub-compartment. In Eqs. 4.4
and 4.6, the first two terms represent the interaction of IgG and FcRn and mAb-FcRn, while the
last terms represent the transit of the IgG-FcRn and mAb-FcRn into the next sub-compartment.
Thefollowingequationsdescribedthecorrespondingconcentrationsforeachspeciesinendothe-
lial sub-compartment for pH=7.0:
dCIgG
SC
7.0
dt
=k
7.0
off
· CIgFc
SC
7.0
− k
7.0
on
· CIgG
SC
7.0
· CFcRn
SC
7.0
+
1
τ · (CIgG
SC
7.4
− CIgG
SC
7.0
)
(4.7)
dCIgFc
SC
7.0
dt
=k
7.0
on
· CIgG
SC
7.0
· CFcRn
SC
7.0
− k
7.0
off
· CIgFc
SC
7.0
+
1
τ · (CIgFc
SC
7.4
− CIgFc
SC
7.0
)
(4.8)
dCmAb
SC
7.0
dt
=k
7.0
off
· CmAbFc
SC
7.0
− k
7.0
on
· CmAb
SC
7.0
· CFcRn
SC
7.0
+
1
τ · (CmAb
SC
7.4
− CmAb
SC
7.0
)
(4.9)
dCmAbFc
SC
7.0
dt
=k
7.0
on
· CmAb
SC
7.0
· CFcRn
SC
7.0
− k
7.0
off
· CmAbFc
SC
7.0
+
1
τ · (CmAbFc
SC
7.4
− CmAbFc
SC
7.0
)
(4.10)
dCFcRn
SC
pH7.0
dt
=k
7.0
off
· CIgFc
SC
7.0
− k
7.0
on
· CIgG
SC
7.0
· CFcRn
SC
7.0
+k
7.0
off
· CmAbFc
SC
7.0
− k
7.0
on
· CmAb
SC
7.0
· CFcRn
SC
7.0
+
1
τ · (CFcRn
SC
pH7.4
− CFcRn
SC
pH7.0
)
(4.11)
The symbols for the concentration variables and model parameters are defined in Table 4.3. The
first two terms of Eqs. 4.7 to 4.10 reflect the interaction of FcRn and IgG at pH = 7.0. The
remaining terms in these equations represent the transit of IgG from the previous, to the next sub-
compartment. In Eq. 4.11, the first two lines represent the interaction of FcRn with endogenous
IgG and administrated mAb, and the third line reflects the transit through the FcRn pool. A
similar set of equations (not shown) can be written for the endothelial sub-compartment for pH =
6.5.
For the endothelial sub-compartment with pH=6.0, the following equations apply:
dCIgG
SC
6.0
dt
=k
6.0
off
· CIgFc
SC
6.0
− k
6.0
on
· CIgG
SC
6.0
· CFcRn
SC
6.0
+
1
τ · (CIgG
SC
6.5
− CIgG
SC
6.0
)
(4.12)
57
dCIgFc
SC
6.0
dt
=k
6.0
on
· CIgG
SC
6.0
· CFcRn
SC
6.0
− k
6.0
off
· CIgFc
SC
6.0
+
1
τ · (CIgFc
SC
6.5
− CIgFc
SC
6.0
)
(4.13)
dCmAb
SC
6.0
dt
=k
6.0
off
· CmAbFc
SC
6.0
− k
6.0
on
· CmAb
SC
6.0
· CFcRn
SC
6.0
+
1
τ · (CmAb
SC
6.5
− CmAb
SC
6.0
)
(4.14)
dCmAbFc
SC
6.0
dt
=k
6.0
on
· CmAb
SC
6.0
· CFcRn
SC
6.0
− k
6.0
off
· CmAbFc
SC
6.0
+
1
τ · (CmAbFc
SC
6.5
− CmAbFc
SC
6.0
)
(4.15)
dCFcRn
SC
pH6.0
dt
=k
6.0
off
· CIgFc
SC
6.0
− k
6.0
on
· CIgG
SC
6.0
· CFcRn
SC
6.0
+k
6.0
off
· CmAbFc
SC
6.0
− k
6.0
on
· CmAb
SC
6.0
· CFcRn
SC
6.0
+
1
τ · (CFcRn
SC
pH6.5
− CFcRn
SC
pH6.0
)
(4.16)
See Table 4.3 for the definition of the concentration variables parameter symbols. The last terms
of Eqs. 4.12 and 4.14 represent free IgG and free mAb transported into lysosome for degradation.
The last terms of Eqs. 4.13 and 4.15 represent the transport of FcRn-bound IgG and FcRn-bound
mAb into recycling endosome.
The following equations describe the concentrations of IgG-FcRn (CIgFc
SC
recyc
) and mAb-FcRn
(CmAbFc
SC
recyc
) in the recycling endosome pool:
dCIgFc
SC
recyc
dt
=
1
τ · CIgFc
SC
6.0
− CL
SC
pino
· CIgFc
SC
recyc
/V
SC
endo,sub
(4.17)
dCmAbFc
SC
recyc
dt
=
1
τ · CmAbFc
SC
6.0
− S
pino
· CL
SC
pino
· CmAbFc
SC
recyc
/V
SC
endo,sub
(4.18)
where the second terms in Eqs. 4.17 and 4.18 represent the recycling of IgG-FcRn and mAb-FcRn
into the vascular and interstitial spaces of the subcutaneous site.
From Figure 4.2a, the following equations can be written to describe the interstitial concentra-
tions of endogenous IgG (CIgG
SC
inter
) and administrated mAb (CmAb
SC
inter
):
V
SC
inter
dCIgG
SC
inter
dt
=(1− fIgG
recyc
)· CL
SC
pino
· CIgFc
SC
recyc
− CL
SC
pino
· CIgG
SC
inter
− L
SC
· CIgG
SC
inter
+PS
SC
L
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
L
e
Pe
L
− 1
+PS
SC
S
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
S
e
Pe
S
− 1
+J
SC
L
· (1− σ L
)· CIgG
SC
vasc
+J
SC
S
· (1− σ S
)· CIgG
SC
vasc
(4.19)
58
V
SC
inter
dCmAb
SC
inter
dt
=S
pino
· (1− fIgG
recyc
)· CL
SC
pino
· CmAbFc
SC
recyc
− S
pino
· CL
SC
pino
· CmAb
SC
inter
− S
SC
LymUpt
· L
SC
· CmAb
SC
inter
+S
diff− conv
· PS
SC
L
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
L
e
PeL
− 1
+S
diff− conv
· PS
SC
S
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
S
e
PeS
− 1
+S
diff− conv
· J
SC
L
· (1− σ L
)· CmAb
SC
vasc
+S
diff− conv
· J
SC
S
· (1− σ S
)· CmAb
SC
vasc
(4.20)
The symbols for the concentration variables and model parameters are defined in Table 4.3. The
initial concentration of mAb in SC interstitial space (CmAb
SC
inter
(0)) is Dose/V
SC
inter
. In the first
lines of these equations, the third term represents lymphatic transport of IgG or mAbs to the local
lymphatic capillaries. Nearly first-order removal of IgG from SC administration site after a short
stay was demonstrated in previous studies [129, 130], thus it is described using a constant SC site
lymph flow ( L
SC
). The differences in lymphatic uptake of mAbs compared to IgG is modeled using
a scaling factor (S
SC
LymUpt
) in Eq. 4.20. Other terms correspond to those in previous equations.
Compared to blood vessels, lymphatic capillaries possess larger gaps between endothelial cells,
have greater vessel diameters, and do not have well-defined basement membranes [131]. Thus,
lymphaticuptakeisbelievedtobethemajorabsorptionprocessforsubcutaneouslydeliveredmAbs,
consideringtheirlargemolecularweight[17]. DegradationofmAbsoccursattheSCsiteinterstitial
space, local lymphatic capillaries and lymphatic vessels, via proteolytic enzymes and/or soluble
and cell-membrane targets [23]. Without quantitative information on the relative contribution of
differentcatabolicpathwaysatthesesites[29],themodelcharacterizesmAbdegradationusinganet
clearanceterm(CL
SC
LymCap
)assignedtothelocallymphaticcapillarycompartment(seeFigure4.2a).
TheequationsbelowdescribethelocalSClymphaticcapillaryconcentrationsofIgG(CIgG
LymCap
)
and mAb (CmAb
LymCap
):
V
LymCap
dCIgG
LymCap
dt
=L
SC
· CIgG
SC
inter
− L
SC
· CIgG
LymCap
(4.21)
V
LymCap
dCmAb
LymCap
dt
=S
SC
LymUpt
· L
SC
· CmAb
SC
inter
− L
SC
· CmAb
LymCap
− CL
SC
LymCap
· CmAb
LymCap
(4.22)
See Table 4.3 for the definition of the concentration variables parameter symbols. In the Eqs. 4.21
and 4.22 , the first terms represent the lymph flow ( L
SC
) from the SC interstitial space and the
second terms are lymph drainage into the lymph node. The last term in Eq. 4.22 is the first-order
pre-systemic degradation of mAb with the clearance rate CL
SC
LymCap
. In contrast, the degradation
59
of endogenous IgG in SC site is assumed to be negligible given that it represents such a small
fraction of the overall degradation of IgG from all the other organs and tissues.
4.2.2.2 Central lymphatic system model
ThemodelusedtodescribethecentrallymphaticsystemisillustratedinFigure4.1. Thelymphflow
draining all organs and tissues is assumed to be collected in a lumped lymph node compartment.
A single lymphatic vessel compartment (Figure 4.2c) is used to represent all central lymphatic
vessels (e.g. lymphatic trunks, thoracic duct, etc). Based on the previous confirmation of the
existence of FcRn in lymphatic vessels [132, 133], FcRn-mediated protection of IgG in lymphatic
vessel endothelium was also included in the model. The following equations describe the amount of
IgG and mAb in the lymph node (Eqs. 4.23 and 4.24 ) and their concentrations in the lymphatic
vessel (Eqs. 4.25 and 4.26) compartments:
dAIgG
LymNode
dt
=
X
(1− σ )· L
Organ
· CIgG
Organ
inter
+L
SC
· CIgG
LymCap
− k
LymNode
· AIgG
LymNode
(4.23)
dAmAb
LymNode
dt
=
X
(1− σ )· L
Organ
· CmAb
Organ
inter
+L
SC
· CmAb
LymphCap
− k
LymNode
· AmAb
LymNode
(4.24)
V
LymVes
dCIgG
LymVes
dt
=k
LymNode
· AIgG
LymNode
− Q
LymVes
· CIgG
LymVes
− CL
LymVes
pino
· CIgG
LymVes
+CL
LymVes
pino
· CIgFc
LymVes
recyc
(4.25)
V
LymVes
dCmAb
LymVes
dt
=k
LymNode
· AmAb
LymNode
− Q
LymVes
· CmAb
LymVes
− S
pino
· CL
LymVes
pino
· CmAb
LymVes
+S
pino
· CL
LymVes
pino
· CmAbFc
LymVes
recyc
(4.26)
The symbols for the concentration variables and model parameters are defined in Table 4.3. The
lasttermsofEqs. 4.23and4.24representtransitofIgGormAbfromthelymphnodecompartment,
corresponding to the first terms of Eqs. 4.25 and 4.26. The second term of Eqs. 4.25 and 4.26
represent the transportation of IgG or mAb from lymphatic vessels. The equations for the transit
andFcRn-mediatedprocessingofIgGinendosomalsub-compartmentsfollowthosepresentedabove
for the SC endothelial space.
60
4.2.2.3 Central venous plasma
In the model, the central plasma compartment represents the systemic central venous pool, not
including the vascular spaces in the organ compartments (Figure 4.1) [134]. The endogenous IgG
concentration (CIgG
Plasma
) in and mAb concentrations (CmAb
Plasma
) in central venous plasma
can be written:
V
Plasma
dCIgG
Plasma
dt
=1000· IgG0+(Q
Liver
− L
Liver
)· CIgG
Liver
vasc
+(Q
Heart
− L
Heart
)· CIgG
Heart
vasc
+(Q
Kidney
− L
Kidney
)· CIgG
Kidney
vasc
+(Q
Skin
− L
Skin
)· CIgG
Skin
vasc
+(Q
Muscle
− L
Muscle
)· CIgG
Muscle
vasc
+(Q
SC
− L
SC
)· CIgG
SC
vasc
− Q
Lung
· CIgG
Plasma
+Q
LymVes
· CIgG
LymVes
(4.27)
V
Plasma
dCmAb
Plasma
dt
=(Q
Liver
− L
Liver
)· CmAb
Liver
vasc
+(Q
Heart
− L
Heart
)· CmAb
Heart
vasc
+(Q
Kidney
− L
Kidney
)· CmAb
Kidney
vasc
+(Q
Skin
− L
Skin
)· CmAb
Skin
vasc
+(Q
Muscle
− L
Muscle
)· CmAb
Muscle
vasc
+(Q
SC
− L
SC
)· CmAb
SC
vasc
− Q
Lung
· CmAb
Plasma
+Q
LymVes
· CmAb
LymVes
+IVmAb(t)
(4.28)
Table 4.3 lists model parameter and variable definitions. In the above equations IgG0 represents
the endogenous IgG production rate and IVmAb(t) is the rate of the intravenous infusion.
4.2.2.4 Peripheral sampling site
To better reflect the sampled peripheral mAb plasma concentration, especially for early measure-
ments, the model incorporates a separate peripheral sampling site compartment, following the
approach reported in [135]. The peripheral sampling is composed of contributions from skin and
muscle (Figure 4.1) as follows: CmAb
peri
=0.7· CmAb
Skin
vasc
+0.3· CmAb
Muscle
vasc
, where CmAb
peri
is
the measured peripheral mAb concentration, CmAb
Skin
vasc
is the administrated mAb concentration
in vascular space of skin and CmAb
Muscle
vasc
is the mAb concentration in muscle vascular space.
4.2.2.5 Models for other organs/tissues
The model structures of other organs and tissues are the same as those presented above for the
subcutaneous tissue (Figure 4.2a and Eqs. 4.1 to 4.20). Since these organs do not receive direct
administration of mAb, the distribution of mAb into local lymphatic capillaries is expected to be
61
low due to significantly slower lymph flow relative to plasma. Thus, a local lymphatic capillary
compartment for these organs was not included. The supplemental materials include the full set of
equations for the complete model presented in this work.
4.2.3 Model parameters
4.2.3.1 Individual organs and tissues
Values for those model parameters that are related to the individual organs/tissues were fixed
basedonliteraturereportedvaluesforhumans[49], includingvascularvolume(V
Organ
vasc
), interstitial
volume(V
Organ
inter
), cellvolume(V
Organ
cell
), plasmaflow( Q
Organ
)andlymphflow( L
Organ
)(Table4.4).
The FcRn concentrations in liver and gastrointestinal tract, which are assumed constant, were
assigned based on reported values obtained using quantitative Western blot [21], while the FcRn
concentrations in all other organs (CFcRn
Organ
7.4
) were fixed at 3.3 x 10
4
nM as in [51] (see FcRn
column in Table 4.4).
4.2.3.2 Central venous and lymphatic systems
Total body plasma volume used in the model was 2750 ml, assuming a total blood volume of 5 L
and hematocrit of 0.45. The sum of individual organ vascular volumes (V
Organ
vasc
) was subtracted
from total plasma volume, yielding the volume of central venous plasma pool (V
Plasma
= 1202 ml).
The total plasma flow ( Q
SC
) was calculated as the sum of the individual organ/tissue plasma flows
(3.152 x 10
6
ml/day). Values of the parameters related to central venous system are listed in Table
4.5. AlsolistedinTable4.5, arethevaluesusedfortheparametersofthecentrallymphaticsystem.
The model used a total number of lymph nodes in humans of 600 [27], with an average lymph node
volume of 0.292 ml [113, 136]. Thus, the volume of the composite lymph node compartment in the
model (V
LymNode
) was calculated as 175.2 ml. After water reabsorption in lymph node, post-nodal
lymph flow ( Q
LymNode
) is 4000 ml/day [27]. The transport of endogenous IgG and mAb in the
lymph node are described by Eqs. 4.23 and 4.24, with a rate constant (k
LymNode
), calculated as
the ratio of Q
LymNode
and V
LymNode
, of 23.83 day
− 1
. The lymphatic vessel compartment includes
62
Table 4.4: Fixed physiological parameters of individual organs/tissues specified in the whole-body
PBPK model
Tissue Vascular volume (ml)
a
Interstitial volume (ml)
a
Cell volume (ml)
a
Plasma flow (ml/day)
a
Lymph flow (ml/day)
b
FcRn (nM)
c
Vascular IgG clearance via pinocytosis (ml/day) Endosomal volume (ml)
V
Organ
vasc V
Organ
inter
V
Organ
cell
Q
Organ
L
Organ
CFcRn
Organ
7.4
CL
Organ
pino
V
Organ
endo
Lung 99.9 300 599 3.152 x 10
6
d 3463 3.300 x 10
4
3.306 0.0124
Heart 15 42.9 242 1.728 x 10
5
190.1 3.300 x 10
4
0.9927 0.003723
Liver 181 362 1270 1.152 x 10
6
1267 3.300 x 10
4
5.986 0.02245
Spleen 17 34.7 122 1.987 x 10
5
218.6 3.300 x 10
4
0.5744 0.002154
GI Tract 43 373 1730 6.739 x 10
5
741.3 4180 7.104 0.02664
Kidney 28.4 96.6 159 9.072 x 10
5
997.9 3.300 x 10
4
0.9397 0.003524
Muscle 700 4560 2.97 x 10
4
5.947 x 10
5
654.2 3.300 x 10
4
115.8 0.4343
Skin 462 2313d 6110 3.168 x 10
5
348.5 3.300 x 10
4
22.5 0.08438
SC site 0.62 3.115 8.23 2952 3.24 3.300 x 10
4
0.03961 0.0001485
a
From Baxter et al [49]
b
Assumed to be 0.11 % of organ plasma flow, resulting in total prenodal lymph flow 8000 ml/day
as reported in [27]
c
Fixed to whole-body load of FcRn concentration measured by Li et al. [21] or from Glassman et
al [51]
Table 4.5: Fixed parameter values for central venous and lymphatic systems
Parameter (unit) Value Description Ref.
Central venous
Q
Plasma
(ml/day) 3.152 x 10
6
Plasma flow (see text)
V
Plasma
(ml) 1202 Central plasma volume (see text)
IgG0 (nmol/day) 1.540 x 10
4
Production rate of endogenous IgG [101]
Central lymphatic
V
LymNode
(ml) 175.2 Volume of lymph node compartment [27, 136]
Q
LymNode
(ml/day) 4000 Post-nodal lymph flow [27]
k
LymNode
(day
− 1
) 23.83 Rate constant of mAb leaving lymph node (see text)
V
LymVes
(ml) 11.35 Volume of lymphatic vessel compartment [113, 137]
Q
LymVes
(ml/day) 4000 Lymph flow rate from lymphatic vessel compartment (see text)
σ 0.2 Lymph reflection coefficient [51, 52]
the lymphatic trunks, thoracic duct and cysterna chyli. The volume of the lymphatic trunk was
calculatedas0.94ml,usingalengthof30cmandradiusof1mm[137]. Thevolumeofthoracicduct
and cysterna chyli were determined to be 8.84 ml and 1.57 ml, respectively [113]. Thus, the volume
oflymphaticvessels(V
LymVes
)summedto11.35ml. Thelymphflowratethroughlymphaticvessel
(Q
LymVes
) was set at 4000 ml/day – the sum of the lymph flow from all organs/tissues. Finally,
the reflection coefficient for lymph ( σ ) was set at 0.2 in the lymphatic vessel and used for all organs
[51, 52].
4.2.3.3 Subcutaneous tissue site
The last row of Table 4.4 also lists values for parameters associated with the SC injection site. The
interstitial space volume at the SC site (V
SC
inter
) was set as 3.115 ml based on the previous work
63
Table 4.6: Fixed parameter values for endosomal processing and trans-capillary transport of mAbs
Parameter (unit) Value Description Ref.
