W J B C
`
`World Journal of
`Biological Chemistry
`
`Online Submissions: http://www.wjgnet.com/1949-8454office
`wjbc@wjgnet.com
`doi:10.4331/wjbc.v3.i4.73
`
`World J Biol Chem 2012 April 26; 3(4): 73-92
` ISSN 1949-8454 (online)
`© 2012 Baishideng. All rights reserved.
`
`Xiaotian Zhong, PhD, MPH, Series Editor
`Pharmacokinetics and toxicology of therapeutic proteins:
`Advances and challenges
`
`TOPIC HIGHLIGHT
`
`Yulia Vugmeyster, Xin Xu, Frank-Peter Theil, Leslie A Khawli, Michael W Leach
`
`Yulia Vugmeyster, Department of Pharmacokinetics, Dynamics,
`and Metabolism, Pfizer Inc., Andover, MA 01810, United States
`Xin Xu, Center for Translational Therapeutics, National Institutes
`of Health, Rockville, MD 20850, United States
`Frank-Peter Theil, UCB Pharma, Braine l’Aleud, B-1420 Brus-
`sels, Belgium
`Leslie A Khawli, Pharmacokinetic and Pharmacodynamic Scienc-
`es, Genentech Inc., South San Francisco, CA 94080, United States
`Michael W Leach, Drug Safety Research and Development,
`Pfizer Inc., Andover, MA 01810, United States
`Author contributions: Vugmeyster Y, Leach MW, and Xu X
`conceived and wrote this review; Theil FP and Khawli LA con-
`tributed to the conception of this review and critically reviewed
`the manuscript.
`Correspondence to: Yulia Vugmeyster, PhD, Senior Princi-
`pal Scientist, Department of Pharmacokinetics, Dynamics, and
`Metabolism, Pfizer Inc., Andover, MA 01810,
`United States. yulia.vugmeyster@pfizer.com
`Telephone: +1-978-2471404 Fax: +1-978-2472842
`Received: November 10, 2011 Revised: January 18, 2012
`Accepted: January 25, 2012
`Published online: April 26, 2012
`
`Abstract
`Significant progress has been made in understand-
`ing pharmacokinetics (PK), pharmacodynamics (PD),
`as well as toxicity profiles of therapeutic proteins in
`animals and humans, which have been in commercial
`development for more than three decades. However, in
`the PK arena, many fundamental questions remain to
`be resolved. Investigative and bioanalytical tools need
`to be established to improve the translation of PK data
`from animals to humans, and from in vitro assays to in
`vivo readouts, which would ultimately lead to a higher
`success rate in drug development. In toxicology, it is
`known, in general, what studies are needed to safely
`develop therapeutic proteins, and what studies do not
`provide relevant information. One of the major com-
`plicating factors in nonclinical and clinical programs for
`therapeutic proteins is the impact of immunogenicity.
`In this review, we will highlight the emerging science
`
`and technology, as well as the challenges around the
`pharmacokinetic- and safety-related issues in drug de-
`velopment of mAbs and other therapeutic proteins.
`
`© 2012 Baishideng. All rights reserved.
`
`Key words: Pharmacokinetics; Toxicology; Therapeutic
`proteins; Biotherapeutics; Monoclonal antibodies
`
`Peer reviewers: Tatjana Abaffy, Dr., Molecular and Cellular
`Pharmacology, University of Miami, Miller School of Medicine,
`1600 NW 10 Ave, Miami, FL 33136, United States; Conceição
`Maria Fernandes, Professor, Department of Escola Superior
`Agrária, Instituto Politécnico de Bragança, Campus de Santa
`Apolónia, Apartado 1172, 5301-854 Bragança, Portugal; Xiao-
`Feng Zheng, Professor, Department of Biochemistry and Molecu-
`lar Biology, Peking University, Beijing 100000, China
`
`Vugmeyster Y, Xu X, Theil FP, Khawli LA, Leach MW. Pharma-
`cokinetics and toxicology of therapeutic proteins: Advances and
`challenges. World J Biol Chem 2012; 3(4): 73-92 Available
`from: URL: http://www.wjgnet.com/1949-8454/full/v3/i4/
`73.htm DOI: http://dx.doi.org/10.4331/wjbc.v3.i4.73
`
`INTRODUCTION
`Biotherapeutics are therapeutic agents that are produced
`from living organisms or their products (including recom-
`binant DNA technology, biotechnological manufacturing,
`and chemical synthesis using nucleotides or amino acids)
`and include monoclonal antibodies (mAbs), antibody
`fragments, peptides, replacement factors, fusion proteins,
`oligonucleotides and DNA preparations for gene therapy,
`as well as vaccines. This is a rapidly growing class of
`therapeutics for a broad spectrum of indications, ranging
`from oncology and autoimmunity to orphan and genetic
`diseases.