All organs, tissues and mAbs
τ (day) 0.001875 Transit time of IgG between endosomal sub-compartments [51]
fIgGrecyc 0.715 IgG recycling fraction of IgG-FcRn complex back to vascular space [50]
k
6.0
on
(nM
− 1
day
− 1
) 13.42 Association rate constant between IgG and FcRn at pH = 6.0 [51]
k
6.5
on
(nM
− 1
day
− 1
) 1.354 Association rate constant between IgG and FcRn at pH = 6.5 [51]
k
7.0
on
(nM
− 1
day
− 1
) 0.6336 Association rate constant between IgG and FcRn at pH = 7.0 [51]
k
7.4
on
(nM
− 1
day
− 1
) 0 Association rate constant between IgG and FcRn at pH = 7.4 [51]
k
6.0
off
(day
− 1
) 573.1 Dissociation rate constant between IgG and FcRn at pH = 6.0 [51]
k
6.5
off
(day
− 1
) 573.1 Dissociation rate constant between IgG and FcRn at pH = 6.5 [51]
k
7.0
off
(day
− 1
) 573.1 Dissociation rate constant between IgG and FcRn at pH = 7.0 [51]
k
7.4
off
(day
− 1
) 0 Dissociation rate constant between IgG and FcRn at pH = 7.4 [51]
rS (nm) 4.44 Small pore radius [19]
rL (nm) 22.85 Large pore radius [19]
α S 0.958 Fractional hydraulic conductance of small pores [19]
α L 0.042 Fractional hydraulic conductance of large pores [19]
Adjusted for mAb MW (150 kDa IgG as an example); All organs and tissues
ae (nm) 4.81 Stokes Einstein radius [19]
σ S 0.998 Small pore vascular reflection coefficient [19]
σ L 0.18 Large pore vascular reflection coefficient [19]
A/A0S 9.280 x 10
− 7
Fractional accessible pore size of small pore [19]
A/A0L 0.349 Fractional accessible pore size of large pore [19]
PeS 9.82 Peclet number of small pores [19]
PeL 4.48 Peclet number of large pores [19]
Adjusted for mAb MW and individual organs and tissues (150 kDa IgG in liver as an example)
PS
Liver
S
(ml/day) 0.156 Permeability-surface area product of small pores [19]
PS
Liver
L
(ml/day) 97.61 Permeability-surface area product of large pores [19]
J
Liver
iso
(ml/day) 481.5 Circular isogravimetric flow [19]
J
Liver
S
(ml/day) 732.3 Lymph flow through small pores [19]
J
Liver
L
(ml/day) 534.7 Lymph flow through large pores [19]
[60], yielding a ratio of SC to skin interstitial volumes (V
SC
inter
/V
Skin
inter
) of 0.00135. The same fraction
was used for the other volumes of the SC site, yielding a SC site total volume and vascular volume
of 11.97 ml and 0.6200 ml, respectively. Based on radiobelled IgG studies [130, 138], the baseline
lymph flow from a SC injection site ( L
SC
) was determined to be 3.240 ml/day. As with the other
organs and tissues, L
SC
was also assumed to be 0.11 % of the SC plasma flow rate ( Q
SC
= 2952
ml/day). Based on lymphangiograms performed in control arms of subjects [129], the maximum
spread of local lymphatic vessels from the injection site was fixed at 1.59 cm and the total length
of lymphatic capillaries in a 1-cm annulus was 385 cm, resulting in a lymphatic capillary length of
612.15 cm. Assuming an average radius of 0.0274 mm [113, 137], the volume of the local lymphatic
capillary compartment (V
LymCap
) was calculated as 0.0144 ml.
4.2.3.4 Endosomal processing and two-pore transcapillary transport
The IgG-FcRn interaction rate constants (k
pHx
on
and k
pHx
off
) and endosomal transit rate (τ ) of IgG
were fixed (see Table 4.6) at values used in the PBPK model in [51]. The transit time ( τ ) of IgG in
the fourth endosomal sub-compartment was assumed to be 0.001875 day, the same as the previous
transit steps. The volume of endosomal sub-compartment for each organ (V
Organ
endo,sub
) was calculated
as one fifth of its endosomal space ( V
Organ
endo,sub
=V
Organ
endo
/5).
64
The details involving calculation of the parameter value of the two pore model are presented in
Li and Shah [19] and summarized below. The values for the four assigned parameters (r
S
, r
L
, α S
,
α L
) related to pore radius and fractional hydraulic conductance of two-pore theory transportation
are list in Table 4.6 [19]. Values for other parameters associated with the two-pore model for
transcapillary transport depend on the MW of the mAb, the specific organ/tissue, or both. The
Stokes Einstein radius in nm of each mAb was calculated as: a
e
= 0.0483· MW
0.386
, with MW in
g/mol. The vascular reflection coefficient of mAb through small pores ( σ S
) and large pores (σ L
)
were calculated:
σ S
=1− 0.8489· e
− 0.00004·MW
,σ L
=0.000035· MW
0.717
(4.29)
Thefractionalaccessibleporesizeofsmallpores(A/A0
S
)andlargepores(A/A0
L
)weredetermined
from the following equations [19]:
A/A0
S
=0.2352· e
− 0.00008295·MW
+0.7767· e
− 0.00053095·MW
(4.30)
A/A0
L
=0.3429· e
− 0.00012175·MW
+0.6571· e
− 0.00000421·MW
(4.31)
With these mAb specific parameter values, the Peclet number of small pores ( Pe
S
) and large pores
(Pe
L
) were calculated for each mAb as:
Pe
S
=
(− X
J
+α S
)· (1− σ S
)· a
e
· r
S
2
X
P
· (A/A0
S
)· α S
,Pe
L
=
(X
J
+α L
)· (1− σ L
)· a
e
· r
L
2
X
P
· (A/A0
L
)· α L
(4.32)
where X
P
was fixed at 13197 nm
3
and was calculated to be 0.38 [19]. The organ-dependent
permeability-surface area product of small pores (PS
Organ
S
) and large pores (PS
Organ
L
) can be
calculated for each mAb as:
PS
Organ
S
=
X
P
a
e
· A
A0
s
· (1− α L
)· L
Organ
r
S
2
,PS
Organ
L
=
X
P
a
e
· A
A0
L
· α L
· L
Organ
r
L
2
(4.33)
65
where L
Organ
is the organ lymph flow as given in Table 4.4. Finally, the circular isogravimetric
flow, lymph flow through small and large pores for each organ/tissue and for each mAb can be
calculated as:
J
Organ
iso
=X
J
· L
Organ
,J
Organ
S
=(1− α L
)· L
Organ
− J
Organ
iso
,J
Organ
L
=α L
· L
Organ
+J
Organ
iso
(4.34)
4.2.3.5 Baseline values for endogenous IgG
The endogenous IgG is produced in the central plasma pool at a constant rate (IgG0) of 1.54 x
10
4
nmol/day, as determined in a previously reported IgG tracer kinetics study [28]. The baseline
endogenous IgG concentration (CIgG
Plasma
) in the plasma pool was reported as 12.1 mg/ml in
[101]. The whole-body pinocytosis rate (CL
Total
pino
) was estimated so as to achieve a steady state
endogenousIgGconcentration(CIgG
Plasma
)of12.1mg/ml, assumingindividualorganpinocytosis
clearances (CL
Organ
pino
) in proportion to the organ/tissue total volumes: CL
Organ
pino
=
V
Organ
V
Total
· CL
Total
pino
.
The endosomal volume of each organ (V
Organ
endo
) was calculated as reported in [102]: V
Organ
endo
=
2· CL
Organ
endo
· τ . Values of CL
Organ
pino
and V
Organ
endo
for each organ are listed in Table 4.4.
4.2.4 Model parameter estimation: IV and SC studies
Prior to mAb administration, the model was simulated to reach steady state for all the endogenous
IgG concentrations. The plasma concentration-time data for each of the 20 IV administered mAbs
(Table 4.1) were used to estimate the drug specific parameters S
pino
and S
diff− conv
. For mAbs
administered at different doses, the data were pooled to obtained a single set of estimates of S
pino
and S
diff− conv
for each mAb. The ADAPT software (version 5) [100] was used to obtain the
maximum likelihood estimates for these parameters assuming an additive plus proportional error
variance model. The Na¨ ıve pooled data (NPD) analysis application was employed for mAbs with
pooled multiple dose data, and the individual estimation (ID) application was used for mAbs with
a single dose.
For the 12 mAbs for which plasma concentration-time data were also available following SC as
well as IV administration (Table 4.2), the subcutaneous site drug specific parameters S
SC
LymUpt
and
66
CL
SC
LymCap
were estimated. In the estimation involving each of these mAbs, the drug specific values
for the parameters S
pino
and S
diff− conv
were fixed at their estimated values as obtained from their
IV data. The maximum likelihood estimates of S
SC
LymUpt
and CL
SC
LymCap
were obtained using either
the NPD or ID application in ADAPT, as described above for the IV data.
4.2.5 Model predicted bioavailability
Using the model with the drug-specific parameters values, the bioavailability for each of the 12
SC administered mAbs was determined. The model-based bioavailability was calculated as the
total amount of the mAb dose absorbed through local SC lymphatic capillary space plus the total
amountdeliveredviatheSCsitebloodflow, dividedbytheSCdose. ForthemAbsadministeredat
different doses, bioavailability was calculated separately for each dose, with the average reported.
BioavailabilitywasalsocalculatedfromthemeandataforeachmAbvianoncompartmentalanalysis
(NCA) using the in-house software.
4.2.6 Biophysical determinants of drug specific model parameters
A central goal of our work was to determine whether biophysical properties of mAbs, beyond MW,
canbeincorporatedinaPBPKmodeltopredictthepharmacokineticcharacteristicsofthedifferent
mAbs. Toward this end, we explored the relationships between the estimated drug-specific model
parameters (S
SC
LymUpt
and CL
SC
LymCap
for SC, and S
pino
and S
diff− conv
for both IV and SC) and
different biophysical characteristics of mAbs. Two charge related attributes, pI and a measure
based on the local charge of the complementarity-determining regions (CDR) of the mAb (patches
of positive charge, PPC metric), were explored. The pI of each mAb was calculated with the
Protein isoelectric point calculator using the mAb amino acid sequence [139]. The PPC metric
for each mAb was calculated with the Therapeutic Antibody Profiler platform using the mAb’s
variable domain sequence [65]. Several mAbs without available values of pI or PPC metric were
excluded from this analysis.
67
4.2.7 Prediction of SC PK of other mAbs
The PBPK model incorporating the regression relationships based on biophysical properties iden-
tified above was used to predict the plasma PK of four additional mAbs not used in the model
development (Omalizumab, Tildrakizumab, Ixekizumab, Lanadelumab). For each of these mAbs,
modelpredictionswerebasedonpopulationssimulations(n=1000)ofthePBPKmodel(SIMappli-
cation in ADAPT), where the drug-specific parameters ( CL
SC
LymCap
, S
SC
LymUpt
, S
pino
andS
diff− conv
)
were assumed to follow a log normal distribution. The mean and standard deviations of these pa-
rameters in the simulation were based on the estimation results from the IV and SC data described
above, as well as identified biophysical property regression relations.
4.2.8 Simulations and sensitivity analyses
For the SC mAbs, simulation analyses were also conducted to quantify the contribution to bioavail-
ability from the lymphatic pathway and blood circulation. In addition, sensitivity analysis was
performed using one of the SC mAbs (Golimumab) to quantify the influence of lymph flow on PK,
by altering lymph flow by 0.1-, 0.5-, 10-, 100-fold of the original value.
4.3 Results
4.3.1 Intravenous pharmacokinetics
ThetotalIgGclearanceviapinocytosisinalltheorgans/tissues(CL
Total
pino
)wasestimatedtobe157.3
ml/day (relative standard errors, RSE% = 11.6) using the baseline endogenous IgG concentration
of 12.1 mg/ml [101]. The ratio of simulated steady-state IgG organ interstitial concentration and
plasma concentration was 0.41, which agrees with previous reported range of 0.1-0.5 in [10].
For each of the 20 IV mAbs, the values for the two estimated parameters, S
pino
and S
diff− conv
,
are listed in the right two columns of Table 4.1 along with their percent relative standard errors.
The results in Figure 4.3 indicate that the model can predict the plasma PK of each of these 20
mAbs for the different doses administered, with R
2
ranging from 0.941 for Siltuximab to 0.998
68
for Belimumab. The parameters representing the mAb specific differences in tissue exchange via
pinocytosis (S
pino
) and diffusion/convection ( S
diff− conv
) were estimated with good precision as
shown in Table 4.1. The estimates of S
pino
range from 0.627 to 2.14, with a mean, median and
standard deviation of 1.07, 0.949 and 0.337, respectively. While the estimates for S
diff− conv
range
from0.452to2.21,withamean,medianandstandarddeviationof1.18,1.16and0.429,respectively.
4.3.2 Subcutaneous pharmacokinetics
The results of estimating the two SC site drug-specific parameters, S
SC
LymUpt
and CL
SC
LymCap
, for
each of the 12 SC administered mAb are listed in Table 4.2 (S
pino
and S
diff− conv
fixed at values
obtainedfromthecorrespondingIVanalysis),whilethemodelpredictedplasmaconcentrationtime
course is shown with the measured plasma concentrations for each mAb in Figure 4.4. The values
of R
2
ranges from 0.948 for Daclizumab to 0.987 for Gevokizumab. As shown in Table 4.2, both
S
SC
LymUpt
and CL
SC
LymCap
were estimated precisely. The value of estimated CL
SC
LymCap
ranges from
0.441 to 3.66, with a mean, median and standard deviation of 1.25, 0.874 and 0.998. The estimates
of S
SC
LymUpt
are in the range of 0.0904 and 0.301, with a mean, median and standard deviation of
0.210, 0.211 and 0.0631, respectively.
The model-based prediction of bioavailability is plotted versus the NCA calculated value for
each of the 12 mAbs in Figure 4.5. As expected, these results are in good agreement (R
2
= 0.91),
given that the PBPK model yielded good predictions of the plasma concentration-time data for
each mAb as shown in Figure 4.4.
The relationship between the observed SC PK characteristics bioavailability and T
max
, and the
estimatedSCsitemodelparametersCL
SC
LymCap
andS
SC
LymUpt
areshowninFigure4.6. Bioavailability
is correlated with CL
SC
LymCap
(p < 0.0001) but not with S
SC
LymUpt
, while T
max
is correlated with
S
SC
LymUpt
(p < 0.001) but not with CL
SC
LymCap
.
4.3.3 Biophysical determinants of mAb SC absorption
The calculated values of the isoelectric points (pI) range from 6.36 to 7.37, while the local charge
of the complementarity-determining regions, PPC metric, ranges from 0 to 1.36 for each of the
69
Figure 4.3: Model predicted and observed mAb plasma concentration versus time profiles after
IV administration. The solid lines are model predictions. The symbols are digitized data from
literature
70
Figure 4.4: Model predicted and observed mAb plasma concentration versus time profiles after
SC administration. The solid lines are model predictions. The symbols are digitized data from
literature
71
Figure 4.5: Linear correlation evaluation between NCA calculated bioavailability (%) and model
predicted bioavailability (%). All mAbs with SC PK data are included. The solid line represents
the regression line (R
2
= 0.91). The dashed line represents the hypothetical line of identity
72
Figure 4.6: Correlation of observed SC PK characteristics versus estimated SC tissue model pa-
rameters. (a) Negative association between observed bioavailability (%) and estimated CL
SC
LymCap
(p = 6.77 x 10
− 05
). (b) No significant association between observed bioavailability (%) and es-
timated S
SC
LymUpt
(p = 0.547). (c) No significant association between observed T
max
(day) and
estimated CL
SC
LymCap
(p = 0.217). (d) Negative correlation between observed T
max
(day) and esti-
mated S
SC
LymUpt
(p = 0.000143)
73
12 mAbs. There was no significant linear relation (results not shown) between pI and any of the
four drug specific model parameters ( S
pino
, S
diff− conv
, CL
SC
LymCap
,S
SC
LymUpt
). However, PPC was
positively correlated with S
pino
and CL
SC
LymCap
, but not with S
diff− conv
and S
SC
LymUpt
, as shown
in Figure 4.7. The resulting regression relationships are: CL
SC
LymCap
= 0.66+1.77· PPC (R
2
=
0.69, p = 0.000764, residual standard error = 0.5789), S
pino
= 0.92+0.64 · PPC (R
2
= 0.58, p
= 0.000661, residual standard error = 0.245). By incorporating PPC into the PBPK model, the
standard deviation of CL
SC
LymCap
decreases from 0.998 to 0.579, while the standard deviation of
S
pino
is reduced from 0.362 to 0.245. These results suggest that this model explains some of the
differences in pinocytosis and SC lymphatic capillary clearance among the mAbs.
4.3.4 Model-based prediction of SC PK of other mAbs
The PBPK model incorporating the regression equations relating the mAb PPC metric and the
model parameters S
pino
and CL
SC
LymCap
was used to predict the plasma PK of the four mAbs
(Omalizumab, Tildrakizumab, Ixekizumab, Lanadelumab) not used in the model development.
Figures 8 and 9 show the resulting 5th – 95th percentile prediction intervals for Omalizumab (PPC
= 0.0057) and Tildrakizumab (PPC = 0), respectively, for two different SC doses in each case.
Corresponding results are shown in Figure 4.10 for Ixekizumab (PPC = 0.193) following one dose
of160mg. ForeachofthesethreemAbs,themeasuredconcentration-timeprofileiscenteredwithin
the model predicted 5th – 95th percentile intervals.
When the model was used to predict the plasma PK of Lanadelumab (PPC = 1.4502) at each
of four different doses (from 7.8 mg to 249 mg), as shown in Figure 4.11, the model underpredicted
plasma concentrations of during the elimination phase. These results suggest that the value used
for S
pino
was likely over estimated based on the PPC regression equation (see Discussion).
4.3.5 Simulations and sensitivity analyses
Using the PBPK model, we compared the contributions to SC absorption attributable to the
lymphatic pathway and the blood perfusion pathway, for each of the 12 mAbs. For these mAbs,
the absorption through lymphatic pathway accounts for between 91.6 % to 99.0 % of the total
74
Figure4.7: CorrelationofestimatedmodelparametersversusmAbPPCmetricatCDRvicinity. (a)
PositiveassociationbetweenestimatedS
pino
valueandmAbPPCmetric: S
pino
=0.92+0.64· PPC
(R
2
= 0.58, p = 0.000661, residual standard error = 0.245). (b) No significant association be-
tween estimated S
diff− conv
value and mAb (p = 0.102). (c) Positive correlation between estimated
CL
SC
LymCap
andmAbPPCmetric: CL
SC
LymCap
=0.66+1.77· PPC (R
2
=0.69,p=0.000764, residual
standard error = 0.579). (d) No significant association between estimated S
SC
LymUpt
and mAb PPC
metric (p = 0.685)
75
Figure 4.8: Model-predicted and digitized concentration versus time profiles of Omalizumab for
each dose. Filled dots indicate digitized Omalizumab PK profiles from literature, after single SC
dose of (a) 150 mg, (b) 300 mg. Solid lines show the predicted median. Dash lines indicate the
predicted 5th and 95th percentiles. The shaded areas display the 5th to 95th percentile population
simulation regions
Figure 4.9: Model-predicted and digitized concentration versus time profiles for Tildrakizumab for
each dose. Filled dots indicate digitized Tildrakizumab PK profiles from literature, after single SC
dose of (a) 50 mg, (b) 200 mg. Solid lines show the predicted median. Dash lines indicate the
predicted 5th and 95th percentiles. The shaded areas display the 5th to 95th percentile population
simulation regions
76
Figure4.10: Model-predictedanddigitizedconcentrationversustimeprofilesforIxekizumab. Filled
dotsindicatedigitizedIxekizumabPKprofilesfromliterature, aftersingleSCdoseof160mg. Solid
lines show the predicted median. Dash lines indicate the predicted 5th and 95th percentiles. The
shaded areas display the 5th to 95th percentile population simulation regions
77
Figure 4.11: Model-predicted and digitized concentration versus time profiles for Lanadelumab for
each dose. Filled dots indicate digitized Lanadelumab PK profiles from literature, after a single
SC dose of (a) 7.8 mg, (b) 26 mg, (c) 71 mg, (d) 249 mg. Solid lines show the predicted median.
Dash lines indicate the predicted 5th and 95th percentiles. The shaded areas display the 5th to
95th percentile population simulation intervals
78
amount of dose absorbed amount. This result showing the primary role of lymphatic uptake in
the absorption of mAbs after SC dose is consistent with the previous studies in sheep [140] and
other studies as reviewed in [23]. The PBPK model was also employed to predict the change of
SC absorption rate over time. The SC absorption is essentially complete 25 days after SC injection
(Figure 4.12). The simulation in Figure 4.12 also demonstrates the major contribution of the
lymphatic system to absorption of subcutaneously administered mAbs using Golimumab as an
example. From Figure 4.12 the peak absorption rate via the lymphatic pathway is predicted to be
of 72.2 nmol/day, compared to 4.53 nmol/day through blood perfusion.
The sensitivity analysis results shown in Figure 4.13 illustrates the role of subcutaneous lymph
flow( L
SC
)ontheplasmapharmacokineticsfollowingSCadministrationofmAbs. ForGolimumab,
as an example, reducing lymph flow by 0.1- and 0.5-fold can significantly delay the SC absorption
and decrease bioavailability to 5.79% and 31.1%, respectively, compared to 49.8% with the original
lymph flow value. In contrast, increasing lymph flow (i.e. 10-fold difference) can facilitate the
absorption of mAb, leading to greater absorption rate and larger bioavailability of 91.3%.
4.4 Discussion
In this work, a whole-body PBPK model is presented with the goal of predicting the clinical
pharmacokinetics of mAbs following SC administration in humans. The model incorporates two
mAb-specific parameters to represent differences in pinocytosis rate and the diffusive/convective
transportratesamongthemAbs. AttheSCadministrationsite,twoadditionalparametersareused
to represent differences in lymphatic capillary uptake and clearance among mAbs. The resulting
model reliably predicted the bioavailability and T
max
, as well as the overall plasma concentration
time profile, of each of the 12 SC mAbs investigated. Further model-based investigation found that
themAb-specificdifferencesinorgan/tissuepinocytosistransport,aswellasSClymphaticcapillary
clearancecouldbeexplained,inpart,basedonthelocalchargeofthecomplementarity-determining
regions of the mAb (patches of positive charge metric), but not the antibody’s isoelectric point.