`Pharmacokinetics (PK) refers to the biological pro-
`cesses determining absorption, distribution, metabolism
`and excretion (ADME) of a drug in an organism. Phar-
`
`WJBC|www.wjgnet.com
`
`73
`
`April 26, 2012|Volume 3|Issue 4|
`
`Page 1 of 20
`
`CSL EXHIBIT 1065
`CSL v. Shire
`
`

`

`Vugmeyster Y et al. Pharmacokinetics and toxicology of therapeutic proteins
`
`macodynamics (PD) refers to drug action on a living
`organism, including the pharmacologic response and the
`duration and magnitude of response observed relative
`to the concentration of the drug at an active site in an
`organism. Significant progress has been made in under-
`standing PK, PD, as well as toxicity profiles of biothera-
`peutics in animals and humans, especially for proteins
`and mAbs, which have been in commercial development
`for more than three decades.
`However, many fundamental ADME questions remain
`to be resolved. Investigative tools need to be established
`to improve the translation of PK data from animals to
`humans and from in vitro assays to in vivo readouts, which
`would ultimately lead to a higher success rate in drug de-
`velopment and provide safer and more effective drugs.
`In addition, commercial considerations, such as cost of
`goods and convenience (including less frequent dosing
`and self-administration), drive the need for a continuous
`advancement of mechanistic ADME evaluations and
`structure activity relations (SAR) for protein therapeutics
`in order to enable rational protein engineering of desired
`ADME profiles.
`The goal of this review is to highlight emerging sci-
`ence and technology, as well as challenges around the
`pharmacokinetic- and safety-related issues in drug devel-
`opment of mAbs and other therapeutic proteins.
`
`WHAT IS KNOWN
`Absorption, distribution, metabolism, and excretion
`Absorption: Unlike small molecules, which are frequent-
`ly delivered via oral administration, therapeutic proteins
`are almost exclusively administered by parenteral routes,
`such as intravenous (IV), subcutaneous (SC) or intramus-
`cular (IM) injection. Molecular size, hydrophilicity, and
`gastric degradation are the main factors that preclude
`gastrointestinal (GI) absorption of therapeutic proteins[1].
`Pulmonary delivery with aerosol formulations or dry
`powder inhalers has been used for selected proteins, e.g.,
`exubera (TM)[2,3]. Intravitreal injections have been used
`for peptides and proteins that require only local activity[4],
`as well as for antisense oligonucleotides[5].
`From the convenience standpoint, SC administration
`of therapeutic proteins is often a preferred route. In par-
`ticular, the suitability of SC dosing for self-administration
`translates into significantly reduced treatment costs. Ab-
`sorption of therapeutic proteins from the SC injection
`site tends to be slow compared to small molecules, and
`the absorption rates depend on the size of the molecule.
`For example, following SC administration, the time to
`reach the maximum systemic concentration (Tmax) in hu-
`mans for peptides is in the range of hours, while the Tmax
`for mAbs is generally several days[6-8]. For mAbs, SC bio-
`availability for currently marketed products is in the range
`of 24% to 95% in humans[1,9,10] (Table 1).
`In general, factors influencing SC absorption parame-
`ters are believed to include intrinsic subject characteristics
`for a given species (such as body weight, sex, age, activity
`
`level); species characteristics with regard to skin morphol-
`ogy and physiology (such as the presence or absence of
`the panniculus carnosus muscle in the skin, maximum SC in-
`jection volume which varies by species, catabolic capacity
`at injection site and/or in the lymphatic system, SC blood
`flow); drug substance and product characteristics [pres-
`ence of an Fc (see below), target interactions, charge, for-
`mulation, dose concentration, total dose]; and mode of
`administration (injection site, injection time, depth of in-
`jection, anesthesia status), as discussed in references[1,9-14].
`However, surprisingly little is known about the mecha-
`nisms and pathways of SC absorption and which path-
`ways are affected by a particular factor described above.
`The emerging science and issues around the mechanisms
`and factors involved in SC absorption that are not known
`are further discussed in the “WHAT IS NOT KNOWN”
`section.
`
`Distribution: Tissue distribution of therapeutic proteins
`usually is limited because of the size of the molecules,
`which is in contrast to small molecule drugs that tend to
`have higher tissue penetration. In addition to size, other
`factors that influence the tissue distribution of a thera-
`peutic protein include the physical and chemical prop-
`erties (e.g., shape and charge), binding properties (e.g.,
`receptor-mediated uptake), the route of administration
`(e.g., IV vs SC, formulation), and the production process
`(which may affect post-translational modifications, such
`as glycosylation). These factors can be modulated via ra-
`tional design to modulate tissue penetration properties of
`a biotherapeutic molecule. For example, a modeling anal-
`ysis of the effects of molecular size and binding affinity
`on tumor targeting was conducted to guide the design of
`new therapeutic protein drugs[15,16]. A similar approach
`was used to engineer a novel human IL-2 analog that an-
`tagonizes the IL-2 receptor[17]. Tissue- or target-specific
`delivery of therapeutic biologics is a challenging, yet a
`very attractive area for pharmaceutical research.
`For mAbs and other large therapeutic proteins, the
`reported volume of distribution after IV administration
`is close to the plasma volume, suggesting limited distribu-
`tion into tissues[18]. However, tissue distribution studies
`with radiolabeled mAbs indicate that many tissues are
`exposed to mAbs, but at lower concentrations than usu-
`ally seen in systemic circulation[19]. Despite the limited
`tissue penetration, large biotherapeutics, such as mAbs,
`often do have efficacy even in cases when the site of ac-
`tion is believed to be the tissue, indicating that it is pos-
`sible to design a therapeutic regimen such that the tissue
`exposure is adequate to modulate the target at the site of
`action. The therapeutic areas for tissue-acting biothera-
`peutics are diverse and examples for autoimmunity and
`oncology are presented in recent reviews[20,21].
`Once in the tissue vasculature, the common transport
`mechanisms for proteins from systemic circulation across
`capillary endothelial cells and into tissues are listed in
`Table 2[22]. The uptake of therapeutic proteins into cells
`may be carried out via receptor-mediated transporters (e.g.,
`
`WJBC|www.wjgnet.com
`
`74
`
`April 26, 2012|Volume 3|Issue 4|
`
`Page 2 of 20
`
`

`

`Vugmeyster Y et al. Pharmacokinetics and toxicology of therapeutic proteins
`
`Table 1 Examples of proteins and peptides administered subcutaneously[10]
`
`INN/BAN (description)
`Buserelin acetate
`(LH-releasing hormone analog)
`
`Trade name MW (kDa)
`Suprefact
`1.30
`
`Absolute bioavailability1
`Human: 70%
`
`SC animal models used in drug development
`Pharm: rat, hamster, guinea pig, rabbit, dog and
`monkey
`Tox: mouse, rat, rabbit and dog
`Pharm: rat and dog
`PK: mouse, rat, rabbit and dog
`Tox: mouse, rat, rabbit and dog
`Pharm: rat, rabbit, dog and pig
`PK: rat and dog
`Tox: rat, rabbit and dog
`Pharm: rat and dog
`PK: rat and dog
`Tox: mouse, rat, rabbit and dog
`Pharm: mouse, rat, guinea pig, rabbit and dog
`PK: rat and dog
`Tox: mouse, rat, guinea pig, rabbit and dog
`Pharm: mouse, rat, rabbit and monkey
`PK: rat, rabbit, dog and monkey