The composite model’s predictive capability was evaluated through Monte Carlo simulations of
the SC PK of another four mAbs not used in model development. Overall, this study provides a
79
Figure 4.12: Simulation of absorption rate of different pathways after the SC dosing, using Goli-
mumab as an example. The solid line represents the total absorption rate after SC administration.
The dashed and dotted lines denote the mAb absorption rate through the lymphatic system and
blood perfusion, respectively
80
Figure 4.13: Sensitivity analysis of SC site lymph flow, using Golimumab as an example. The solid
line represents the original Golimumab PK profile after SC administration. The two-dashed line,
long-dashed line, dotted line and dashed line denote PK profiles after changing SC site lymph flow
to 100-, 10-, 0.5- and 0.1-fold of the original parameter value (0.00225 ml/min), respectively
81
platform PBPK model to characterize and predict the plasma disposition of monoclonal antibodies
following SC administration.
After SC administration, mAbs transverse the ECM to reach lymphatic capillaries or blood
capillaries before absorption [17]. In our model, the ECM of the SC space was represented by
a single homogenous interstitial compartment, with the transport and subsequent lymphatic up-
take of the mAb expressed by a single term representing the mAb-specific interstitial to lym-
phatic capillary transport rate (S
SC
LymUpt
· L
SC
). Previous evidence suggests that the uptake of
IgG from the depot following its SC administration is first-order [130, 138], further supporting the
tissue-level model assumptions used in the SC model. While previous studies have investigated
the pre-systemic degradation of biotherapeutics [141–143], quantitative information on local tis-
sue catabolism/degradation is lacking [23]. Thus, in the model, the overall mAb clearance due to
catabolism/degradation in the SC site and/or in the draining lymphatics was represented as a net
clearance from the local lymphatic capillary space (CL
SC
LymCap
). Despite these simplifying assump-
tions, the model presented is qualitatively consistent with other studies reported in the literature.
For example, the increase of lymph flow by massage or heat at the injection site has been shown to
aid in the absorption of biotherapeutics [144, 145], which is consistent with the model sensitivity
analysis results in Figure 4.13. Also, the model prediction demonstrates the dominant role of the
lymphatic system in the SC absorption (more than 90%), in line with experimental results in sheep
[140]. The model predictions are also consistent with the central role of ECM transport and lym-
phaticuptakeindeterminingtherateofmAbabsorption(Figure4.6),asreportedinpreclinicaland
clinical studies using hyaluronidase to enhance the subcutaneous absorption [45, 146, 147]. In ad-
dition, the extent of SC absorption is largely determined by pre-systemic degradation (CL
SC
LymCap
),
as identified by the correlation analysis (Figure 4.6). This suggests that inhibiting pre-systemic
elimination can increase SC bioavailability, which can be achieved by co-administration of protease
inhibitors or via a dose saturation strategy, as mentioned in [17].
Various biophysical properties of mAbs have been considered in an effort to explain the sub-
stantial differences in PK behavior observed with different mAbs [32, 148]. Since the net charge of
a mAb in vivo will influence its interaction with tissue components and/or targets, the antibody’s
82
isoelectricpointhasbeeninvestigatedasadeterminantofitsPK[32,33,149]. Ithasbeenreported,
however, that differences in pI value of more than one unit are required to produce a detectable
change in an antibody’s systemic clearance and tissue distribution [33, 150, 151]. In this study, the
calculated pI values of the 12 mAbs examined range from 6.36 to 7.37. Given this narrow range it
is not unexpected that we did not find a significant relationship between mAb pI value and any of
the four mAb-specific model parameters.
It has been recently reported that the local charge of a mAb’s complementarity determining
regions may impact the charge balance and clearance of these large molecules [5, 34, 149]. Since
a mAb’s CDR directly contacts the antigen and mediates its binding, we investigated whether the
positive charge of mAb CDR vicinity (PPC metric) could explain some of the differences observed
in the four mAb-specific model parameters, and thus the variability in mAb PK behavior of the
mAbs investigated. As shown in Figure 4.7, we found that PPC could explain some of the inter-
mAb differences in S
pino
and CL
SC
LymCap
, but not S
diff− conv
or S
SC
LymUpt
. Futher investigation with
additionalmAboverarangeofPPCvaluesisneeded, however, tomorefullyevaluatetheseresults.
The positive correlation found between mAb PPC and S
pino
and between PPC and CL
SC
LymCap
can be understood as follows. Increased positive charge of CDR vicinity can lead to increased
interaction between mAbs and cell membranes, which consist of negatively charged components
such as heparin and sialic acid. This stronger binding then facilitates the entry of mAb into
endothelial cells through pinocytosis, which is represented in the model by the mAb specific scale
factor S
pino
that alters the pinocytosis clearance. This result is consistent with the study of Datta-
Mannanetal[34],inwhichtheyfoundthatbalancingchargeintheCDRofhumanizedabswithout
changing pI can reduce mAb binding to cells, and thus decrease mAb clearance.
The PPC metric was also found to be positively correlated with the SC site pre-systemic mAb
clearance (CL
SC
LymCap
). The pre-systemic degradation of mAbs can be mediated by proteases, as
well as soluble and cell-membrane receptors [17]. With higher positive charge in the CDR vicinity,
mAbsaremorelikelytobeexposedtosolubleorcell-surfacetargetswithnegativecharge,whichcan
leadtoreceptor-mediateduptakeanddegradationortheformationofimmunecomplexeswhichwill
be cleared [25]. In a study in cynomolgus monkeys [152], it was reported that increasing the mAbs
83
positive charge can lead to higher clearance at the administration site and to lower bioavailability,
in line with our findings. In another report, Sharma et al found that extreme variable domain
charges could result in faster mAb clearance [153].
It was also expected that the mAb-specific differences in the uptake rate from SC interstitial
space to lymphatic capillaries, represented by the scaling factor S
SC
LymUpt
in the model, would cor-
relate with PPC; but as shown in Figure 4.7b, this is not the case. The rate-limiting step of SC
absorption, mAb transport through ECM, is known to be effected by electrostatic interactions
and steric exclusion [154]. Steric exclusion renders diffusion rate slower with increased content of
hyaluronic acid in ECM and larger molecular weight of the drug, also with the possible influence
of molecular shape [154]. Therefore, more positively charged mAbs would be expected to be trans-
ported within the ECM more slowly due to their increased interaction with tissue components [17].
However, the differences in some external and internal factors (e.g. steric exclusion, formulation,
physical activity level, location of the SC injection, anesthesia) may obscure this relationship. As
a result, any influence of positive charge of CDR vicinity of mAbs may not be detectable using the
model.
Based on our modeling and the mAbs analyzed, the PPC metric did not correlate with the
differences estimated in the mAb-specific model parameter related to diffusive/convective trans-
port (S
diff− conv
), suggesting that other non-charged based factors may be responsible for these
differences. Also, the differences in S
diff− conv
are beyond those due to the differences in mAb
molecular weight, since the model incorporates the MW-based two-pore theory. Other factors not
considered may influence the convective transport that depends on the difference between the hy-
drostatic pressure gradient and osmotic pressure gradient, such as structure of the capillary wall
(e.g. diameter of paracellular pores) and physiological conditions of subjects [10, 155].
In developing the model presented in this report, pooled mean plasma mAb concentration-time
profiles were used as obtained from more than 30 published literature reports. Thus, differences in
these studies due to, for example, subject’s age, sex, weight, physiological state, as well as mAb
formulation and administration site could not be incorporated as explanatory covariates. Instead
the effects of these differences are represented as variability in the four estimated mAb-specific
84
model parameters, thus resulting in an overestimation of the variability in these properties. Given
the availability of only plasma mAb concentrations, we were unable to evaluate the ability of the
model to predict organ/tissue PK.
In summary, in this study, we established a whole-body PBPK model for predicting mAb PK
following SC delivery to support translational studies. Given its physiological basis, the developed
model enables us to: (1) gain mechanistic insights of events determining mAb SC absorption
and facilitate further optimization of SC delivery, (2) explain the influence on systemic PK by the
alternationofantibodyvariableregion,(3)supportfirst-in-humantranslationalstudiesforSCmAb
administration. Thismodelalsoprovidesthepotentialofpredictingperipheraltissueconcentration,
which can guide the design of effective dosing ranges and strategies in mAb development process.
The supplementary material of this chapter is available on:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7529865/
85
5 MonoclonalAntibodyPharmacoki-
neticsinCynomolgusMonkeysFol-
lowingSubcutaneousAdministration:
Physiologically Based Model Pre-
dictionsfromPhysiochemicalProp-
erties
S.Hu,A.Datta-MannanandD.Z.D’Argenio. MonoclonalAntibodyPharmacokineticsinCynomol-
gus Monkeys Following Subcutaneous Administration: Physiologically Based Model Predictions
from Physiochemical Properties. AAPS Journal (submitted)
86
Abstract
Anintegratedphysiologicallybasedmodelingframeworkispresentedforpredictingpharmacoki-
netics and bioavailability of subcutaneously administered monoclonal antibodies in cynomol-
gus monkeys, based on in silico structure-derived metrics characterizing antibody size, overall
charge, local charge and hydrophobicity. The model accounts for antibody specific differences
in pinocytosis, transcapillary transport, local lymphatic uptake and pre-systemic degradation
at the subcutaneous injection site, and reliably predicts the pharmacokinetics of five different
wild type mAbs and their Fc variants following intravenous and subcutaneous administration.
Significant associations were found between subcutaneous injection site degradation rate and
the antibody’s local positive charge of its complementarity-determining region (R = 0.61, p =
0.00037), antibody pinocytosis rate and its overall positive charge (R = 0.59, p = 0.00063), and
antibody paracellular transport and its overall charge together with hydrophobicity (R = 0.63,
p = 0.00096). Based on these results, populations simulations were performed to predict the
relationshipbetweenbioavailabilityandantibodylocalpositivecharge. Inaddition,modelsimu-
lations were conducted to calculate the relative contribution of absorption pathways (lymphatic
andblood), pre-systemicdegradationpathways(interstitialandlysosomal), andtheinfluenceof
injection site lymph flow on antibodybioavailability andpharmacokinetics. Theproposed phys-
iologically based modeling framework integrates fundamental mechanisms governing antibody
subcutaneous absorption and disposition, with structured-based physiochemical properties, to
predict antibody bioavailability and pharmacokinetics in vivo.
Keywords Subcutaneous monoclonal antibody absorption; antibody bioavailability; PBPK
model; FcRn interaction; non-specific binding; antibody physiochemical properties
87
5.1 Introduction
The lack of reliable in vitro or preclinical in vivo methodologies for predicting the bioavailability
and pharmacokinetics (PK) of subcutaneously (SC) administered monoclonal antibodies (mAbs)
and other biotherapeutics in humans has been identified as one of the major impediments to the
efficient development of subcutaneous delivery of therapeutic proteins [14, 156]. Although our
understanding of the mechanisms that govern the SC absorption of mAbs is incomplete, it has
been established that injection site physiology, formulation and antibody specific physiochemical
properties can affect absorption and pharmacokinetics of SC mAbs [29, 36].
Antibody physiochemical properties related to charge, hydrophobicity, aggregation, and FcRn
binding have been found to contribute to observed differences in SC absorption [14, 36]. For
example, mAbs with higher isoelectric points (higher overall positive charge) have stronger inter-
actions with SC tissues and lower bioavailability [31, 32]. More recently, it has been reported that
bothincreasedlocalpositivechargeinantibody’scomplementarity-determiningregions(CDR)and
strongerhydrophobicinteractionsnegativelyaffecttheextentandrateofSCabsorptioninratsand
cynomolgus monkeys [36]. The role of FcRn binding on SC bioavailability is less clear, with studies
in mice showing marked differences in bioavailability in animals with or without FcRn binding
[24, 42], while reports using cynomolgus monkeys found no change in bioavailability after Fc engi-
neering [43, 44]. Despite the insights provide by these and related studies, a reliable framework for
predicting mAb SC PK from antibody physiochemical properties or other preclinical information,
including animal models, remains elusive [14, 15, 157].
Theuseofphysiologicallybasedpharmacokinetic(PBPK)modelshascontributedtoadvancing
ourunderstandingofthedispositionofintravenously(IV)administeredmAbsandothertherapeutic
proteins in humans and in other species over the past 30 years [49–52]. Previous work has also
suggested that information on antibody non-specific interactions derived from in vitro assays can
be incorporated into a PBPK model in an effort to predict mAb IV PK in humans [158]. PBPK
models have been further applied to describe the absorption and PK following SC administration
of biotherapeutics including mAbs. A SC depot compartment was added to a PBPK model to
describe the SC PK of a pegylated peptide in monkeys and scaled to humans to predict plasma
88
PK of this peptide in a clinical trial [59]. A physiological absorption site model was included in
a PBPK model and used to describe the PK of different therapeutic proteins in humans [60]. A
PBPKmodelwithaSCtissuespacewasusedtoexploretheeffectsofproteinsizeonbioavailability
and pre-systemic degradation in mice [61]. There is, however, no report of PBPK models used to
predict bioavailability and PK of SC administered mAbs that incorporate information on antibody
specific physiochemical properties.
To address this deficiency, we extended our previously reported PBPK model developed for IV
mAb administration [158] to include a SC delivery site, based on our study of antibody PK in
cynomolgus monkeys following both IV and SC administration [43]. The SC injection site model
includes antibody extravasation via transcapillary transport and pinocytosis, pH dependent FcRn
processing,pre-systemicdegradation,andlymphaticuptake. Estimatedparametersassociatedwith
these processes were explored for their associations with antibody structure-based physiochemical
metrics related to their overall and local charge, hydrophobicity, and others. The antibodies used
were engineered with sequence differences in their heavy-chain and light-chain variable regions to
target differentsoluble antigens[43]. In addition, Fc variantsof eachantibody withT250Q/M428L
mutations were also evaluated. The developed model was applied to predict bioavailability of
the different mAbs based on identified structure-based physiochemical properties, calculate the
relative contribution of pre-systemic degradation and the effects of local lymph flow on SC PK
and bioavailability. Results illustrate the utility of the proposed PBPK model-based framework to
explain some differences in antibody pharmacokinetics and bioavailability following subcutaneous
administration.
5.2 Materials and Methods
5.2.1 Pharmacokinetics in cynomolgus monkeys
IV and SC PK of five wild type (WT) IgG4 antibodies and their variants with Fc mutation (listed
in Table 5.1) were investigated in cynomolgus monkeys as described previously [40, 43] and serve
as the basis for the modeling analysis in this study. Briefly, 12 male cynomolgus monkeys were
89
Table 5.1: Selected physiochemical properties of the antibodies evaluated in this study. See sup-
plemental material for complete list of values for all 26 properties
Antibody PSH
a
PPC
a
PNC
a
patch pos
b
patch hyd
b
patch cdr pos
b
A WT 120.5 0.25 0 150 1490 40
B WT 156.8 0.193 0 150 1650 40
C WT 162.4 0.1309 0 240 1430 60
D WT 177.9 0.0539 0.0734 280 1450 160
E WT 149.2 1.939 2.121 260 1350 80
A Var 120.5 0.25 0 160 1520 40
B Var 156.8 0.193 0 140 1670 80
C Var 162.4 0.1309 0 240 1410 60
D Var 177.9 0.0539 0.0734 300 1310 120
E Var 149.2 1.939 2.121 290 1350 90
a
calculated using Therapeutic Antibody Profiler (TAP) [65]
b
calculated using Molecular Operating Environment (MOE) [66]
PSH: patches of surface hydrophobicity across CDR
PPC: patches of positive charge across CDR
PNC: patches of negative charge across CDR
patch pos: area of positive protein patches
patch hyd: area of hydrophobic protein patches
patch cdr pos: area of positive protein patches near CDRs
equally divided into four groups. Each group was assigned to receive a single mixture dose of the
fiveWTmAbsorthefiveantibodyvariantsviaIVorSCadministration. Antibodiesweredissolved
in PBS (pH 7.4) with the concentration of 1.0 mg/kg of each. Concentrations of antibodies were
measuredforeachmonkeyatprespecifiedsamplingtime,usingvalidatedELISAs[43]. ThefiveWT
antibodies have sequence differences in their heavy-chain and light-chain variable regions, targeting
differentsolubleantigens[40]. TheFcvariantsinvolvedaT250Q/M428Lmutation[43]. Thetarget
antigen levels in cynomolgus monkeys were insignificant, so the clearances of these mAbs were not
affected by target-mediated drug disposition [43].
5.2.2 Structure-based physiochemical properties
Based on the sequence of each antibody, the Therapeutic Antibody Profiler (TAP) was used to
calculate five structure-based physiochemical metrics related to antibody variable domains [65].
90
Figure 5.1: PBPK model used for mAbs following subcutaneous (SC) administration in cynomol-
gus monkeys (overall model structure follows (24)). (a) Organ-level structure of the SC site, with
vascular space, endothelial space, interstitial space, local lymphatic capillary and cell space shown.
Pre-systemic degradation (CL
SC
) is included in the SC site interstitial space, along with local
lymphatic uptake (S
SC
LymUpt
). Transcapillary transport through diffusion and convection through
gaps between endothelial cells is depicted with endothelial pinocytosis uptake and endosomal pro-
cessing, which are applied also for other organs. (b) Endosomal subcompartments at a specific pH
= x, representing the competition between endogenous IgG and exogenous mAb for FcRn. Similar
model used to represent other organs (see supplemental information).
These five in silico determined metrics characterize antibodies in terms of the total length of the
CDR, patches of surface hydrophobicity across the CDR (PSH), patches of positive charge across
the CDR (PPC), patches of negative charge across the CDR (PNC) and Fv charge symmetry
parameter (SFvCSP) (supplementary material).
Antibody sequences were also used to determine 21 physiochemical properties using the Molec-
ular Operating Environment (MOE) software platform [66]. These structure-based metrics in-
clude the area of positive protein patches (patch pos), the area of hydrophobic protein patches
(patch hyd), and the area of positive protein patches near CDRs (patch cdr pos), among others
(see supplementary material).
5.2.3 PBPK model
The complete PBPK model used to analyze the monkey PK is presented in the supplementary
material, and is a modified version of the human PBPK model we reported previously for IV
91
administration[74,158]. Figure5.1depictsthesubsystemmodeladdedtorepresenttheSCinjection
site, including the vascular, endothelial, interstitial, cellular spaces and a local lymphatic capillary.
Other organs in the model have the same components as the SC site, without the local lymphatic
capillary uptake or pre-systemic degradation pathways. Within each organ, antibodies and IgG
are extravasated via transcapillary transport and pinocytosis, with the former described using the
a two-pore theory model [19, 158]. The equations governing the concentrations of endogenous IgG
(CIgG
SC
vasc
) and mAb (CmAb
SC
vasc
) within the SC site vascular space are:
V
SC
vasc
· dCIgG
SC
vasc
dt
=Q
SC
· CIgG
Lung
vasc
− (Q
SC
− L
SC
)· CIgG
SC
vasc
− CL
SC
pino
· CIgG
SC
vasc
+fIgG
recyc
· CL
SC
pino
· CIgFc
SC
recyc
− PS
SC
L
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
L
e
Pe
L
− 1
− PS
SC
S
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
S
e
Pe
S
− 1
− J
SC
L
· (1− σ L
)· CIgG
SC
vasc
− J
SC
S
· (1− σ S
)· CIgG
SC
vasc
(5.1)
V
SC
vasc
· dCmAb
SC
vasc
dt
=Q
SC
· CmAb
Lung
vasc
− (Q
SC
− L
SC
)· CmAb
SC
vasc
− S
pino
· CL
SC
pino
· CmAb
SC
vasc
+fIgG
recyc
· CL
SC
pino
· CmAbFc
SC
recyc
− S
diff− conv
· PS
SC
L
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
L
e
PeL
− 1
− S
diff− conv
· PS
SC
S
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
S
e
PeS
− 1
− S
diff− conv
· J
SC
L
· (1− σ L
)· CmAb
SC
vasc
− S
diff− conv
· J
SC
S
· (1− σ S
)· CmAb
SC
vasc
(5.2)
Allsymbolsforthemodelparametersandvariablesaredefinedinthesupplementarymaterial. Two
antibody specific scale factors were used to reflect antibody differences in transport, including one
scale factor (S
pino
) representing differences in pinocytosis and another scale factor ( S
diff− conv
) rep-
resenting differences in diffusion and convection [51, 74]. It is assumed that exocytosis is antibody
independent.
The corresponding equations for the concentrations of endogenous IgG (CIgG
SC
inter
) and mAbs
(CmAb
SC
inter
) in the SC interstitial space are:
V
SC
inter
dCIgG
SC
inter
dt
=(1− fIgG
recyc
)· CL
SC
pino
· CIgFc
SC
recyc
− CL
SC
pino
· CIgG
SC
inter
− L
SC
· CIgG
SC
inter
+PS
SC
L
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
L
e
Pe
L
− 1
+PS
SC
S
· (CIgG
SC
vasc
− CIgG
SC
inter
)· Pe
S
e
Pe
S
− 1
+J
SC
L
· (1− σ L
)· CIgG
SC
vasc
+J
SC
S
· (1− σ S
)· CIgG
SC
vasc
(5.3)
92
V
SC
inter
dCmAb
SC
inter
dt
=(1− fIgG
recyc
)· CL
SC
pino
· CmAbFc
SC
recyc
− S
pino
· CL
SC
pino
· CmAb
SC
inter
− S
SC
LymUpt
· L
SC
· CmAb
SC
inter
− CL
SC
· CmAb
SC
inter
+S
diff− conv
· PS
SC
L
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
L
e
PeL
− 1
+S
diff− conv
· PS
SC
S
· (CmAb
SC
vasc
− CmAb
SC
inter
)· Pe
S
e
PeS
− 1
+S
diff− conv
· J
SC
L
· (1− σ L
)· CmAb
SC
vasc
+S
diff− conv
· J
SC
S
· (1− σ S
)· CmAb
SC
vasc
(5.4)
The pre-systemic degradation of mAb is assumed to occur in the SC interstitial space (second line
of Equation 5.4). Two additional antibody specific parameters in the SC site ( S
SC
LymUpt
and CL
SC
)
describe the antibody’s local lymphatic uptake and pre-systemic degradation after SC administra-
tion (second line of Equation 5.4). The degradation of endogenous IgG in the interstitial space of
the SC site is assumed to be negligible (second line of Equation 5.3), because it is a small fraction
of the overall turnover of endogenous IgG.
The model structure used to describe the endosomal transit and processing (Figure 5.1) has
been detailed previously [158]. Briefly, three compartments were used to represent different stages
of antibody and IgG endosomal transit and processing (Figure 5.1). Antibodies and IgG interact
with FcRn in a pH-dependent manner, governed by mass action association and dissociation rate
constants k
pHx
on
and k
pHx
off
. Unbound IgG in the late endosome compartment (pH 6.0) is mainly
sorted to lysosome for degradation (represented by the K
e
term in Figure 5.1), while FcRn-bound
IgG is sorted to recycling endosome (pH 7.4b) for exocytosis (5). For the late endosome compart-
ment (pH 6.0), it was reported that about 95% of free IgG will be degraded in the lysosome [73].
Thus, we added another transport pathway for free IgG to the recycling endosome (pH 7.4b) with
a transit rate of 0.02*1/τ (1/day), in which τ (day) is the transit time of IgG between endosomal
sub-compartments.
5.2.4 Model parameter values – fixed
5.2.4.1 Individual organ/tissue values
Parameter values for different organ/tissues were fixed by pooling information from various sources
[52, 159, 160]. First, total organ volumes of a typical 3.8-kg cynomolgus monkey were based on
values reported in a previous PBPK model in monkeys (6.2 kg) by direct weight-based scaling [52].
The volumes of vascular and interstitial spaces of each organ (Table 5.2) were then calculated using
93
Table 5.2: Values of the fixed physiological parameters for individual organs/tissues
Organ/tissue Interstitial volume (ml)
a
V
Organ
inter
Cell volume (ml)
a
V
Organ
cell
Vascular volume (ml)
a
V
Organ
vasc Plasma flow (ml/day)
a
Q
Organ
Lung 5.405 13.72 2.756 5.332 x 10
5
Heart 2.567 13.51 1.266 2.128 x 10
4
Liver 11.46 92.61 10.54 1.302 x 10
5
Spleen 0.2589 2.772 0.6163 1.596 x 10
4
GI Tract 25.16 115.8 5.359 7.982 x 10
4
Kidney 2.51 11.08 3.146 1.011 x 10
5
Muscle 216.7 1175 14.04 9.046 x 10
4
Skin 110.7 297 5.37 2.661 x 10
4
Other 139.7 584.9 18.58 1.623 x 10
5
SC site 0.1107 0.297 0.00537 26.6
Organ/tissue Lymph flow (ml/day)
b
L
Organ
FcRn (nM)
c
CFcRn Vascular IgG clearance via pinocytosis (ml/day)
d
CL
Organ
pino
Endosomal volume (ml)
d
V
Organ
endo
Lung 1066 1.483 x 10
6
0.3978 2.990 x 10
− 3
Heart 42.57 1.483 x 10
6
0.3153 2.370 x 10
− 3
Liver 260.4 1.483 x 10
6
2.084 1.566 x 10
− 2
Spleen 31.93 1.483 x 10
6
0.0663 4.983 x 10
− 4
GI Tract 159.6 1.483 x 10
6
2.66 1.999 x 10
− 2
Kidney 202.2 1.483 x 10
6
0.3042 2.286 x 10
− 3
Muscle 180.9 1.483 x 10
6
36.47 0.2741
Skin 53.21 1.483 x 10
6
7.51 0.05644
Other 324.6 1.483 x 10
6
13.51 0.1015
SC site 0.05321 1.483 x 10
6
0.003755 2.822 x 10
− 5
a
Based on previous literature (see text)
b
Fixed to be 2% of organ plasma flow
c
Estimated (see text)
d
Calculated (see text)
volume fractions reported previously [160]. The total plasma volume was calculated as 108.0 ml
based on a body weight (BW) of 3.8 kg, the weight-normalized blood volume of 47.7 ml/kg (V
blood
)
[160] and an hematocrit of 40.4% [161]: V
Total
Plasma
=BW · V
blood
· (1− HCT%). The venous plasma
volume was then computed to be 46.4 ml using the sum of plasma volumes of all organs and the
total plasma volume (V
Total
Plasma
).
The total cardiac output of a 3.8-kg reference monkey was fixed at 0.62 L/min [162, 163], with
the total plasma flow ( Q
Total
= 5.33 x 10
5
ml/day, Table 5.3) distributed to different organs in
the same proportions as reported in humans (Table 5.2) [158]. The lymph flow of each organ was
assumed to be 2% of the plasma flow [52] as listed in Table 5.2. The volume and flow rate of lymph
node and lymphatic vessel were calculated via allometric scaling (K
LymNode
= 68.9 · BW
− 0.25
,
V
LymVes
=0.163· BW, Q
LymVes
=165· BW
0.75
) and the resulting values are listed in Table 5.3.
Based on reported radiolabelled IgG studies in humans [130, 138], the volume of a SC injection
site depot (assumed as a sphere) for antibodies in humans was calculated to be 2 ml. The volume
of a SC depot in cynomolgus monkeys was scaled to 0.11 ml (V
SC
inter
= 0.0291 · BW), which was
set as the interstitial volume of the SC site (V
SC
inter
) in this model. By treating the SC site as a
proportion (0.1%) of skin, the vascular space volume (V
SC
vasc
), total volume (V
SC
Total
) and plasma flow
94
Table 5.3: Fixed parameter values for central venous and lymphatic systems (see Materials and
Methods)
Parameter (unit) Value Description Ref.
Central venous
Q
Total
(ml/day) 5.332 x 10
5
Plasma flow [161–163]
V
Plasma
(ml) 46.35 Central plasma volume (see text)
IgG0 (nmol/day) 1793 Production rate of endogenous IgG [102, 164]
Central lymphatic
k
LymNode
(day
− 1
) 49.37 Rate constant of mAb leaving lymph node (see text)
V
LymVes
(ml) 0.6161 Volume of lymphatic vessel compartment (see text)
Q
LymVes
(ml/day) 449.9 Lymph flow rate from lymphatic vessel compartment (see text)
(Q
SC
) of the SC site were calculated proportionally (Table 5.2). The volume of local lymphatic
capillary was obtained via allometrically scaled (V =2.06· 10
− 4
· BW) as 0.000782 ml.
Given the steady-state concentration of endogenous IgG of 11.3 mg/ml in cynomolgus monkeys
[164], and assuming a half-life of 8.3 days and a distribution volume of 75 ml/kg [102], the pro-
duction rate of endogenous IgG (IgG0) was calculated as 1792 nmol/day (Table 5.3). The whole-
body pinocytosis clearance (CL
Total
pino
) was fixed as 63.3 ml/day, which is based on the reported
weight-normalized clearance of IgG [102] and is consistent with another report [165]. Individual
organ pinocytosis clearances (Table 5.2) were then determined by proportioning the whole-body
pinocytosisclearancebasedonorganvolumesasfollows: CL
Organ
pino
=(V
Organ
Total
/V
Total
)· CL
Total
pino
. The
endosomalvolumeofeachorgan(Table5.2)wasthencalculatedasfollows: V
Organ
endo
=2· CL
Organ
pino
· τ .
Previous experiments have indicated that endothelial cells contribute approximately equally
to antibody clearance as do hematopoietic cells in interstitial space and draining lymphatics [24],
Accordingly, the clearance of the SC space as calculated above was partitioned equally into the
two processes CL
SC
pino
and CL
SC
(Figure 5.1a). The later represents SC pre-systemic degradation
of antibodies via hematopoietic cells, proteolytic enzymes, and cell-membrane or soluble targets.
5.2.4.2 Transcapillary transport and antibody-FcRn binding affinities
Parameters related to the transcapillary transport (Table 5.4) were calculated using the two-pore
theory model [19, 57], and depend on antibody molecular weights, or organ lymph flows, or both.
95
Table 5.4: Parameters values related to endosomal processing and two-pore transport of mAbs
Parameter (unit) Value Description Ref.
All organs, tissues and mAbs
τ (day) 0.00375 Transit time of IgG between endosomal sub-compartments [51]
fIgGrecyc 0.715 IgG recycling fraction of IgG-FcRn complex back to vascular space [50]
k
6.0
on
(nM
− 1
day
− 1
) 1.96 Association rate constant between IgG and FcRn at pH = 6.0 [51]
k
6.0
off
(day
− 1
) calculated Dissociation rate constant between IgG and FcRn at pH = 6.0 [51]
k
7.4
on
(nM
− 1
day
− 1
) 45.08 Association rate constant between IgG and FcRn at pH = 7.4 [51]
k
7.4
off
(day
− 1
) calculated Dissociation rate constant between IgG and FcRn at pH = 7.4 [51]
rS (nm) 4.44 Small pore radius [19]
rL (nm) 22.85 Large pore radius [19]
α S 0.958 Fractional hydraulic conductance of small pores [19]
α L 0.042 Fractional hydraulic conductance of large pores [19]
Adjusted for mAb MW (150 kDa IgG as an example); All organs and tissues
ae (nm) 4.81 Stokes Einstein radius [19]
σ S 0.998 Small pore vascular reflection coefficient [19]
σ L 0.18 Large pore vascular reflection coefficient [19]
A/A0S 9.280 x 10-7 Fractional accessible pore size of small pore [19]
A/A0L 0.349 Fractional accessible pore size of large pore [19]
PeS 9.82 Peclet number of small pores [19]
PeL 4.48 Peclet number of large pores [19]
Adjusted for mAb MW and individual organs and tissues (150 kDa IgG in kidney as an example)
PS
Kidney
S
(ml/day) 0.02502 Permeability-surface area product of small pores [19]
PS
Kidney
L
(ml/day) 15.57 Permeability-surface area product of large pores [19]
J
Kidney
iso
(ml/day) 76.84 Circular isogravimetric flow [19]
J
Kidney
S
(ml/day) 116.9 Lymph flow through small pores [19]
J
Kidney
L
(ml/day) 85.33 Lymph flow through large pores [19]
The equilibrium dissociation constant (K
d
) at different pH were obtained as follows. For en-
dogenous IgG and WT mAbs, K
6.0
d
and K
7.4
d
were fixed at 132 nM and 23000 nM, respectively
[165]. K
6.0
d
of 132 nM is consistent with the measurements from previous experiments [40]. For
the antibody variants with T250Q/M428L mutation, the K
6.0
d
was fixed at 6.2 nM, which is the
mean value of these five variants based on an surface plasmon resonance kinetic analysis (range
3 to 13 nM) [40]. The K
7.4
d
of the variants were then calculated using the log-linear relationship
between K
d
and pH reported previously [166]: log
10
(K
7.4
d
)=1.8· pH − 10. This equation was used
to derive the K
7.4
d
as 2053 nM. The k
6.0
on
and k
7.4
on
of endogenous IgG and WT mAbs were fixed at
1.96 day
− 1
nM
− 1
and 45.1 day
− 1
nM
− 1
respectively [165]. The same k
on
values were used for the
antibody variants, under the assumption that changes in K
d
are attributed mostly to differences in
dissociationrates[40]. Then,k
off
ataspecificpHwasderivedusingtheequation: K
off
=K
on
· K
d
.
5.2.5 Model parameter values – estimated
All model simulations and parameter estimations were conducted using the ADAPT software (ver-
sion 5) [100]. The lysosomal degradation rate constant (K
e
) and FcRn initial concentrations were
determined based on previous studies [102, 167, 168]. First, the endogenous IgG concentration
without FcRn protection was fixed as 3.39 mg/ml, which is 30% of the IgG baseline concentration
[168]. This value was used as the steady state concentration of endogenous IgG in the complete
96
PBPK model, resulting in an estimated = 58 day
− 1
in the absence of FcRn. This value results
in approximately 95% free IgG within late endosome being degraded, which is consistent with a
previous report [73].
After fixing K
e
, the nominal concentration of endogenous IgG (11.3 mg/ml) was then used as
the steady state condition in our PBPK model to estimate the initial FcRn concentration within
the endosomal compartments as 1.48 x 10
6
nM.
Foreachantibodyandforeachanimal,twoscalefactors(S
pino
,S
diff− conv
)wereestimatedusing
the IV concentration-time measurements, via maximum likelihood estimation (additive plus pro-
portional error variance model) using the ID application in ADAPT. Then two SC site parameters
(S
SC
LymUpt
, CL
SC
) were estimated for each antibody using SC concentration-time measurements,
with S
pino
and S
diff− conv
for each antibody fixed at their median values obtained from the IV
analysis. The model was allowed to reach steady state of endogenous IgG before the antibody IV
or SC administration.
The estimates of these four antibody specific parameters ( S
pino
, S
diff− conv
, S
SC
LymUpt
, CL
SC
)
were used to calculate model-predicted SC bioavailability (F) as follows. Simulations of IV PK
and SC PK were conducted after 1 mg/kg dose administration, followed by calculating areas under
curves (AUC) of each monkey. The ratio between the AUC after a SC dose and the AUC after an
IV dose was calculated as the predicted F.
5.2.6 Biophysical determinants of antibody specific model parameters
Structure-based physiochemical properties from TAP and MOE were explored as potential predic-
tors of each of the four antibody specific model parameters ( S
pino
, S
diff− conv
, CL
SC
, S
SC
LymUpt
).
Univariate and multivariate linear associations were examined using the “lm” (linear model) func-
tion in the “stats” package in R [103]. The selection of significant physiochemical properties was
basedonpvalues. The“Proteinpatches”functionofMOEwasusedtovisualizethesurfacepatches
of positive charge, negative charge and hydrophobicity of antibodies [66].
97
5.2.7 Model simulations
The PBPK model was used to simulate the impact of FcRn blockade (k
6.0
on
and k
7.4
on
of IgG were
set to zero) on endogenous IgG concentration-time profiles, with comparison to results from pre-
vious studies following IV administration of FcRn antagonists in cynomolgus monkeys [167, 168].
Model simulations were also used to calculate the relative contributions of SC absorption pathways
(lymphatic and blood) to systemic antibody delivery, to determine the extent of pre-systemic SC
site antibody degradation attributable to interstitial (CL
SC
) versus endothelial lysosomal (K
e
)
degradation, and to evaluate the effects of local lymph flow ( L
SC
) on SC PK and bioavailability.
5.3 Results
5.3.1 Intravenous pharmacokinetics
The antibody serum concentration-time measurements in each monkey were used to estimate the
model parameters (S
pino
and S
diff− conv
) that characterize the antibody specific differences in ex-
travasation. Representative model fits are shown in Figure 5.2a for each of the five WT and five
variant antibodies, with the complete results for all animals provided in the supplementary mate-
rial. Overall, the model provides a good description of the observed PK data from the different
antibodies following IV administration (R > 0.91). Table 5.5 (columns two and three) lists the me-
dians and standard deviations of S
pino
and S
diff− conv
obtained for each antibody from all animals,
while parameter estimates of the all individual animals along with their relative standard error
(RSE%) are provided in the supplementary material. These two parameters were estimated with
good precision in 29 of the 30 animals (RSE% < 25%), with the exception of one monkey admin-
istered the B variant that resulted in an RSE% for S
diff− conv
=95%. The median values of S
pino
estimates range from 0.792 to 3.033 and those of S
diff− conv
from 0.331 to 1.023. S
pino
estimates
are significantly higher in T250Q/M428L variants compared to their wildtype counterparts for the
A, B and C antibodies. But the difference in S
pino
between WT and variants was not significant
for the D and E antibodies (Table 5.5). No obvious trend is observed in S
diff− conv
between the
WT mAbs and their variants (Table 5.5).
98
Figure 5.2: Model predicted (solid lines) and observed (blank dots) antibody concentration versus
time profiles. (a) – IV results: top row shows results from each of the five WT antibodies from
one representative monkey; second row shows results from each variant in a selected animal. (b)
– SC results: top row shows results from each of the five WT antibodies from one representative
monkey; second row shows results from each variant in a selected animal. The title line for each
graph provides the animal ID as a reference to the complete results in all animals presented in the
supplemental material
99
Table 5.5: Median values (n=3) and standard deviations across animals of the antibody specific
parameters and model predicted bioavailability (F, %) for each antibody
Antibody S
pino
S
diff− conv
CL
SC
(ml/day) S
SC
LymUpt
F (%)
A WT 1.03 (0.257) 0.921 (0.144) 0.0155 (0.00404) 2.20 (0.420) 87.0 (4.29)
B WT 0.792 (0.370) 0.356 (0.320) 0.0282 (0.0235) 0.863 (0.409) 61.5 (8.52)
C WT 1.47 (0.123) 0.577 (0.0673) 0.00503 (0.00444) 0.622 (0.534) 78.8 (11.5)
D WT 2.19 (0.661) 0.550 (0.059) 0.134 (0.0853) 1.27 (1.30) 45.0 (7.72)
E WT 3.03 (2.04) 0.977 (0.342) 0.0889 (0.0685) 0.685 (0.788) 38.5 (9.94)
A Var 1.27 (0.211) 0.526 (0.149) 0.0109 (0.0121) 0.522 (0.193) 51.6 (12.9)
B Var 0.951 (0.181) 0.331 (0.222) 0.0198 (0.00850) 0.594 (0.236) 62.0 (4.73)
C Var 1.94 (0.293) 1.02 (0.104) 0.00599 (0.00185) 0.568 (0.153) 68.8 (6.29)
D Var 2.00 (0.754) 0.960 (0.254) 0.0497 (0.00577) 0.956 (0.285) 48.7 (4.83)
E Var 1.73 (0.536) 0.694 (0.323) 0.0414 (0.0203) 0.622 (0.258) 45.7 (5.38)
S
pino
: scale factor for antibody specific endothelial pinocytosis uptake rate
S
diff− conv
: scale factor for antibody specific paracellular transport rate
CL
SC
: pre-systemic degradation of antibodies within SC interstitial space
S
SC
LymUpt
: scale factor for antibody specific lymphatic uptake from SC injection site into local
lymphatic capillary
F (%): model-predicted bioavailability for each antibody
5.3.2 Subcutaneous pharmacokinetics
The SC site antibody specific parameters CL
SC
and S
SC
LymUpt
were estimated in each animal after
fixing and to their median values for the corresponding antibody as listed in Table 5.5. Figure 5.2b
displays representative fits for each of the WT and variant antibodies following SC administration.
The complete fits of SC PK data for all monkeys are provided in the supplementary material and
suggest adequate description of observed SC PK data (R > 0.84). The median values and standard
deviations of and are listed in Table 5.5 (columns four and five), with the parameter estimates for
each animal and their RSE% given in the supplementary material. These two SC site parameters
were estimated with reasonable precision (RSE% < 50%), except one monkey administered the
C WT antibody resulted in an RSE% for CL
SC
= 174%. Median values of CL
SC
vary over a
wide range from 0.00503 to 0.134 ml/day, while S
SC
LymUpt
are in the range of 0.522 to 2.20. No
significant change of CL
SC
and S
SC
LymUpt
was observed when comparing the WT antibodies versus
their variants, although values of both parameters are generally lower in variants. Table 5.5 also
shows the medians and standard deviations of the resulting calculated values of F, which range
100
from 39% (E WT) to 87% (A WT). No consistent trend is observed when comparing the predicted
F of WT antibodies to their variants (see Discussion).
5.3.3 Biophysical determinants of antibody specific model parameters
Associations between the estimated antibody specific model parameters ( S
pino
, S
diff− conv
, CL
SC
and S
SC
LymUpt
) and the 26 structure-based physiochemical properties were explored using linear re-
gression. Analysis using each property individually, identified several significant relationships. The
parameter S
pino
was found to positively correlate with a metric of area of positive protein patches
(patch pos, R = 0.59, p = 0.00063) as shown in Figure 5.3a. The addition of other individu-
ally significant physiochemical properties (patch hyd, R = 0.58, p = 0.0007; pI seq, R = 0.58,
p = 0.0008) in a multiple regression with patch pos was not found to be collectively significant
(p > 0.05). Thus, patch pos alone was found to be the best predictor of S
pino
in our analysis.
The parameter S
diff− conv
was found to be negatively correlated with the area of hydrophobic
protein patch (patch hyd, R = 0.54, p = 0.002) as shown in Figure 5.3b. In a multiple regres-
sion analysis, both patch hyd and patch pos were found to be significant predictors of S
diff− conv
:
S
diff− conv
= − 0.0028 · patch hyd − 0.0032 · patch pos+5.47 (R = 0.63, p = 0.00096). The SC
absorption site parameter CL
SC
was found to be positively correlated with a metric of area of pos-
itive protein patches near CDRs (patch cdr pos, R = 0.61, p = 0.00037) as shown in Figure 5.3c.