`Tox: rat, dog, rabbit and monkey
`Pharm: monkey
`PK: monkey
`Tox: rabbit and monkey
`Pharm: rat
`PK: rat and monkey
`Tox: mouse, rat, dog and monkey
`Pharm: mouse and monkey
`PK: rat and monkey
`Tox: monkey
`Pharm: rat and monkey
`PK: rat and monkey
`Tox: mouse, rat, rabbit and monkey
`Pharm: mouse, rat and dog
`PK: mouse, rat and monkey
`Tox: rat and monkey
`Pharm: mouse and monkey
`PK: mouse, rat, rabbit and monkey
`Tox: mouse, rat, rabbit and monkey
`Pharm: mouse
`PK: rat and monkey
`Tox: mouse, rat and monkey
`PK: rat and monkey
`Tox: monkey
`Pharm: mouse, rat and monkey (marmoset)
`PK: mouse and monkey
`Tox: mouse and monkey
`PK: monkey
`Tox: rabbit and monkey
`Pharm: monkey
`PK: mouse and monkey
`Tox: monkey
`PK: monkey
`Tox: mouse and monkey
`Pharm: monkey
`PK: monkey
`Tox: monkey
`Pharm: mouse
`PK: mouse, rat and monkey
`Tox: mouse, rat, rabbit and monkey
`Pharm: mouse and monkey
`PK: mouse, rat and monkey
`Tox: monkey
`
`Pramlintide acetate
`(amylin analog)
`
`Insulin lispro
`(insulin analog)
`
`Insulin glulisine
`(insulin analog)
`
`Insulin glargine
`(insulin analog)
`
`Mecasermin
`(IGF-1)
`
`IFNβ-1b
`(cytokine)
`
`Somatropin
`(GH)
`
`IFNβ-1a
`(cytokine)
`
`PEG-IFNα-2b
`(cytokine variant)
`
`Pegfilgrastim
`(PEG-G-CSF)
`
`Pegvisomant
`(PEG-GH)
`
`PEG-IFNα-2a
`(cytokine variant)
`
`Certolizumab pegol
`(PEG-anti-TNFα Fab' fragment)
`Canakinumab
`(anti-IL-1β mAb)
`
`Adalimumab
`(anti-TNF mAb)
`Omalizumab
`(anti-IgE mAb)
`
`Golimumab
`(anti-TNF mAb)
`Ustekinumab
`(anti-p40 mAb)
`
`Pegasys
`
`Cimzia
`
`Ilaris
`
`Humira
`
`Xolair
`
`Simponi
`
`Stelara
`
`60
`
`91
`
`145
`
`148
`
`149
`
`150
`
`150
`
`150
`
`251
`
`Etanercept
`(TNF receptor-Fc-IgG1 fusion protein)
`
`Enbrel
`
`Rilonacept
`(IL-1 inhibitor, fusion protein)
`
`Arcalyst
`
`Symlin
`
`3.95
`
`Human: 30 to 40%
`
`Humalog
`
`5.81
`
`Human: 55 to 77%
`
`Apidra
`
`Lantus
`
`5.82
`
`6.06
`
`Increlex
`
`7.65
`
`Betaseron
`
`18.5
`
`Human: about 70%
`Dog: 42%
`Rat: 96%2
`Precipitates in skin-slow
`uptake in human, dog and rat
`
`Human: about 100%
`Rabbit: 47%
`Rat: 38 to 57%
`Human: 50%
`Monkey: 31 to 44%
`
`Nutropin
`
`22
`
`Human: 81%
`
`Rebif
`
`22.5
`
`PEG-Intron
`
`Neulasta
`
`31
`
`39
`
`Somavert
`
`42, 47 and 523
`
`Human: 6 to 62%
`Monkey: 12 to 38%
`Rat: 16%
`Monkey: 57 to 89%
`Rat: 43 to 51%
`
`Monkey: 49 to 68%
`Rat: < 10% to 30%
`
`Human: 49 to 65%
`Monkey: 70 to 81%
`Mouse: 45 to 73%
`Human: 61 to 80%
`
`Human: 76 to 88%
`Rat: 24 to 34%
`Human: 63 to 67%
`Monkey: 60%
`
`Human: 64%
`Monkey: 96%
`Human: 53 to 71%
`Monkey: 64 to 104%
`Mouse: 90%
`Human: 53%
`Monkey: 77%
`Human: 24 to 95%
`Monkey: 97%
`
`Human: 76%
`Monkey: 73%
`Mouse: 58%
`Human: 43%
`Monkey: 70%
`Rat: 60%
`Mouse: 78%
`
`1Systemic dose following subcutaneous (SC) injection relative to systemic dose following intravenous injection; 2Assumes linearity of AUC/dose;
`3Product is a mixture of three distinct protein variants. GH: Growth hormone; LH: Luteinizing hormone; MW: Molecular weight; Pharm: Pharmacology;
`PK: Pharmacokinetics; Tox: Toxicology (including safety pharmacology); INN: International nonproprietary name; BAN: British approved name; SC:
`Subcutaneous; PEG: Polyethylene glycol; IFN: Interferon; TNF: Tumor necrosis factor; IL: Interleukin.
`
`WJBC|www.wjgnet.com
`
`75
`
`April 26, 2012|Volume 3|Issue 4|
`
`Page 3 of 20
`
`

`

`Vugmeyster Y et al. Pharmacokinetics and toxicology of therapeutic proteins
`
`Table 2 Transport mechanisms for proteins from systemic
`circulation across capillary endothelia[22]
`
`Type of capillary
`endothelium
`
`Barrier/transport
`mechanism
`
`Continuous
`(non-fenestrated)
`
`Fenestrated
`
`Discontinuous
`(sinusoidal)
`
`Basal lamina
`membrane
`supported by
`collagen
`Large pores,
`open fenestrae,
`intracellular
`junctions, basal
`lamina
`
`Large pores
`(fenestrae),
`pinocytotic
`vesicles
`
`Typical tissues
`
`Particle size
`subject to
`passage
`50-110 nm
`
`Muscle, central
`nervous system,
`bone, skin,
`cardiac muscle
`50-800 nm Renal glomeruli,
`intestinal villi,
`synovial tissue,
`endocrine
`glands, choroid
`plexus (brain)
`1000-10 000 nm Liver, spleen,
`bone marrow,
`postcapillary
`venules of
`lymph nodes
`
`Fc receptors, often leading to recycling of the molecule)
`or other internalization processes, such as endocytosis
`or pinocytosis (often leading to degradation of the mol-
`ecule). Target-mediated tissue distribution has also been
`reported for some mAbs[23,24]. High drug concentrations
`in kidney and liver have been reported for peptides, low
`molecular proteins, and oligonucleotides[25,26]. Upon tis-
`sue uptake, metabolism/catabolism of protein drugs will
`occur in tissues before the remnants of the molecules are
`excreted from the body as smaller peptides and amino
`acid degradants, or they are recycled for synthesis into
`other proteins in the body.
`The high vascular concentrations of the test article
`provide a potential source for interference of tissue drug
`concentrations, and should be considered when interpret-
`ing biodistribution data for therapeutic proteins. To mini-
`mize vascular interference, whole body perfusion is often
`performed before tissue analysis in biodistribution studies
`of therapeutic proteins, especially for rodents[19]. Other
`methods to correct for the contribution of residual drug
`in tissue blood vessels, such as the use of radiolabeled
`erythrocytes or the use of dual isotopes of 125I- and/or
`131I-labeled proteins, have also been applied[27,28].
`
`Metabolism/Catabolism: Therapeutic proteins are re-
`moved from circulation or interstitial fluid via several
`pathways: degradation by proteolysis, Fcγ receptor-me-
`diated clearance, target-mediated clearance, nonspecific
`endocytosis, and formation of immune-complexes (ICs)
`followed by complement- or Fc receptor-mediated clear-
`ance mechanisms. While proteolysis occurs widely in the
`body, its kinetics and mechanistic details are poorly un-
`derstood, especially for large therapeutic proteins such as
`mAbs. In vitro incubations with plasma, liver and kidney
`homogenates have been used for peptides to facilitate the
`selection of leads in discovery research; however the in
`vitro in vivo correlations for such an approach remain to
`be established (see additional discussions in the “WHAT
`IS NOT KNOWN” section). Once taken up into cells, a
`
`biotherapeutic may be metabolized to peptides or amino
`acids. This may occur in circulation by circulating phago-
`cytic cells or by their target antigen-containing cells, or
`may occur in tissues by various cells. For molecules with
`an Fc (including therapeutic mAbs, endogenous Abs, and
`fusion proteins), binding of the Fc domain to Fc gamma-
`receptors may result into the internalization and subse-
`quent degradation by lysosomes in the reticuloendothelial
`system (e.g., macrophages and monocytes)[1,29,30].
`Alternatively, molecules with an Fc may be protected
`from degradation by binding to protective receptors [i.e.,
`the neonatal Fc-receptor (FcRn)] in endothelial cells,
`explaining the long half-lives (up to 4 wk) of these pro-
`teins. The following references provide excellent reviews
`on the scholarship in this field[1,31-33]. The FcRn receptor
`is a 52-kDa membrane-bound heterodimeric glycopro-
`tein comprising a heavy chain and a light chain (beta2-
`microglobulin). Structurally, the FcRn receptor varies
`only subtly from conventional major histocompatibility
`complex (MHC) class I proteins protein. Its physiological
`function and expression in different tissues have been de-
`scribed[1,31-33]. In particular, the FcRn receptor, located in
`endosomes of endothelial cells, is known to bind to the
`Fc domain of IgG at pH 6.0-6.5, but only weakly or not
`at all at pH 7.0-7.5. This unique property allows FcRn
`to protect Fc-containing molecules from degradation by
`binding to them in acidic endosomes after uptake into en-
`dothelial cells via nonspecific endocytosis or fluid-phase
`pinocytosis. The IgG-FcRn complex is then transported
`back to the cell surface and disassociated at physiologi-
`cal pH, releasing the intact Fc-containing molecule back
`to the circulation. In contrast, Fc-containing molecules
`that are not bound to FcRn are degraded to amino acids
`by lysosomes in the cells. The correlation between FcRn
`binding affinity and systemic half-live has been investi-
`gated for a number of mAbs[33-43]. While the contribu-
`tion of FcRn in prolonging half-lives of Fc-containing
`proteins is well recognized, other factors may also play a
`role in determining the elimination rate of these proteins,
`because the binding affinity to FcRn alone could not ex-
`plain the variation of half-lives observed for all approved
`Fc-containing therapeutic proteins (see additional discus-
`sions in the “WHAT IS NOT KNOWN” section).