No additional physiochemical property was found to be significant when added to patch cdr pos
in multiple regression (p > 0.05), nor was S
SC
LymUpt
found to be related to any of the antibody
specific physicochemical properties. When visualizing the surface patches of two antibodies with
large differences in both S
pino
and S
SC
LymUpt
(A WT and D WT), larger positive charge patches
(blue area) can be observed in D WT (Figure 5.3e) than those of A WT (Figure 5.3d).
5.3.4 Model simulations
Usingthedevelopedmodelincorporatingtheidentifiedrelationshipbetween CL
SC
andpatch cdr pos
(Figure 5.3c), population simulations (n=1000) were conducted to predict the expected distribu-
tion of F versus patch cdr pos (SIM application in ADAPT). Simulated SC PK for a hypothetical
101
Figure 5.3: Differences in antibody specific parameters can be partly explained by their structure-
basedmetrics. (a)positivecorrelationbetweenS
pino
andpatch pos: S
pino
=0.0096· patch pos− 0.34
(R = 0.59, p = 0.00063, residual standard error = 0.83). (b) negative correlation between
S
diff− conv
and patch hyd: S
diff− conv
= − 0.0013 · patch hyd + 2.58 (R = 0.54, p = 0.002,
residual standard error = 0.24). (c) positive association between CL
SC
and patch cdr pos:
CL
SC
= 0.00080· patch cdr pos− 0.021(R = 0.61, p = 0.00037, residual standard error = 0.040).
Black dots are for wild type antibodies, while grey dots are for T250Q/M428L variants. In (a), (b),
and (c), the shaded areas represent 95% confidence intervals of the regression lines. Top-oriented
surface patch visualization (MOE (23)) of (d) A WT and (e) D WT Fv domains of one arm at
pH 7.4. Blue patches indicate positive charge, red negative charge and green patches represent
hydrophobic areas
102
antibody with patch cdr pos = 40 (black) and patch cdr pos = 160 (grey) are displayed in Figure
5.4a, indicating significant differences caused by local CDR positive charge. Figure 5.4b shows
the expected distribution of F in the population for these two hypothetical antibodies. The pre-
dicted relationship between F and patch cdr pos is depicted in Figure 5.4c, and shows that as
patch cdr pos increases from 30 to 240, median values of the predicted bioavailability decrease
from 91.6% to 24.5% (5-95th percentile ranges of the simulated values also shown). In these simu-
lations, the other antibody specific parameters were set at fixed values: S
pino
= 1.78, S
diff− conv
=
0.67, S
SC
LymUpt
= 1.02.
Model simulation indicates that following FcRn blockade, endogenous IgG concentration drops
to about 30% of its basal concentration within 16 days. This time course and change are consistent
with the concentration-time courses from previous experiments conducted in cynomolgus monkeys
[167,168],inwhichtheIVadministrationofFcRnantagonistsledtoreductionofIgGconcentration
to 20 to 30% of the pre-administration values.
Using median values of four antibody specific parameters of each antibody, model simulations
indicate that the SC site lymph flow contributes 93.5% to 98.8% of total (lymph flow plus blood
perfusion) absorbed amount of these five WT mAbs and their variants. This primary contribution
of lymphatic drainage to SC absorption is in line with previous studies [23, 140]. Among the 10
mAbs evaluated in this study, model simulations also show that the degradation in the SC site
interstitial space (represented by CL
SC
) contributes between 51.0% (C Var) to 96.1% (D WT) of
thetotalpre-systemicdegradation(CL
SC
plusSCsitelysosomaldegradation). Thisresultsuggests
that the degradation in the SC site interstitial space is a major determinant of antibody SC F.
A sensitivity analysis of the SC site lymph flow ( L
SC
) reveals significant effects of local lymph
flow on SC PK profiles (Figure 5.4d). Using the A WT antibody as an example, we find that a
5-fold increase in L
SC
results in an increased F from 87% to 97%, while decreasing L
SC
to 0.5-
and 0.1-fold will reduce the F to 77% and 42%, respectively. This significant impact of the SC site
lymphflowonFis consistentwithareviewofclinicalstudies, inwhichtheauthorssummarizethat
higher SC bioavailability occurs at injection sites with faster average lymph flows (4.2, 2.8, and 2.0
cm/min for thigh, anterior trunk, and upper arm, respectively) [169].
103
Figure 5.4: Simulations illustrating the use of the model to predict bioavailability based on physic-
ochemical properties. (a) Concentration-time profiles from 1000 simulations following a 1 mg/kg
SC dose for a hypothetical antibody with patch cdr pos = 40 (black) and one with patch cdr pos
= 160 (grey), respectively. (b) Distribution of the bioavailability calculated from the simulation
of the two antibodies: patch cdr pos = 40 (black) and patch cdr pos = 160 (grey), respectively.
(c) Model-predicted bioavailability versus patch cdr pos. The dots are medians of bioavailability
from 1000 population simulations at a specific patch cdr pos. The bars represent 5th and 95th
percentiles of the simulated values. (d) Sensitivity analysis of SC site lymph flow ( L
SC
) using A
WT in monkey ID = I02877-SC as an example. The solid line represents the original SC PK profile
of A WT PK. The dot dashed, dotted and dashed lines represent SC PK profiles after updating
SC site lymph flow to 5-, 0.5- and 0.1-fold of the original value, respectively
104
5.4 Discussion
An incomplete understanding of the factors that determine antibody subcutaneous absorption
(e.g., physiological, physiochemical, formulation), and how they might interact, has limited the
development of reliable in vitro methodologies and animal models to predict antibody SC PK and
bioavailability in humans [14]. In this work, a predictive PBPK modeling framework is developed
thatincorporatesstructure-derivedmetricscharacterizingantibodysize,overallcharge,localcharge
and hydrophobicity, using results from mAb IV and SC PK studies in cynomolgus monkeys. The
premise of the model-based approached is that antibody physiochemical properties can be better
related to underlying antibody transport mechanisms (e.g., SC site lymphatic uptake and pre-
systemic degradation, paracellular exchange, nonspecific binding), then directly to measures of
systemic PK (clearance, bioavailability) using empirical relationships. The proposed PBPK model
framework also allows for prediction of a complete antibody concentration-time course in plasma
as well as in tissues of interest. Moreover, given the mechanistic basis of the model, the effects of
pathophysiological changes at the injection site on antibody absorption and PK can be explored.
Pre-systemic degradation of antibodies in the interstitial space and the local lymphatic vessels
within the SC injection site is represented in the model by a composite first-order clearance process
(CL
SC
in Figure 5.1). Based on our analysis of 26 structure-based physiochemical metrics, CL
SC
shows the most significant positive association with a metric describing local positive charge area
near CDRs (patch cdr pos, R = 0.61, p = 0.00037, Figure 5.3c). This association can be explained
in terms of charge-mediated interactions: more positively charged mAbs have stronger binding
withSCtissuecomponents[36]and/ortargetswithnegativecharges,leadtoincreasedpre-systemic
degradation[17]. BecausehigherCL
SC
yieldslowerbioavailability,thispositiveassociationbetween
CL
SC
and patch cdr pos is in line with previous studies in cynomolgus monkeys and rats in which
mAbswithincreasedpositivechargepatchesinCDRsorFvregionsshowedadecreasedextentofSC
absorption [36, 152]. This positive correlation is also consistent with our previous work in humans
(discussed below) that found a positive relationship between pre-systemic antibody clearance in a
PBPK model and a metric of patches of positive charge near CDRs [74].
105
In the model, antibody uptake in the local SC lymphatic vessels was assumed to be antibody
specific, represented by scaling ( S
SC
LymUpt
) the lymphatic transport rate that is governed by lym-
phaticflowandantibodyinterstitialconcentration. Forthe10antibodiesconsidered, theestimated
values of S
SC
LymUpt
vary over a four-fold range (0.522-2.20). The transport of mAb though extra-
cellular matrix is affected by electrostatic interactions and steric exclusion [74], thus it would be
expected that S
SC
LymUpt
might correlate with metrics of charge due to antibody-SC tissue interac-
tions [36]. However, no significant correlation was found in this study, possible due to the influence
of steric exclusion or the limited number of antibodies evaluated.
Differences in extravasation among antibodies due to non-specific interactions and pinocytosis
uptake, reflected by the model scale factor S
pino
, show significant association with the area of
positive patch metric (R = 0.59, p = 0.00063, Figure 5.3a). This positive association is likely due
to charged-mediated antibody-cell membrane interactions (mainly negatively charged) leading to
greater antibody uptake [34, 170]. Other metrics related to hydrophobicity, isoelectric point and
positive charge in the CDR region, also show significant correlation with S
pino
individually (see
supplementarymaterial)butarenotsignificantwhencombinedwithpatch pos(p>0.05),likelydue
to correlations between these metrics. These mechanism-based modeling results are qualitatively
consistent with previous studies reporting empirical relationships between charge properties and
antibody clearance [72, 153].
We found that differences in antibody transcapillary transport via diffusion and convection,
represented by the model scale factor S
diff− conv
, could be explained in part by area of hydrophobic
patches and patch pos (S
diff− conv
was negatively correlated with both). Paracellular convection of
antibodies depends on vascular-tissue hydrostatic gradient and sieving coefficient, with the latter
determined by pore physiology, along with antibody size, shape and charge [10]. With higher posi-
tivecharge,antibodiesshouldbindmorestronglywiththecellmembrane,whichmayslowantibody
transcapillarytransportasreflectedbysmallervaluesof S
diff− conv
. Moreover, increasedhydropho-
bic interactions of antibodies can lead to increased aggregation potential [36], which can alter their
size and shape and hinder their transcapillary transport as represented by lower S
diff− conv
.
106
The results in Figure 5.4 illustrate how the composite model, including the identified relations
between antibody dependent model scale factors and associated physicochemical properties, can
be used to predict bioavailability based on in silico metrics. Figure 5.4c shows a 2.5-fold drop
in the predicted bioavailability over the range of the positive patch metric values (40-160) for the
antibodies used in this work. We note, however, that this result uses fixed values for the other scale
factors: S
pino
, S
diff− conv
, S
SC
LymUpt
. A more complete prediction would need to include antibody
specific values for these other scale factors reflecting the correlations between the in silico derived
antibody metrics that are unavailable based on the small number of antibodies used in this study.
Efforts to improve bioavailability of subcutaneously administered monoclonal antibodies by
engineering FcRn binding have shown mixed results. Previous studies in mice reported a reduced
bioavailability after eliminating FcRn-antibody interaction [24, 42]. However, no improvement in
bioavailability was observed in two studies in cynomolgus monkeys after Fc engineering to yield
higher affinity to FcRn [43, 44]. In this study, we did not see a consistent change in predicted F
between WT antibodies and their variants (Table 5.5). In our model, both CL
SC
and S
SC
LymUpt
determine F. Although both CL
SC
and S
SC
LymUpt
are generally lower in variants compared to their
WT,thisdifferenceisnotsignificant. FurtheranalysisofthesetwoSCsiteparametersinadditional
monkeys may help reduce the influence of inter-animal variability on the ability to detect any
differences between WT and variants.
In a previous modeling study [74] using human mAb SC PK data from 12 human studies
reported in the literature, we found a relationship between pre-systemic antibody degradation at
the SC injection site and an in silico derived metric representing patched of positive charge near
the antibody CDR. This previous work, however, was subject to a number of limitations. The
set of 12 antibodies used in the study involved a selection bias in that all mAbs were approved
or in clinical trials, and thus might not reflect an appropriately broad range of physical chemical
characteristics. Also, the antibodies were studied in 12 separate trials involving subjects/patients
with different ages, weight, sex, physiological states, and antibody injection sites. Moreover, only
mean plasma concentration-time data were available from the publications, thus the effects of
interindividual variability could not be assessed. The work reported herein addresses some of these
107
limitations. Between antibody experimental variability was controlled to an extent, given that a
mixtureformulationoffiveWTorvariantantibodieswasadministeredtoeachanimal. Inaddition,
the WT antibodies were engineered to a variety of targets leading to antibodies with a range of
physicochemical properties. Also, replicate experiments allowed the modeling analysis to reflect
the inter-animal variability in predicted results.
Although statistically significant correlations were found between antibody specific model pa-
rametersandseveral in silico structure-basedmetrics, notablevariabilityoftheregressionlinescan
be observed (Figure 5.3), and some study limitations and modeling assumptions warrant further
consideration. The work was limited to a relatively small number of antibodies resulting in a lim-
ited number of distinct values of the physicochemical properties. Also, the plasma concentrations
showed some considerable between animal variability for several of the antibodies. These limita-
tions can be addressed in future studies with additional antibodies and a larger number of animals
per group. Such expanded studies would also provide for external model validation, which could
not be pursued in the current work. We have detailed the assumptions that underly all aspects of
the model, including those that relate to the subcutaneous site model. In the latter, pre-systemic
degradation of mAbs in the SC interstitial space and the local lymphatic vessels is represented by a
composite first-order clearance ( CL
SC
). This composite clearance could be portioned into its two
components, to further explore the role of antibody physicochemical characteristics on these two
mechanisms. Finally, additional antibodies may reveal other than the linear relationships between
model parameters and physiochemical metrics reported in our results.
An integrated physiologically based modeling, incorporating in silico structure-derived anti-
body properties, is presented for predicting pharmacokinetics and bioavailability of subcutaneously
administered monoclonal antibodies in cynomolgus monkeys. The model provides a framework for
predicting bioavailability and pharmacokinetics of an antibody based on its local positive charge
in the complementarity-determining region, overall charge, and hydrophobicity. Given the model’s
mechanistic basis, it can also be used to obtain insights about the rate and extent of antibody SC
absorption and, potentially, to predict peripheral tissue concentrations.
108
6 Summary and future work
In this dissertation research, we developed and evaluated a comprehensive physiologically-based
pharmacokinetic model platform for predicting the disposition and absorption of mAbs following
intravenousandsubcutaneousadministrationthatincorporatesphysiochemicalantibodyproperties
derived from in vitro assays and in silico structure-based metrics. The model includes the funda-
mental mechanisms of transport and processing that govern antibody disposition and absorption,
including pinocytosis uptake, endosomal transit and processing, and paracellular transport and
others. Relevant antibody specific properties such as molecular weight, non-specific interaction,
overall charge, local charge, hydrophobicity, FcRn binding and others are integrated in the model.
Current and past approaches for predicted mAb PK in humans using antibody physiochemical
properties, have focused on relating these properties directly to clinical pharmacology end points
such as drug clearance and bioavailability, via empirical relationships developed using traditional
statistical methods [68] or via machine learning tools [72], have shown little generalizable success.
Also, inter-species scaling of PK from animal models to humans, which has been applied with high
success for small molecule drugs, has been considerably less successful for protein drugs, especially
after SC administration. The underlying premise of our approach is that antibody physiochemical
properties are best related to the associated physiological mechanisms (e.g., paracellular exchange,
nonspecific binding, SC site lymphatic uptake and pre-systemic degradation) that are responsible
fortheirtransportandprocessing. TheproposedPBPKmodelframeworkalsoallowsforprediction
of the complete antibody concentration-time course in plasma and in tissues of interest, in addition
to the clinical pharmacology end points of drug clearance and bioavailability. Moreover, the PBPK
model allows us to gain mechanistic insights into the factors that determine mAb disposition and
109
absorption and how pathophysiological factors influence antibody pharmacokinetics. The major
findings, limitations and future work of this dissertation are summarized below.
6.1 Major findings
We have introduced a PBPK model that incorporates the fundamental mechanisms governing
the absorption of SC administered mAbs. The resulting model describes plasma PK and SC
bioavailability of 12 mAbs in humans (approved or in clinical trials) and 10 mAbs in cynomolgus
monkeys(wildtypeandFcengineered),byaccountingfordifferencesinspecifictransportprocesses.
Basedonmodelsimulations, weconfirmedthatthelymphaticpathwayisresponsibleformorethan
90 % of total SC absorption of antibodies in both humans and monkeys. Sensitivity analysis using
the developed model shows that increasing the local lymph flow significantly improves SC PK and
increases bioavailability.
We extended a previously reported PBPK model and applied it to describe the PK of IV ad-
ministered antibodies in humans and monkeys. The updated model incorporates new experimental
insights related to endosomal transit and processing. The model also includes the contribution of
additional organs/tissues to antibody disposition with updated physiological parameters. In addi-
tion to antibody molecular weight and FcRn-IgG binding affinity, the model includes two antibody
specific parameters for extravasation to describe antibody differences in disposition.
We established that antibody physiochemical properties derived from in vitro studies and in
silico methods can inform underlying antibody transport processes, which in turn can be used
to make model-based predictions of antibody PK. An in vitro assay of heparin chromatography
relative retention time (Hep RT) was found to associate with both antibody pinocytosis uptake
and paracellular transport. Several in silico metrics of charge and hydrophobicity are related to
antibody pre-systemic degradation, pinocytosis uptake and paracellular transport. Also, Hep RT
was found to be related to in silico metrics of charge and hydrophobicity, indicating that in vitro
assays may reflect more composite information. Results from both in vitro assays and in silico
metrics can be used for model based PK predictions.
110
6.2 Limitations and future work
Although statistically significant correlations were found between antibody specific model param-
eters and several physiochemical properties, notable variability in predictions can be observed (see
Chapters 3, 4 and 5). Several limitations related to study designs and modeling analysis warrant
further consideration and could be the subject of future work.
TheanalysisinChapter3andChapter4waslimitedtoPKdataofmAbsapprovedorinclinical
trials, and thus might not reflect the broad range of physiochemical characteristics that would be
encountered in the early drug candidate selection process. Also, the antibodies were studied in
separate trials involving subjects/patients with different ages, weight, sex, physiological states and
antibodyinjectionsites,whichcouldcontributetodifferencesinantibodyPK.Moreover,onlymean
plasma concentration time data were available from the publications and used in our analysis, thus
the effects of inter-individual variability could not be assessed. Further work incorporating a larger
numberofantibodiesalongwithindividualsubjectplasmaconcentration-timemeasurementshould
be considered.
The study in Chapter 5 was intended to address some of these limitations. In this work,
the antibodies used were engineered to a variety of targets leading to antibodies with a range of
physiochemical properties. In addition, mixture doses of multiple antibodies were used to control
for some between experiment variability. Also, replicate experiments allowed the modeling analysis
to reflect the inter-animal variability in predicted results. However, the work in Chapter 5 was
still limited to a small number of antibodies resulting in a limited number of distinct values of the
physiochemical properties.
Because of the relatively small number of antibodies with complete pharmacokinetic data avail-
able, we could not conduct any formal external validation of the predictive ability of the proposed
modeling framework, although internal and other validation of the PBPK model was conducted as
detail in Chapters 3-5. Results from ongoing clinical trials with other antibodies can be used as a
basis for external model validation. Although the PBPK model can predict tissue drug concentra-
tions, only plasma concentration-time data were available in our analyses, thus we could not assess
the model’s ability to predict tissue concentrations.
111
We have detailed the assumptions that underlies all aspects of the model in Chapters 3-5, in-
cluding those that relate to the subcutaneous site model (Chapter 4, 5). In the latter, pre-systemic
degradation of mAbs in the SC interstitial space and the local lymphatic vessels is represented by
a composite first-order clearance ( CL
SC
), which warrants further consideration. This composite
clearance could be portioned into its two components (SC site and lymphatic system), to further
explore the role of antibody physiochemical characteristics on these two mechanisms. Finally, we
have used linear relationships to relate antibody physiochemical properties to model transport pro-
cesses. Results from additional antibodies may provide insights into more mechanistic approaches
to couple antibody properties and transport processes.
The model can be applied to explore other factors that influence mAb ADME processes. For
example, target binding was not incorporated in the model, which can be added given available
information about target turnover and binding processes. The effects of immunogenicity on anti-
body clearance can also be investigated. To overcome the limitations on SC dose/volume of large
molecules (such as mAbs) caused by the hypodermis ECM, hyaluronidase has been employed as a
permeation enhancer. Hyaluronidase acts by transiently degrading hyaluronan, a component of SC
ECM,thusreducingtheviscosityoftheECM.Recombinanthumanhyaluronidasehasbeenusedin
the clinic, which has allowed for increased SC infusion volumes and resulted in shortened absorp-
tiontimes. Theinfluenceofhyaluronidaseco-administrationinSCadministrationcanbedescribed
using antibody specific parameters at the SC site in our model. Quantitative relationships between
hyaluronidase dose and model parameters can guide future clinical studies using hyaluronidase to
facilitate SC absorption.
112
Bibliography
[1] P.Valent,B.Groner,U.Schumacher,G.Superti-Furga,M.Busslinger,R.Kralovics,C.Zielin-
ski, J. M. Penninger, D. Kerjaschki, G. Stingl, J. S. Smolen, R. Valenta, H. Lassmann,
H. Kovar, U. J¨ ager, G. Kornek, M. M¨ uller, and F. S¨ orgel, “Paul Ehrlich (1854-1915) and
His Contributions to the Foundation and Birth of Translational Medicine,” Journal of Innate
Immunity, vol. 8, no. 2, pp. 111–120, 2016.
[2] P. J. Carter and G. A. Lazar, “Next generation antibody drugs: Pursuit of the ’high-hanging
fruit’,” Nature Reviews Drug Discovery, vol. 17, no. 3, pp. 197–223, 2018.
[3] S. Ait-Oudhia, M. Ayse Ovacik, and D. E. Mager, “Systems pharmacology
and enhanced pharmacodynamic models for understanding antibody-based drug ac-
tion and toxicity,” mAbs, vol. 9, no. 1, pp. 15–28, 2017. [Online]. Available:
http://www.tandfonline.com/action/journalInformation?journalCode=kmab20
[4] A. C. Chan and P. J. Carter, “Therapeutic antibodies for autoimmunity and inflammation,”
Nature Reviews Immunology, vol. 10, no. 5, pp. 301–316, 2010.
[5] A. Datta-Mannan, “Mechanisms Influencing The Disposition Of Monoclonal Antibodies And
Peptides,” Drug Metabolism and Disposition, vol. 47, no. 10, pp. 1100–1110, 2019.
[6] H. Kaplon, A. Chenoweth, S. Crescioli, and J. M. Reichert, “Antibodies to
watch in 2022,” mAbs, vol. 14, no. 1, pp. 1–45, 2022. [Online]. Available:
https://doi.org/10.1080/19420862.2021.2014296
[7] A. L. Grilo and A. Mantalaris, “The Increasingly Human and Profitable Monoclonal
Antibody Market,” Trends in biotechnology, vol. 37, no. 1, pp. 9–16, 2019. [Online].
Available: http://www.ncbi.nlm.nih.gov/pubmed/29945725
[8] L. M. Weiner, R. Surana, and S. Wang, “Monoclonal antibodies: Versatile platforms for
cancer immunotherapy,” Nature Reviews Immunology, vol. 10, no. 5, pp. 317–327, 2010.
[9] E. A. Van Erp, W. Luytjes, G. Ferwerda, and P. B. Van Kasteren, “Fc-mediated antibody
effector functions during respiratory syncytial virus infection and disease,” Frontiers in Im-
munology, vol. 10, no. MAR, 2019.
[10] J. T. Ryman and B. Meibohm, “Pharmacokinetics of monoclonal antibodies,” CPT: Phar-
macometrics and Systems Pharmacology, vol. 6, no. 9, pp. 576–588, 2017.
113
[11] F. Kausar, K. Mustafa, G. Sweis, R. Sawaqed, K. Alawneh, R. Salloum, M. Badaracco, T. B.
Niewold, and N. J. Sweiss, “Ocrelizumab: A step forward in the evolution of B-cell therapy,”
Expert Opinion on Biological Therapy, vol. 9, no. 7, pp. 889–895, 2009.
[12] V.A.ThomasandJ.P.Balthasar,“UnderstandingInter-IndividualVariabilityinMonoclonal
Antibody Disposition,” Antibodies (Basel), vol. 8, no. 4, p. 56, 2019.
[13] G. Z. Ferl, F. P. Theil, and H. Wong, “Physiologically based pharmacokinetic models of
small molecules and therapeutic antibodies: A mini-review on fundamental concepts and
applications,” Biopharmaceutics and Drug Disposition, vol. 37, no. 2, pp. 75–92, 3 2016.
[14] M. S´ anchez-F´ elix, M. Burke, H. H. Chen, C. Patterson, and S. Mittal, “Predicting bioavail-
ability of monoclonal antibodies after subcutaneous administration: Open innovation chal-
lenge,” Advanced Drug Delivery Reviews, 2020.
[15] C.Bender,S.Eichling,L.Franzen,V.Herzog,L.M.Ickenstein,D.Jere,L.Nonis,G.Schwach,
P. Stoll, M. Venczel, and S. Zenk, “Evaluation of in vitro Tools to Predict the in vivo Ab-
sorption of Biopharmaceuticals Following Subcutaneous Administration,” Journal of Phar-
maceutical Sciences, vol. 000, 2022.
[16] H.WongandT.W.Chow, “PhysiologicallyBasedPharmacokineticModelingofTherapeutic
Proteins,” Journal of Pharmaceutical Sciences, vol. 106, no. 9, pp. 2270–2275, 2017. [Online].
Available: http://dx.doi.org/10.1016/j.xphs.2017.03.038
[17] M. Viola, J. Sequeira, R. Sei¸ ca, F. Veiga, J. Serra, A. C. Santos, and A. J. Ribeiro, “Sub-
cutaneous delivery of monoclonal antibodies: How do we get there?” Journal of Controlled
Release, vol. 286, pp. 301–314, 9 2018.
[18] G. Z. Ferl, A. M. Wu, and J. J. DiStefano, “A predictive model of therapeutic monoclonal
antibodydynamicsandregulationbytheneonatalFcreceptor(FcRn),” AnnalsofBiomedical
Engineering, vol. 33, no. 11, pp. 1640–1652, 2005.
[19] Z. Li and D. K. Shah, “Two-pore physiologically based pharmacokinetic model with de novo
derived parameters for predicting plasma PK of different size protein therapeutics,” Journal
of Pharmacokinetics and Pharmacodynamics, vol. 46, no. 3, pp. 305–318, 2019.
[20] E. S. Ward and R. J. Ober, “Targeting FcRn to Generate Antibody-Based Therapeutics,”
Trends in Pharmacological Sciences, vol. 39, no. 10, pp. 892–904, 2018.
[21] T. Li and J. P. Balthasar, “FcRn Expression in Wildtype Mice, Transgenic Mice, and in
Human Tissues,” Biomolecules, vol. 8, no. 4, p. 115, 2018.
[22] M. Ovacik and K. Lin, “Tutorial on Monoclonal Antibody Pharmacokinetics and Its Con-
siderations in Early Development,” Clinical and Translational Science, vol. 11, no. 6, pp.
540–552, 2018.
[23] W. F. Richter and B. Jacobsen, “Subcutaneous Absorption of Biotherapeutics: Knowns
and Unknowns,” Drug Metabolism and Disposition, vol. 42, pp. 1881–1889, 2014. [Online].
Available: http://dx.doi.org/10.1124/dmd.114.059238
114
[24] W. F. Richter, G. J. Christianson, N. Frances, H. P. Grimm, G. Proetzel, and D. C. Roope-
nian, “Hematopoietic cells as site of first-pass catabolism after subcutaneous dosing and
contributors to systemic clearance of a monoclonal antibody in mice,” mAbs, vol. 10, no. 5,
pp. 803–813, 2018.
[25] J. Tibbitts, D. Canter, R. Graff, A. Smith, and L. A. Khawli, “Key factors influencing
ADME properties of therapeutic proteins: A need for ADME characterization in drug
discovery and development,” mAbs, vol. 8, no. 2, pp. 229–245, 2016. [Online]. Available:
http://dx.doi.org/10.1080/19420862.2015.1115937
[26] A. V. Kamath, “Translational pharmacokinetics and pharmacodynamics of monoclonal anti-
bodies,” Drug Discovery Today: Technologies, vol. 21-22, no. Iv, pp. 75–83, 2016.
[27] J. E. Moore and C. D. Bertram, “Lymphatic System Flows,” Annual Review of Fluid Me-
chanics, vol. 50, no. 1, pp. 459–482, 2018.
[28] M. Dostalek, I. Gardner, B. M. Gurbaxani, R. H. Rose, and M. Chetty, “Pharmacokinet-
ics, pharmacodynamics and physiologically-based pharmacokinetic modelling of monoclonal
antibodies,” Clinical Pharmacokinetics, vol. 52, no. 2, pp. 83–124, 2013.
[29] W. F. Richter, S. G. Bhansali, and M. E. Morris, “Mechanistic Determinants of Biothera-
peutics Absorption Following SC Administration,” The AAPS Journal, vol. 14, no. 3, pp.
559–570, 2012.
[30] N. Ferrell, K. O. Cameron, J. J. Groszek, C. L. Hofmann, L. Li, R. A. Smith,
A. Bian, A. Shintani, A. L. Zydney, and W. H. Fissell, “Effects of pressure and
electrical charge on macromolecular transport across bovine lens basement membrane,”
Biophysical Journal, vol. 104, no. 7, pp. 1476–1484, 2013. [Online]. Available:
http://dx.doi.org/10.1016/j.bpj.2013.01.062
[31] H. MacH, S. M. Gregory, A. MacKiewicz, S. Mittal, A. Lalloo, M. Kirchmeier, and
M. Shameem, “Electrostatic interactions of monoclonal antibodies with subcutaneous tis-
sue,” Therapeutic Delivery, vol. 2, no. 6, pp. 727–736, 2011.
[32] Y. Zheng, D. B. Tesar, L. Benincosa, H. Birnb¨ ock, C. A. Boswell, D. Bumbaca, K. J. Cowan,
D. M. Danilenko, A. L. Daugherty, P. J. Fielder, H. P. Grimm, A. Joshi, N. Justies, G. Ko-
laitis,N.Lewin-Koh,J.Li,S.McVay,J.O’Mahony,M.Otteneder,M.Pantze,W.S.Putnam,
Z. J. Qiu, J. Ruppel, T. Singer, O. Stauch, F. P. Theil, J. Visich, J. Yang, Y. Ying, L. A.
Khawli, and W. F. Richter, “Minipig as a potential translatable model for monoclonal an-
tibody pharmacokinetics after intravenous and subcutaneous administration,” mAbs, vol. 4,
no. 2, pp. 243–255, 2012.
[33] T. Igawa, H. Tsunoda, T. Tachibana, A. Maeda, F. Mimoto, C. Moriyama, M. Nanami,
Y. Sekimori, Y. Nabuchi, Y. Aso, and K. Hattori, “Reduced elimination of IgG antibodies by
engineering the variable region.” Protein engineering, design & selection, vol. 23, no. 5, pp.
385–392, 5 2010.
[34] A. Datta-Mannan, A. Thangaraju, D. Leung, Y. Tang, D. R. Witcher, J. Lu, and V. J.
Wroblewski, “Balancing charge in the complementarity-determining regions of humanized
115
mAbs without affecting pl reduces non-specific binding and improves the pharmacokinetics,”
mAbs, vol. 7, no. 3, pp. 483–493, 2015.
[35] A. Schoch, H. Kettenberger, O. Mundigl, G. Winter, J. Engert, J. Heinrich, and T. Emrich,
“Charge-mediated influence of the antibody variable domain on FcRn-dependent pharma-
cokinetics,” PNAS, vol. 112, no. 19, pp. 5997–6002, 2015.
[36] A. Datta-Mannan, S. Estwick, C. Zhou, H. Choi, N. E. Douglass, D. R. Witcher, J. Lu,
C. Beidler, and R. Millican, “Influence of physiochemical properties on the subcutaneous
absorption and bioavailability of monoclonal antibodies,” mAbs, vol. 12, no. 1, pp. 1–14,
2020.
[37] C. L. Dobson, P. W. Devine, J. J. Phillips, D. R. Higazi, C. Lloyd, B. Popovic, J. Arnold,
A.Buchanan, A.Lewis, J.Goodman,C.F.VanDerWalle, P.Thornton, L.Vinall,D.Lowne,
A. Aagaard, L. L. Olsson, A. R. Wollberg, F. Welsh, T. K. Karamanos, C. L. Pashley, M. G.
Iadanza, N. A. Ranson, A. E. Ashcroft, A. D. Kippen, T. J. Vaughan, S. E. Radford, and
D. C. Lowe, “Engineering the surface properties of a human monoclonal antibody prevents
self-association and rapid clearance in vivo,” Scientific Reports , vol. 6, no. June, pp. 1–14,
2016.
[38] C. J. Bosques and A. M. Manning, “Fc-gamma receptors: Attractive targets
for autoimmune drug discovery searching for intelligent therapeutic designs,” Au-
toimmunity Reviews, vol. 15, no. 11, pp. 1081–1088, 2016. [Online]. Available:
http://dx.doi.org/10.1016/j.autrev.2016.07.035
[39] L. Abuqayyas and J. P. Balthasar, “Application of knockout mouse models to investigate
the influence of Fc γ R on the tissue distribution and elimination of 8C2, a murine IgG1
monoclonal antibody,” International Journal of Pharmaceutics, vol. 439, no. 1-2, pp. 8–16,
2012. [Online]. Available: http://dx.doi.org/10.1016/j.ijpharm.2012.09.042
[40] A. Datta-Mannan, C. K. Chow, C. Dickinson, D. Driver, J. Lu, D. R. Witcher, and V. J.
Wroblewski, “FcRn affinity-pharmacokinetic relationship of five human IgG4 antibodies en-
gineered for improved in vitro FcRn binding properties in cynomolgus monkeys,” Drug
Metabolism and Disposition, vol. 40, no. 8, pp. 1545–1555, 2012.
[41] M.J.Borrok,Y.Wu,N.Beyaz,X.Q.Yu,V.Oganesyan,W.F.Dall’Acqua,andP.Tsui,“PH-
dependentbindingengineeringrevealsanFcRnaffinitythresholdthatgovernsIgGrecycling,”
Journal of Biological Chemistry, vol. 290, no. 7, pp. 4282–4290, 2015.
[42] R. Deng, Y. G. Meng, K. Hoyte, J. Lutman, Y. Lu, S. Iyer, L. E. DeForge, F. P. Theil, P. J.
Fielder, and S. Prabhu, “Subcutaneous bioavailability of therapeutic antibodies as a function
of FcRn binding affinity in mice,” mAbs, vol. 4, no. 1, pp. 101–109, 2012.
[43] A. Datta-Mannan, D. R. Witcher, J. Lu, and V. J. Wroblewski, “Influence of improved
FcRn binding on the subcutaneous bioavailability of monoclonal antibodies in cynomolgus
monkeys,” mAbs, vol. 4, no. 2, pp. 267–273, 2012.
[44] I. P. Nnane, C. Han, Q. Jiao, S. H. Tam, H. M. Davis, and Z. Xu, “Modification of the
Fc Region of a Human Anti-oncostatin M Monoclonal Antibody for Higher Affinity to FcRn
116
ReceptorandExtensionofHalf-lifeinCynomolgusMonkeys,” Basic & Clinical Pharmacology
& Toxicology, vol. 121, pp. 13–21, 2017.
[45] L. H. Bookbinder, A. Hofer, M. F. Haller, M. L. Zepeda, G. A. Keller, J. E. Lim, T. S.
Edgington, H. M. Shepard, J. S. Patton, and G. I. Frost, “A recombinant human enzyme
for enhanced interstitial transport of therapeutics,” Journal of Controlled Release, vol. 114,
no. 2, pp. 230–241, 2006.
[46] M.J.Dolton, A.Chesterman, A.Moein, K.M.Sink, A.Waitz, K.Blondeau, G.A.Kerchner,
N. Hu, L. Brooks, M. K. Wetzel-Smith, A. Roden, A. Deshmukh, K. Peng, M. Carrasco-
Triguero, J. Smith, S. Ostrowitzki, and A. Quartino, “Safety, Tolerability, and Pharmacoki-
netics of High-Volume Subcutaneous Crenezumab, With and Without Recombinant Human
Hyaluronidase in Healthy Volunteers,” Clinical Pharmacology and Therapeutics, 2021.
[47] P.N.Morcos,X.Zhang,C.McIntyre,B.Bittner,L.Rowell,andZ.Hussain,“Pharmacokinet-
ics and pharmacodynamics of single subcutaneous doses of tocilizumab administered with or
without rHuPH20,” Int. Journal of Clinical Pharmacology and Therapeutics, vol. 51, no. 07,
pp. 537–548, 2013.
[48] A. Bass, A. Plotka, K. Mridha, C. Sattler, A. M. Kim, and D. R. Plowchalk, “Pharma-
cokinetics, pharmacodynamics, and safety of bococizumab, a monoclonal antibody against
proprotein convertase subtilisin/kexin type 9, in healthy subjects when administered in co-
mixture with recombinant human hyaluronidase: A phase 1 randomized tria,” Health Science
Reports, vol. 1, no. 9, pp. 1–10, 2018.
[49] L. T. Baxter, H. Zhu, R. K. Jain, H. Zhu, D. G. Mackensen, and W. F. Butler, “Biodistri-
bution of Monoclonal Antibodies: Scale-up from Mouse to Human Using a Physiologically
Based Pharmacokinetic Model,” Cancer Research, vol. 55, no. 20, pp. 4611–4622, 1995.
[50] A. Garg and J. P. Balthasar, “Physiologically-based pharmacokinetic (PBPK) model to pre-
dict IgG tissue kinetics in wild-type and FcRn-knockout mice,” J Pharmacokinet Pharmaco-
dyn, vol. 34, pp. 687–709, 2007.
[51] P. M. Glassman and J. P. Balthasar, “Physiologically-based pharmacokinetic modeling to
predicttheclinicalpharmacokineticsofmonoclonalantibodies,” Journal of Pharmacokinetics
and Pharmacodynamics, vol. 43, no. 4, pp. 427–446, 2016.
[52] D. K. Shah and A. M. Betts, “Towards a platform PBPK model to characterize the plasma
and tissue disposition of monoclonal antibodies in preclinical species and human,” Journal of
Pharmacokinetics and Pharmacodynamics, vol. 39, no. 1, pp. 67–86, 2012.
[53] D. G. Covell, J. Barbet, O. D. Holton, C. D. Black, J. N. Weinstein, and R. J. Parker,
“Pharmacokinetics of monoclonal immunoglobulin G1, F(ab’)2, and Fab’ in mice,” Cancer
Research, vol. 46, no. 8, pp. 3969–3978, 1986.
[54] Yang Chen and Joseph P. Balthasar, “Evaluation of a Catenary PBPK Model for Predicting
the In Vivo Disposition of mAbs Engineered for High-Affinity Binding to FcRn,” The AAPS
Journal, vol. 14, no. 4, 2012.
117
[55] B. RIPPE and B. HARALDSSON, “Fluid and protein fluxes across small and large pores
in the microvasculature. Application of two-pore equations,” Acta Physiologica Scandinavica,
vol. 131, no. 3, pp. 411–428, 1987.
[56] B. Rippe and B. Haraldsson, “Transport of macromolecules across microvascular walls: the
two-pore theory,” Physiological Reviews, vol. 74, no. 1, pp. 163–219, 1994.
[57] A. Sepp, A. Berges, A. Sanderson, and G. Meno-Tetang, “Development of a physiologically
based pharmacokinetic model for a domain antibody in mice using the two-pore theory,”
JOURNAL OF PHARMACOKINETICS AND PHARMACODYNAMICS, vol. 42, no. 2, pp.
97–109, 4 2015.
[58] L. Zhao, P. Ji, Z. Li, P. Roy, and C. G. Sahajwalla, “The antibody drug absorption following
subcutaneous or intramuscular administration and its mathematical description by coupling
physiologicallybasedabsorptionprocesswiththeconventionalcompartmentpharmacokinetic
model,” Journal of Clinical Pharmacology, vol. 53, no. 3, pp. 314–325, 2013.
[59] E. Offman and A. N. Edginton, “A PBPK workflow for first-in-human dose selection of a
subcutaneously administered pegylated peptide,” Journal of Pharmacokinetics and Pharma-
codynamics, vol. 42, no. 2, pp. 135–150, 2015.
[60] K. L. Gill, I. Gardner, L. Li, and M. Jamei, “A Bottom-Up Whole-Body Physiologically
Based Pharmacokinetic Model to Mechanistically Predict Tissue Distribution and the Rate
of Subcutaneous Absorption of Therapeutic Proteins,” The AAPS Journal, vol. 18, no. 1, pp.
156–170, 2016.
[61] Z. Li, X. Yu, Y. Li, A. Verma, H. P. Chang, and D. K. Shah, “A Two-Pore Physiologi-
cally Based Pharmacokinetic Model to Predict Subcutaneously Administered Different-Size
Antibody/Antibody Fragments,” The AAPS Journal, vol. 23, no. 62, 2021.
[62] T. E. Kraft, W. F. Richter, T. Emrich, A. Knaupp, M. Schuster, A. Wolfert, and H. Ketten-
berger,“Heparinchromatographyasaninvitropredictorforantibodyclearanceratethrough
pinocytosis,” mAbs, vol. 12, no. 1, pp. 1–9, 2020.
[63] A. Datta-Mannan and V. J. Wroblewski, “Application of FcRn binding assays to guide mAb
development,” Drug Metabolism and Disposition, vol. 42, no. 11, pp. 1867–1872, 2014.
[64] T. Jain, T. Sun, S. Durand, A. Hall, N. R. Houston, J. H. Nett, B. Sharkey, B. Bobrowicz,
I. Caffry, Y. Yu, Y. Cao, H. Lynaugh, M. Brown, H. Baruah, L. T. Gray, E. M. Krauland,
Y.Xu, M.V´ asquez, andK.D.Wittrup, “Biophysicalpropertiesoftheclinical-stageantibody
landscape,” Proceedings of the National Academy of Sciences, vol. 114, no. 5, pp. 944–949,
2017.
[65] M. I. J. Raybould, C. Marks, K. Krawczyk, B. Taddese, J. Nowak, A. P. Lewis, A. Bujotzek,
J. Shi, and C. M. Deane, “Five computational developability guidelines for therapeutic an-
tibody profiling,” Proceedings of the National Academy of Sciences, vol. 116, no. 10, pp.
4025–4030, 2019.
[66] U. Chemical Computing Group, “Molecular Operating Environment (MOE),” Montreal. QC
(Canada), 2019.
118
[67] D. R. Goulet, M. J. Watson, S. H. Tam, A. Zwolak, M. L. Chiu, W. M. Atkins, and A. Nath,
“Toward a Combinatorial Approach for the Prediction of IgG Half-Life and Clearance,” Drug
Metabolism and Disposition, vol. 46, no. 12, pp. 1900–1907, 2018.
[68] L. B. Avery, J. Wade, M. Wang, A. Tam, A. King, N. Piche-Nicholas, M. S. Kavosi, S. Penn,
D. Cirelli, J. C. Kurz, M. Zhang, O. Cunningham, R. Jones, B. J. Fennell, B. McDonnell,
P. Sakorafas, J. Apgar, W. J. Finlay, L. Lin, L. Bloom, and D. M. O’Hara, “Establishing
in vitro in vivo correlations to screen monoclonal antibodies for physicochemical properties
related to favorable human pharmacokinetics,” mAbs, vol. 10, no. 2, pp. 244–255, 2018.
[69] S. Chung, V. Nguyen, Y. Linda, J. Lafrance-vanasse, S. J. Scales, and K. Lin, “An in vitro
FcRn- dependent transcytosis assay as aa screening tool for predictive assessment of non-
specific clearance of antibody therapeutics in humans,” mAbs, vol. 11, no. 5, pp. 942–955,
2019.
[70] A. Datta-Mannan, J. Lu, D. R. Witcher, D. Leung, Y. Tang, and V. J. Wroblewski, “The
interplay of non-specific binding, targetmediated clearance and FcRn interactions on the
pharmacokinetics of humanized antibodies,” mAbs, vol. 7, no. 6, pp. 1084–1093, 2015.
[71] R. L. Kelly, Y. Yu, T. Sun, I. Caffry, H. Lynaugh, M. Brown, T. Jain, Y. Xu, and
K. D. Wittrup, “Target-independent variable region mediated effects on antibody clearance
can be FcRn independent,” mAbs, vol. 8, no. 7, pp. 1269–1275, 2016. [Online]. Available:
http://dx.doi.org/10.1080/19420862.2016.1208330
[72] B.Grinshpun,N.Thorsteinson,J.N.Pereira,F.Rippmann,D.Nannemann,V.D.Sood,and
Y. Fomekong Nanfack, “Identifying biophysical assays and in silico properties that enrich for
slow clearance in clinical-stage therapeutic antibodies,” mAbs, vol. 13, no. 1, pp. 1–12, 2021.
[73] H. M. Jones, Z. Zhang, P. Jasper, H. Luo, L. B. Avery, L. E. King, H. Neubert, H. A.
Barton, A. M. Betts, and R. Webster, “A Physiologically-Based Pharmacokinetic Model
for the Prediction of Monoclonal Antibody Pharmacokinetics From In Vitro Data,” CPT:
Pharmacometrics & Systems Pharmacology, vol. 8, pp. 738–747, 2019.
[74] S.HuandD.Z.D’Argenio,“Predictingmonoclonalantibodypharmacokineticsfollowingsub-
cutaneous administration via whole-body physiologically-based modeling,” Journal of Phar-
macokinetics and Pharmacodynamics, vol. 47, no. 5, pp. 385–409, 2020.
[75] M.H.Weisman,L.W.Moreland,D.E.Furst,M.E.Weinblatt,E.C.Keystone,H.E.Paulus,
L. S. Teoh, R. B. Velagapudi, P. A. Noertersheuser, G. R. Granneman, S. A. Fischkoff, and
E. K. Chartash, “Efficacy, pharmacokinetic, and safety assessment of adalimumab, a fully
human anti-tumor necrosis factor-alpha monoclonal antibody, in adults with rheumatoid
arthritis receiving concomitant methotrexate: A pilot study,” Clinical Therapeutics, vol. 25,
no. 6, pp. 1700–1721, 2003.
[76] M. Stroh, H. Winter, M. Marchand, L. Claret, S. Eppler, J. Ruppel, O. Abidoye, S. L. Teng,
W. T. Lin, S. Dayog, R. Bruno, J. Jin, and S. Girish, “Clinical Pharmacokinetics and Phar-
macodynamics of Atezolizumab in Metastatic Urothelial Carcinoma,” Clinical Pharmacology
and Therapeutics, vol. 102, no. 2, pp. 305–312, 2017.
119
[77] R. S. Black, “A Single Ascending Dose Study of Bapineuzumab in Patients With Alzheimer
Disease,” Alzheimer Dis Assoc Disord, vol. 24, no. 2, pp. 198–203, 2010.
[78] W. W. Busse, R. Katial, D. Gossage, S. Sari, B. Wang, R. Kolbeck, A. J. Coyle, M. Koike,
G.L.Spitalny,P.A.Kiener,G.P.Geba,andN.A.Molfino,“Safetyprofile,pharmacokinetics,
and biologic activity of MEDI-563, an anti-IL-5 receptor α antibody, in a phase I study of
subjects with mild asthma,” Journal of Allergy and Clinical Immunology, vol. 125, no. 6, pp.
1237–1244, 2010.
[79] B. Knight, D. Rassam, S. Liao, and R. Ewesuedo, “A phase I pharmacokinetics study com-
paring PF-06439535 (a potential biosimilar) with bevacizumab in healthy male volunteers,”
Cancer Chemotherapy and Pharmacology, vol. 77, no. 4, pp. 839–846, 2016.
[80] A. Chakraborty, S. Tannenbaum, C. Rordorf, P. J. Lowe, D. Floch, H. Gram, and S. Roy,
“Pharmacokinetic and Pharmacodynamic Properties of Canakinumab, a Human Anti-
Interleukin-1b Monoclonal Antibody,” Clinical Pharmacokinetics, vol. 51, no. 6, pp. e1–e18,
2012.
[81] A. A. Othman, J. Q. Tran, M. T. Tang, and S. Dutta, “Population Pharmacokinetics of Da-
clizumab High-Yield Process in Healthy Volunteers: Integrated Analysis of Intravenous and
Subcutaneous,Single-andMultiple-DoseAdministrationKeyPoints,” ClinicalPharmacology,
vol. 53, pp. 907–918, 2014.
[82] L. S. Rosen, J. W. Goldman, A. P. Algazi, P. K. Turner, B. Moser, T. Hu, X. A. Wang,
J. Tuttle, V. Wacheck, J. E. Wooldridge, and M. Banck, “A first-in-human phase I study of
a bivalent MET antibody, emibetuzumab (LY2875358), as monotherapy and in combination
with erlotinib in advanced cancer,” Clinical Cancer Research, vol. 23, no. 8, pp. 1910–1919,
2017.
[83] B. White, F. Leon, W. White, and G. Robbie, “Two first-in-human, open-label, phase I dose-
escalation safety trials of MEDI-528, a monoclonal antibody against interleukin-9, in healthy
adult volunteers,” Clinical Therapeutics, vol. 31, no. 4, pp. 728–740, 4 2009.
[84] Y.Sasaki,K.Miwa,K.Yamashita,Y.Sunakawa,K.Shimada,H.Ishida,K.Hasegawa,K.Fu-
jiwara, M. Kodaira, Y. Fujiwara, M. Namiki, M. Matsuda, Y. Takeuchi, and N. Katsumata,
“A phase I study of farletuzumab, a humanized anti-folate receptor α monoclonal antibody,
in patients with solid tumors,” Investigational New Drugs, vol. 33, no. 2, pp. 332–340, 2015.
[85] D. Yin, B. Sleight, C. Alvey, A. G. Hansson, and A. Bello, “Pharmacokinetics and pharma-
codynamics of figitumumab, a monoclonal antibody targeting the insulin-like growth factor
1 receptor, in healthy participants,” Journal of Clinical Pharmacology, vol. 53, no. 1, pp.
21–28, 1 2013.
[86] J. M. Gow, W. H. Tsuji, G. J. Williams, D. Mytych, D. Sciberras, S. L. Searle, T. Mant,
and J. P. Gibbs, “Safety, tolerability, pharmacokinetics, and efficacy of AMG 403, a human
anti-nerve growth factor monoclonal antibody, in two phase I studies with healthy volunteers
and knee osteoarthritis subjects,” Arthritis Research and Therapy, vol. 17, no. 1, pp. 1–11,
2015.
120
[87] C. Cavelti-Weder, A. Babians-Brunner, C. Keller, M. A. Stahel, M. Kurz-Levin, H. Zayed,
A. M. Solinger, T. Mandrup-Poulsen, C. A. Dinarello, and M. Y. Donath, “Effects of gevok-
izumab on glycemia and inflammatory markers in type 2 diabetes,” Diabetes Care, vol. 35,
no. 8, pp. 1654–1662, 8 2012.
[88] Y. Zhuang, C. Calderon, S. J. Marciniak, E. Bouman-Thio, P. Szapary, T.-Y. Yang,
A. Schantz, H. M. Davis, H. Zhou, and Z. Xu, “First-in-human study to assess guselkumab
(anti-IL-23 mAb) pharmacokinetics/safety in healthy subjects and patients with moderate-
to-severepsoriasis,”European Journal of Clinical Pharmacology,vol.72,pp.1303–1310,2016.
[89] H. Ortega, S. Yancey, and S. Cozens, “Pharmacokinetics and absolute bioavailability of
mepolizumab following administration at subcutaneous and intramuscular sites,” Clinical
Pharmacology in Drug Development, vol. 3, no. 1, pp. 57–62, 1 2014.
[90] K. Kretsos, G. Golor, A. Jullion, M. Hickling, S. Mccabe, S. Shaw, J. Jose, and R. Oliver,
“Safety and Pharmacokinetics of Olokizumab , an Anti-IL-6 Monoclonal Antibody , Admin-
istered to Healthy Male Volunteers : A Randomized Phase I Study,” Clinical Pharmacology
in Drug Development, 2014.
[91] V.Meininger,P.F.Pradat,A.Corse,S.Al-Sarraj,B.R.Brooks,J.B.Caress,M.Cudkowicz,
S.J.Kolb,D.Lange,P.N.Leigh,T.Meyer,S.Milleri,K.E.Morrison,R.W.Orrell,G.Peters,
J.D.Rothstein,J.Shefner,A.Lavrov,N.Williams,P.Overend,J.Price,S.Bates,J.Bullman,
D. Krull, A. Berges, B. Abila, G. Meno-Tetang, and J. Wurthner, “Safety, pharmacokinetic,
and functional effects of the Nogo-A monoclonal antibody in amyotrophic lateral sclerosis: A
randomized, first-in-human clinical trial,” PLoS ONE, vol. 9, no. 5, 2014.
[92] D. B. Agus, M. S. Gordon, C. Taylor, R. B. Natale, B. Karlan, D. S. Mendelson, M. F. Press,
D. E. Allison, M. X. Sliwkowski, G. Lieberman, S. M. Kelsey, and G. Fyfe, “Phase I clinical
studyofpertuzumab, anovelHERdimerizationinhibitor, inpatientswithadvancedcancer,”
Journal of Clinical Oncology, vol. 23, no. 11, pp. 2534–2543, 2005.
[93] S. Khalilieh, P. Hodsman, C. Xu, A. Tzontcheva, S. Glasgow, and D. Montgomery, “Phar-
macokinetics of Tildrakizumab (MK-3222), an Anti-IL-23 Monoclonal Antibody, After Intra-
venous or Subcutaneous Administration in Healthy Subjects,” Basic and Clinical Pharma-
cology and Toxicology, vol. 123, no. 3, pp. 294–300, 2018.
[94] C. K. Oh, R. Faggioni, F. Jin, L. K. Roskos, B. Wang, C. Birrell, R. Wilson, and N. A.
Molfino, “An open-label, single-dose bioavailability study of the pharmacokinetics of CAT-
354 after subcutaneous and intravenous administration in healthy males,” British Journal of
Clinical Pharmacology, vol. 69, no. 6, pp. 645–655, 6 2010.
[95] J. Morita, M. Tanaka, M. Nomoto, S. Matsuki, T. Tsuru, K. Matsuguma, and M. Shiramoto,
“PharmacokineticBioequivalence, Safety, andImmunogenicityofDMB-3111, aTrastuzumab
Biosimilar, and Trastuzumab in Healthy Japanese Adult Males: Results of a Randomized
Trial,” BioDrugs, vol. 30, pp. 17–25, 2016.
[96] C. L. Kauffman, N. Aria, E. Toichi, T. S. McCormick, K. D. Cooper, A. B. Gottlieb, D. E.
Everitt, B. Frederick, Y. Zhu, M. A. Graham, C. E. Pendley, and M. A. Mascelli, “A phase
I study evaluating the safety, pharmacokinetics, and clinical response of a human IL-12 p40
121
antibody in subjects with plaque psoriasis,” Journal of Investigative Dermatology, vol. 123,
no. 6, pp. 1037–1044, 2004.
[97] S. Bosgra, J. Van Eijkeren, P. Bos, M. Zeilmaker, and W. Slob, “An improved model to
predict physiologically based model parameters and their inter-individual variability from
anthropometry,” Critical Reviews in Toxicology, vol. 42, no. 9, pp. 751–767, 2012.
[98] R. P. Brown, M. D. Delp, S. L. Lindstedt, L. R. Rhomberg, and R. P. Beliles, “Physio-
logical parameter values for physiologically based pharmacokinetic models,” Toxicology and
Industrial Health, vol. 13, no. 4, pp. 407–484, 1997.
[99] F. D. Sowby, “Basic anatomical and physiological data for use in radiological protection:
reference values,” Annals of the ICRP, vol. 6, no. 1, p. 1, 2002.
[100] D. Z. D’Argenio, S. Alan, and X. Wang, ADAPT 5 User’s Guide: Pharmacoki-
netic/Pharmacodynamic Systems Analysis Software, Biomedical Simulations Resources, Ed.
Los Angeles: Biomedical Simulations Resources, 2009.
[101] T. A. WALDMANN and W. STROBER, “Metabolism of Immunoglobulins,” in Clinical Im-
munobiology, 1976, pp. 71–95.
[102] T. Li and J. P. Balthasar, “Development and Evaluation of a Physiologically Based Phar-
macokinetic Model for Predicting the Effects of Anti-FcRn Therapy on the Disposition of
Endogenous IgG in Humans,” Journal of Pharmaceutical Sciences, vol. 108, no. 1, pp. 714–
724, 1 2019.
[103] RCoreTeam,“R:ALanguageandEnvironmentforStatisticalComputing,”Vienna,Austria,
2021.
[104] C. G. Starr and P. M. Tessier, “Selecting and engineering monoclonal antibodies with drug-
like specificity,” Current Opinion in Biotechnology, vol. 60, pp. 119–127, 2019.
[105] E. K. Makowski, L. Wu, P. Gupta, and P. M. Tessier, “Discovery-stage identification of drug-
like antibodies using emerging experimental and computational methods,” mAbs, vol. 13,
no. 1, p. 1895540, 2021.
[106] S. Liu and D. K. Shah, “Mathematical models to characterize the absorption , distribution
, metabolism , and excretion ( ADME ) of protein therapeutics,” Drug Metabolism and
Disposition, vol. 50, no. 8, 2022.
[107] A. Mahmuda, F. Bande, K. J. K. Al-Zihiry, N. Abdulhaleem, R. A. Majid, R. A. Hamat,
W.O.Abdullah,andZ.Unyah,“Monoclonalantibodies: Areviewoftherapeuticapplications
and future prospects,” Tropical Journal of Pharmaceutical Research, vol. 16, no. 3, pp. 713–
722, 2017.
[108] “FDA Approved Drug Products.” [Online]. Available:
https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm
[109] “European Medicines Agency: Medicines.” [Online]. Available:
https://www.ema.europa.eu/en/medicines
122
[110] B. Bittner, W. Richter, and J. Schmidt, “Subcutaneous Administration of Biotherapeutics:
An Overview of Current Challenges and Opportunities,” BioDrugs, vol. 32, no. 5, pp.
425–440, 2018. [Online]. Available: https://doi.org/10.1007/s40259-018-0295-0
[111] P. M. Glassman and J. P. Balthasar, “Physiologically-based modeling of monoclonal
antibody pharmacokinetics in drug discovery and development,” Drug Metabolism
and Pharmacokinetics, vol. 34, no. 1, pp. 3–13, 2 2019. [Online]. Available:
https://linkinghub.elsevier.com/retrieve/pii/S1347436718301332
[112] E. Offman, C. Phipps, and A. N. Edginton, “Population physiologically-based
pharmacokinetic model incorporating lymphatic uptake for a subcutaneously administered
pegylated peptide,” In Silico Pharmacology, vol. 4, no. 1, pp. 1–14, 2016. [Online]. Available:
http://dx.doi.org/10.1186/s40203-016-0018-5
[113] N. Varkhede and M. L. Forrest, “Understanding the monoclonal antibody disposition after
subcutaneous administration using a minimal physiologically based pharmacokinetic model,”
Journal of Pharmacy and Pharmaceutical Sciences, vol. 21, no. 1S, pp. 130s–148s, 2018.
[114] I. Lowy, P. L. Hibberd, C. P. Taylor, L. Davidson, A. Lembo, S. Tummala, R. J. Farrell,
D. Molrine, and C. P. Kelly, “Open-label, dose escalation phase I study in healthy volunteers
to evaluate the safety and pharmacokinetics of a human monoclonal antibody to Clostridium
difficile toxin A,” Vaccine, vol. 26, no. 27-28, pp. 3404–3409, 2008.
[115] d. B. A, v. d. P. L, R. R, S. M, V. R. P, S. O, B. C, F. H, B. Y, V. R, K. J, and K. H, “A
single dose, placebo controlled study of the fully human anti-tumor necrosis,” J Rheumatol,
vol. 29, no. September 2015, pp. 2288–2298, 2002.
[116] W. W. Cai, M. Fiscella, C. Chen, Z. J. Zhong, W. W. Freimuth, and D. C. Subich, “Bioavail-
ability, pharmacokinetics, and safety of belimumab administered subcutaneously in healthy
subjects,” Clinical Pharmacology in Drug Development, vol. 2, no. 4, pp. 349–357, 2013.
[117] T. Derfuss, F. Curtin, C. Guebelin, C. Bridel, M. Rasenack, A. Matthey, R. Du Pasquier,
M. Schluep, J. Desmeules, A. B. Lang, H. Perron, R. Faucard, H. Porchet, H. P. Hartung,
L. Kappos, and P. H. Lalive, “A phase IIa randomized clinical study testing GNbAC1, a
humanized monoclonal antibody against the envelope protein of multiple sclerosis associated
endogenous retrovirus in multiple sclerosis patients - A twelve month follow-up,” Journal of
Neuroimmunology, vol. 285, pp. 68–70, 2015.
[118] Z. Xu, Q. Wang, Y. Zhuang, B. Frederick, H. Yan, E. Bouman-Thio, J. C.
Marini, M. Keen, D. Snead, H. M. Davis, and H. Zhou, “Subcutaneous
Bioavailability of Golimumab at 3 Different Injection Sites in Healthy Subjects,”
Journal of Clinical Pharmacology, vol. 50, pp. 276–284, 2010. [Online]. Available:
https://accp1.onlinelibrary.wiley.com/doi/pdf/10.1177/0091270009340782
[119] W. Park, S. J. Lee, J. Yun, and D. H. Yoo, “Comparison of the pharmacokinetics and safety
of three formulations of infliximab (CT-P13, EU-approved reference infliximab and the US-
licensed reference infliximab) in healthy subjects: A randomized, double-blind, three-arm,
parallel-group, single-dose, ,” Expert Review of Clinical Immunology, vol. 11, pp. S25–S31,
2015.
123
[120] G. M. Subramanian, P. W. Cronin, G. Poley, A. Weinstein, S. M. Stoughton, J. Zhong,
Y. Ou, J. F. Zmuda, B. L. Osborn, and W. W. Freimuth, “A Phase 1 Study of PAmAb, a
FullyHumanMonoclonalAntibodyagainstBacillusanthracisProtectiveAntigen, inHealthy
Volunteers,” Clinical Infectious Diseases, vol. 41, no. 1, pp. 12–20, 6 2005.
[121] A. Khatri, D. Eckert, R. Oberoi, A. Suleiman, Y. Pang, L. Cheng, and A. A.
Othman, “Pharmacokinetics of Risankizumab in Asian Healthy Subjects and Patients With
Moderate to Severe Plaque Psoriasis, Generalized Pustular Psoriasis, and Erythrodermic
Psoriasis,” The Journal of Clinical Pharmacology, pp. 1–13, 2019. [Online]. Available:
http://doi.wiley.com/10.1002/jcph.1473
[122] T.Puchalski, U.Prabhakar, Q.Jiao, B.Berns, andH.M.Davis, “Pharmacokineticandphar-
macodynamic modeling of an anti-interleukin-6 chimeric monoclonal antibody (siltuximab)
in patients with metastatic renal cell carcinoma,” Clinical Cancer Research, vol. 16, no. 5,
pp. 1652–1661, 3 2010.
[123] S. Hetherington, M. Texter, E. Wenzel, J. M. Patti, L. Reynolds, T. Shamp, and S. Swan,
“Phase I dose escalation study to evaluate the safety and pharmacokinetic profile of
tefibazumab in subjects with end-stage renal disease requiring hemodialysis,” Antimicrobial
Agents and Chemotherapy, vol. 50, no. 10, pp. 3499–3500, 10 2006.
[124] E. L. L´ opez, M. M. Contrini, E. Glatstein, S. G. Ayala, R. Santoro, D. Allende, G. Ezcurra,
E.Teplitz,T.Koyama,Y.Matsumoto,H.Sato,K.Sakai,S.Hoshide,K.Komoriya,T.Morita,
R. Harning, and S. Brookman, “Safety and pharmacokinetics of urtoxazumab, a human-
ized monoclonal antibody, against Shiga-like toxin 2 in healthy adults and in pediatric pa-
tients infected with Shiga-like toxin-producing Escherichia coli,” Antimicrobial Agents and
Chemotherapy, vol. 54, no. 1, pp. 239–243, 1 2010.
[125] J. Hillson, T. Mant, . M. Rosano, C. Huntenburg, . Mehrshid Alai-Safar, . Siddhesh Darne,
D.Palmer,.Borislava,G.Pavlova,J.Doralt,R.Reeve,N.Goel,D.Weilert,.Paul,W.Rhyne,
K.Chance,J.Caminis,J.Roach,andT.Ganguly,“Pharmacokineticequivalence,comparable
safety,andimmunogenicityofanadalimumabbiosimilarproduct(M923)toHumirainhealthy
subjects,” Pharmacol Res Perspect, p. 380, 2018.
[126] S. Schreiber, “Development of a novel auto-injector of subcutaneous CT-P13 infliximab:
Phase I randomised, open-label, single-dose trial to compare the pharmacokinetics and safety
topre-filledsyringeinhealthysubjects,” Abstracts of the 14th Congress of ECCO – European
Crohn’s and Colitis Organisation, pp. 458–459, 2019.
[127] H. M. Kinnunen and R. J. Mrsny, “Improving the outcomes of biopharmaceutical delivery
via the subcutaneous route by understanding the chemical, physical and physiological
properties of the subcutaneous injection site,” Journal of Controlled Release, vol. 182, no. 1,
pp. 22–32, 2014. [Online]. Available: http://dx.doi.org/10.1016/j.jconrel.2014.03.011
[128] Y.-B. Hu, E. B. Dammer, R.-J. Ren, and G. Wang, “The endosomal-lysosomal system:
from acidification and cargo sorting to neurodegeneration,” Translational Neurodegeneration,
vol. 4, no. 18, 2015.
124
[129] R. H. Mellor, A. W. B. Stanton, P. Azarbod, M. D. Sherman, J. R. Levick, and P. S.
Mortimer, “Enhanced Cutaneous Lymphatic Network in the Forearms of Women with
Postmastectomy Oedema,” J Vasc Res, vol. 37, pp. 501–512, 2000. [Online]. Available:
www.karger.com/journals/jvr
[130] A. Stanton, W. Svensson, R. Mellor, A. Peters, J. Levick, and P. Mortimer, “Differences in
lymph drainage between swollen and non-swollen regions in arms with breast-cancer-related
lymphoedema,” Clinical Science, vol. 101, no. 2, pp. 131–140, 2001.
[131] M. A. Swartz, “The physiology of the lymphatic system,” Advanced Drug Delivery Reviews,
vol. 50, pp. 3–20, 2001. [Online]. Available: www.elsevier.com/locate/drugdeliv
[132] H. Kim, R. N. Fariss, C. Zhang, S. B. Robinson, M. Thill, and K. G.
Csaky, “Mapping of the Neonatal Fc Receptor in the Rodent Eye,” Invest
Ophthalmol Vis Sci, vol. 49, no. 5, pp. 2025–2029, 2008. [Online]. Available:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3950474/pdf/nihms-548613.pdf
[133] S. Latvala, B. Jacobsen, M. B. Otteneder, A. Herrmann, and S. Kronenberg, “Distribution
of FcRn across species and tissues,” Journal of Histochemistry & Cytochemistry, vol. 65,
no. 6, pp. 321–333, 2017. [Online]. Available: https://doi.org/10.1369/0022155417705095
[134] J. R. Chabot, D. E. Dettling, P. J. Jasper, and B. C. Gomes, “Comprehensive mechanism-
based antibody pharmacokinetic modeling,” in Proceedings of the Annual International Con-
ference of the IEEE Engineering in Medicine and Biology Society, EMBS. IEEE, 2011, pp.
4318–4323.
[135] A. Rostami-Hodjegan, M. Jamei, M. Chetty, H. Musther, M. Rowland, and K. L. Gill, “Are
Physiologically Based Pharmacokinetic Models Reporting the Right Cmax? Central Venous
Versus Peripheral Sampling Site,” The AAPS Journal, vol. 17, no. 5, pp. 1268–1279, 2015.
[136] M.YingandB.S.Pang,“Three-dimensionalultrasoundmeasurementofcervicallymphnode
volume,” British Journal of Radiology, vol. 82, no. 980, pp. 617–625, 2009.
[137] K. N. Margaris and R. A. Black, “Modelling the lymphatic system: Challenges and oppor-
tunities,” Journal of the Royal Society Interface, vol. 9, no. 69, pp. 601–612, 2012.
[138] A. Stanton, S. Modi, R. Mellor, A. Peters, W. Svensson, J. Levick, and P. Mortimer, “A
quantitative lymphoscintigraphic evaluation of lymphatic function in the swollen hands of
women with lymphoedema following breast cancer treatment,” Clinical Science, vol. 110,
no. 5, pp. 553–561, 2006.
[139] L. P. Kozlowski, “IPC - Isoelectric Point Calculator,” Biology Direct, vol. 11, no. 1, pp.
1–16, 2016. [Online]. Available: http://dx.doi.org/10.1186/s13062-016-0159-9
[140] D. N. McLennan, C. J. Porter, and S. A. Charman, “Subcutaneous drug delivery and the
role of the lymphatics,” Drug Discovery Today: Technologies, vol. 2, no. 1, pp. 89–96, 2005.
[141] M. Berger, P. A. Halban, L. Girardier, J. Seydoux, R. E. Offord, and A. E. Renold, “Ab-
sorptionkineticsof subcutaneously injectedinsulin.Evidencefor degradation attheinjection
site.” Diabetologia, vol. 17, no. 2, pp. 97–99, 8 1979.
125
[142] B. Law and C.-H. Tung, “Proteolysis: a biological process adapted in drug delivery, therapy,
and imaging.” Bioconjugate chemistry, vol. 20, no. 9, pp. 1683–1695, 9 2009.
[143] R. M. Watanabe, A. Volund, and R. N. Bergman, “Intravenous insulin infusion to simu-
late subcutaneous absorption. Bioavailability and metabolic sequelae.” Diabetes care, vol. 14,
no. 11, pp. 1021–1030, 11 1991.
[144] G. Freckmann, S. Pleus, A. Westhoff, L. G. Krinelke, A. Buhr, N. Jendrike, and C. Haug,
“Clinical Performance of a Device That Applies Local Heat to the Insulin Infusion Site: A
Crossover Study,” Journal of Diabetes Science and Technology, vol. 6, no. 2, pp. 320–327,
2012. [Online]. Available: www.journalofdst.org
[145] V.S.Trubetskoy,K.R.Whiteman,V.P.Torchilin,andG.L.Wolf,“Massage-inducedrelease
of subcutaneously injected liposome-encapsulated drugs to the blood,” Journal of Controlled
Release, vol. 50, no. 1-3, pp. 13–19, 1998.
[146] O. Shpilberg and C. Jackisch, “Subcutaneous administration of rituximab (MabThera) and
trastuzumab (Herceptin) using hyaluronidase,” British Journal of Cancer, vol. 109, no. 6,
pp. 1556–1561, 2013. [Online]. Available: http://dx.doi.org/10.1038/bjc.2013.371
[147] J. R. Thomas, R. C. Yocum, M. F. Haller, and J. Flament, “The INFUSE-Morphine IIB
Study: Use of Recombinant Human Hyaluronidase (rHuPH20) to Enhance the Absorption of
Subcutaneous Morphine in Healthy Volunteers,” Journal of Pain and Symptom Management,
vol. 38, no. 5, pp. 673–682, 2009. [Online]. Available: www.ClinicalTrials.gov,
[148] I. H¨ otzel, F. P. Theil, L. J. Bernstein, S. Prabhu, R. Deng, L. Quintana, J. Lutman, R. Sibia,
P. Chan, D. Bumbaca, P. Fielder, P. J. Carter, and R. F. Kelley, “A method for detecting
off-target elimination,” mAbs, vol. 4, no. 6, pp. 753–760, 2012.
[149] B. Li, D. Tesar, C. A. Boswell, H. S. Cahaya, A. Wong, J. Zhang, M. Gloria, C. Eigen-
brot, H. Pantua, J. Diao, S. B. Kapadia, R. Deng, and R. F. Kelley, “Framework selection
can influence pharmacokinetics of a humanized therapeutic antibody through differences in
molecule charge,” mAbs, vol. 6, no. 5, pp. 1255–1264, 2014.
[150] C.A.Boswell,D.B.Tesar,K.Mukhyala,F.-P.Theil,P.J.Fielder,andL.A.Khawli,“Effects
of charge on antibody tissue distribution and pharmacokinetics.” Bioconjugate chemistry,
vol. 21, no. 12, pp. 2153–2163, 12 2010.
[151] L.A.Khawli,S.Goswami,R.Hutchinson,Z.W.Kwong,J.Yang,X.Wang,Z.Yao,A.Sreed-
hara, T. Cano, D. Tesar, I. Nijem, D. E. Allison, P. Y. Wong, Y.-H. Kao, C. Quan, A. Joshi,
R. J. Harris, and P. Motchnik, “Charge variants in IgG1: Isolation, characterization, in vitro
binding properties and pharmacokinetics in rats.” mAbs, vol. 2, no. 6, pp. 613–624, 2010.
[152] D.BumbacaYadav,V.K.Sharma,C.AndrewBoswell,I.Hotzel,D.Tesar,Y.Shang,Y.Ying,
S. K. Fischer, J. L. Grogan, E. Y. Chiang, K. Urban, S. Ulufatu, L. A. Khawli, S. Prabhu,
S. Joseph, and R. F. Kelley, “Evaluating the Use of Antibody Variable Region (Fv) Charge
as a Risk Assessment Tool for Predicting Typical Cynomolgus Monkey Pharmacokinetics,”
THE JOURNAL OF BIOLOGICAL CHEMISTRY, vol. 290, no. 50, pp. 29732–29741, 2015.
126
[153] V. K. Sharma, T. W. Patapoff, B. Kabakoff, S. Pai, E. Hilario, B. Zhang, C. Li, O. Borisov,
R. F. Kelley, I. Chorny, J. Z. Zhou, K. A. Dill, and T. E. Swartz, “In silico selection of thera-
peutic antibodies for development: Viscosity, clearance, and chemical stability,” Proceedings
of the National Academy of Sciences, vol. 111, no. 52, pp. 18601–18606, 2014.
[154] J. A. Sequeira, A. C. Santos, J. Serra, C. Estevens, R. Sei¸ ca, F. Veiga, and A. J.
Ribeiro, “Subcutaneous delivery of biotherapeutics: challenges at the injection site,”
Expert Opinion on Drug Delivery, vol. 16, no. 2, pp. 143–151, 2019. [Online]. Available:
https://doi.org/10.1080/17425247.2019.1568408
[155] Takakura, Mahato, andHashida, “Extravasationofmacromolecules.” Advanced drug delivery
reviews, vol. 34, no. 1, pp. 93–108, 10 1998.
[156] D. S. Collins, M. S´ anchez-F´ elix, A. V. Badkar, and R. Mrsny, “Accelerating the development
of novel technologies and tools for the subcutaneous delivery of biotherapeutics,” Journal of
Controlled Release, vol. 321, no. January, pp. 475–482, 2020.
[157] W. F. Richter, H.-P. Grimm, M.-H. Gouy, S. Søgaard, C. Kreuzer, U. Wessels, D. Draganov,
C. Muenzer, and T. Hoche, “Subcutaneous Site-of-Absorption Study with the Monoclonal
Antibody Tocilizumab in Minipigs: Administration Behind Ear Translates Best to Humans,”
The AAPS Journal, vol. 22, no. 3, 2020.
[158] S. Hu, A. Datta-mannan, D. Z. D. Argenio, and S. Hu, “Physiologically Based Modeling
to Predict Monoclonal Antibody Pharmacokinetics in Humans from in vitro Physiochemical
Properties,” mAbs, vol. 14, no. 1, 2022.
[159] P. M. Glassman, Y. Chen, and J. P. Balthasar, “Scale-up of a physiologically-based pharma-
cokinetic model to predict the disposition of monoclonal antibodies in monkeys,” Journal of
Pharmacokinetics and Pharmacodynamics, vol. 42, no. 5, pp. 527–540, 2015.
[160] D. Mandikian, I. Figueroa, A. Oldendorp, H. Rafidi, S. Ulufatu, M. G. Schweiger, J. A.
Couch, N. Dybdal, S. B. Joseph, S. Prabhu, G. Z. Ferl, and C. A. Boswell, “Tissue Physiol-
ogy of Cynomolgus Monkeys: Cross-Species Comparison and Implications for Translational
Pharmacology,” AAPS Journal, vol. 20, no. 6, pp. 1–13, 2018.
[161] B.-S. Koo, D.-H. Lee, P. Kang, K.-J. Jeong, S. Lee, K. Kim, Y. Lee, J.-W. Huh, Y.-H. Kim,
S.-J. Park, Y. B. Jin, S.-U. Kim, J.-S. Kim, Y. Son, and S.-R. Lee, “Reference values of
hematological and biochemical parameters in young-adult cynomolgus monkey (Macaca fas-
cicularis) and rhesus monkey (Macaca mulatta) anesthetized with ketamine hydrochloride,”
Laboratory Animal Research, vol. 35, no. 1, pp. 1–6, 2019.
[162] S. Gelman, K. C. Fowler, S. P. Bishop, and L. R. Smith, “Cardiac output distribution and
regional blood flow during hypocarbia in monkeys,” Journal of Applied Physiology, vol. 58,
no. 4, pp. 1225–1230, 1985.
[163] S. Nakayama, H. Koie, C. Pai, Y. Ito-Fujishiro, K. Kanayama, T. Sankai, Y. Yasutomi, and
N. Ageyama, “Echocardiographic evaluation of cardiac function in cynomolgus monkeys over
a wide age range,” Experimental Animals, vol. 69, no. 3, pp. 336–344, 2020.
127
[164] A. Datta-Mannan, D. R. Witcher, Y. Tang, J. Watkins, and V. J. Wroblewski, “Monoclonal
antibody clearance: Impact of modulating the interaction of IgG with the neonatal Fc recep-
tor,” Journal of Biological Chemistry, vol. 282, no. 3, pp. 1709–1717, 2007.
[165] D. Hardiansyah and C. M. Ng, “Effects of the FcRn developmental pharmacology on the
pharmacokineticsoftherapeuticmonoclonalIgGantibodyinpediatricsubjectsusingminimal
physiologically-based pharmacokinetic modelling,” mAbs, vol. 10, no. 7, pp. 1144–1156, 2018.
[166] Y.A.Yeung,M.K.Leabman,J.S.Marvin,J.Qiu,C.W.Adams,S.Lien,M.A.Starovasnik,
andH.B.Lowman,“EngineeringHumanIgG1AffinitytoHumanNeonatalFcReceptor: Im-
pactofAffinityImprovementonPharmacokineticsinPrimates,” The Journal of Immunology,
vol. 182, no. 12, pp. 7663–7671, 2009.
[167] A. R. Mezo, K. A. McDonnell, C. A. Tan Hehir, S. C. Low, V. J. Palombella, J. M. Stattel,
G.D.Kamphaus, C.Fraley, Y.Zhang, J.A.Dumont, andA.J.Bitonti, “ReductionofIgGin
nonhuman primates by a peptide antagonist of the neonatal Fc receptor FcRn,” Proceedings
of the National Academy of Sciences, vol. 105, no. 7, pp. 2337–2342, 2008.
[168] B. Smith, A. Kiessling, R. Lledo-Garcia, K. L. Dixon, L. Christodoulou, M. C. Catley,
P. Atherfold, L. E. D’Hooghe, H. Finney, K. Greenslade, H. Hailu, L. Kevorkian, D. Light-
wood,C.Meier,R.Munro,O.Qureshi,K.Sarkar,S.P.Shaw,R.Tewari,A.Turner,K.Tyson,
S.West,S.Shaw,andF.R.Brennan,“Generationandcharacterizationofahighaffinityanti-
human FcRnantibody, rozanolixizumab, and the effects ofdifferent molecular formats onthe
reduction of plasma IgG concentration,” mAbs, vol. 10, no. 7, pp. 1111–1130, 2018.
[169] P. Zou, F. Wang, J. Wang, Y. Lu, D. Tran, and S. K. Seo, “Impact of injection
sites on clinical pharmacokinetics of subcutaneously administered peptides and proteins,”
Journal of Controlled Release, vol. 336, no. June, pp. 310–321, 2021. [Online]. Available:
https://doi.org/10.1016/j.jconrel.2021.06.038
[170] S.Liu,A.Verma,H.Kettenberger,W.F.Richter,andD.K.Shah,“Effectofvariabledomain
charge on in vitro and in vivo disposition of monoclonal antibodies,” mAbs, vol. 13, no. 1,
2021.
128
Abstract (if available)
Abstract
Therapeutic antibodies (mAbs) are effective in treating many diseases including cancers, auto-immune diseases and others. Currently more than 100 mAbs have been approved for clinical use in the United States and Europe with more than 600 in different stages of clinical development. Given their complex in vivo disposition mechanisms, however, antibodies can display significant differences in their pharmacokinetics (PK) behaviors, representing a challenge for antibody design, screening and development. Although traditionally antibodies have been administered via intravenous (IV) infusion, there is a rapid acceleration in the number of antibodies formulated for subcutaneous (SC) administration to improve convenience and adherence compared to IV dosing. Subcutaneous administration, however, is associated with an additional level of variability in mAb PK related to antibody differences in bioavailability and absorption kinetics. To date, no in vitro or preclinical in vivo methodology has been reported for reliably predicting the bioavailability and PK of SC administered mAbs. In this dissertation research, we developed and evaluated a comprehensive physiologically-based pharmacokinetic (PBPK) model platform for predicting disposition and absorption of mAbs following IV and SC administration that incorporates physiochemical antibody properties derived from in vitro assays and in silico structure-based metrics. The model includes the fundamental mechanisms of transport and processing that govern antibody disposition and absorption, including pinocytosis uptake, endosomal transit and processing, and paracellular transport. Relevant antibody specific properties such as molecular weight, non-specific interaction, overall charge, local charge, hydrophobicity, FcRn binding and others, as measured by in vitro assays and in silico methods, are integrated in the model. The PBPK model prediction framework was developed using available PK data from 22 clinical trials in humans involving IV administration and 12 trials with SC administered antibodies. In addition, results from IV and SC studies in cynomolgus monkeys using nonclinical antibodies and their Fc variants were also used in model development. Using these studies, we have demonstrated that antibody physiochemical properties derived from in vitro studies and in silico methods can inform underlying antibody transport processes, which in turn can be used to make model-based predictions of antibody PK. We have found that an in vitro assay of heparin chromatography relative retention time can be used to predict an antibody’s pinocytosis uptake and paracellular transport processes in the model. In addition, several in silico metrics of charge and hydrophobicity can be used to quantify antibody pre-systemic degradation, pinocytosis and paracellular transport processes. The proposed PBPK model framework also allows for predicting complete antibody concentration-time course in plasma and in tissues of interest, in addition to the clinical pharmacology end points of drug clearance and bioavailability. The framework presented in this dissertation can serve as a platform for incorporating other antibody specific information for more accurate PK predictions and contribute to the design, engineering and development of new antibody-based therapies.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
In-silico physiological based pharmacokinetic modeling of prodrugs
PDF
Understanding anti-angiogenic signaling and treatment for cancer through mechanistic modeling
Asset Metadata
Creator
Hu, Shihao
(author)
Core Title
Physiologically-based modeling to predict monoclonal antibody pharmacokinetics in humans from physiochemical properties
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Biomedical Engineering
Degree Conferral Date
2022-12
Publication Date
09/20/2022
Defense Date
09/09/2022
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
antibody biophysical characterization,antibody convective transport,antibody physiochemical properties,FcRn interaction,in vitro-in vivo prediction,nonspecific binding,OAI-PMH Harvest,PBPK model,two-pore theory
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
D'Argenio, David (
committee chair
), Finley, Stacey (
committee member
), Khawli, Leslie (
committee member
), McCain, Megan (
committee member
)
Creator Email
hushihao1@gmail.com,shihaoh@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC112013941
Unique identifier
UC112013941
Legacy Identifier
etd-HuShihao-11233
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Hu, Shihao
Type
texts
Source
20220921-usctheses-batch-983
(batch),
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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
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
Repository Email
cisadmin@lib.usc.edu
Tags
antibody biophysical characterization
antibody convective transport
antibody physiochemical properties
FcRn interaction
in vitro-in vivo prediction
nonspecific binding
PBPK model
two-pore theory