`Target-mediated clearance is one of main causes of
`non-linear elimination kinetics. Upon binding to target
`on cells, the therapeutic proteins are internalized into
`the cells and subjected to degradation in lysosomes. For
`targets such as the endothelial growth factor receptor
`(EGFR), target-mediated clearance is the predominant
`clearance pathway at clinical doses, as illustrated by the
`nonlinear kinetic characteristics of cetuximab[44]. Target-
`mediated clearance could be demonstrated by comparing
`the disposition kinetics between normal healthy animals
`vs animals over expressing the target[23,45]. PK/PD mod-
`els are usually established to describe saturable kinetic
`profiles that are associated with the target-mediated clear-
`ance in humans[46-50]. Formation of anti-drug antibodies
`(ADA) followed by formation of biotherapeutic/ADA
`ICs, is another main cause for the non-linear elimination
`
`WJBC|www.wjgnet.com
`
`76
`
`April 26, 2012|Volume 3|Issue 4|
`
`Page 4 of 20
`
`

`

`Vugmeyster Y et al. Pharmacokinetics and toxicology of therapeutic proteins
`
`kinetics, including time-dependent clearance, which is
`often evidenced by a rapid concentration drop in the PK
`profiles (discussed below).
`It should also be noted that many general factors that
`contribute to inter-subject variability in PK profiles for
`small molecule compounds may also apply to therapeutic
`proteins. These factors can be categorized into intrinsic
`factors (such as age, sex, body weight, activity level, renal
`and hepatic impairment) and to extrinsic factors (e.g.,
`concomitant drugs, diet) and there are several examples
`in the literature describe the role of some of these fac-
`tors for mAbs[51,52].
`
`Excretion: Renal excretion is thought to play an important
`role in the elimination of protein degradation products and
`low molecular weight (MW) biologics (MW < 30 kDa). The
`process of renal filtration, transport, and metabolism of
`low-MW proteins has been well discussed in literature[26].
`Proteins are hindered at the glomerular filter in proportion
`to their molecular size, structure, and net charge. However,
`the mechanisms of reabsorption of peptides and proteins
`in the kidney need further investigation.
`When radiolabeled mAbs or Fc fusion proteins were
`used in animal disposition studies, a majority of the ra-
`dioactive dose was recovered in the urine[19]. The sodium
`dodecyl sulfate polyacrylamide gel electrophoresis (SDS-
`PAGE) analysis and trichloroacetic acid (TCA) soluble
`counts indicated that the radioactive materials in urine
`were associated with low molecular fragments, suggesting
`that the excretion of intact parent drug was negligible. Bili-
`ary excretion of therapeutic proteins, such as insulin and
`epidermal growth factor has been reported[53]. It appeared
`that proteins were subjected to degradation in the liver,
`and the degradants were subsequently excreted into bile[54].
`It has also been reported that plasma protein binding
`plays an important role in the tissue distribution of several
`new modalities of biologic therapeutics (e.g., oligomers),
`resulting in altered excretion profiles. Modification of the
`lipophilicity of the backbone for oligomers has been used
`to prolong the in vivo half-life by increasing plasma protein
`binding in order to reduce the renal excretion[5].
`
`Anti-drug antibodies
`Immunogenicity, specifically formation of ADAs, is one
`of the major complicating issues in nonclinical and clinical
`programs for therapeutic proteins. There are many factors
`that contribute to the ability of a therapeutic protein to
`elicit ADA production[55]. Intrinsic factors affecting im-
`munogenicity are protein sequence (including similarity
`to endogenous proteins and the presence of T and B cell
`epitopes), post-translational modification (glycosylation,
`oxidation), and tertiary structure (including aggregation
`propensity). Extrinsic factors include the route, dose, and
`type of formulation (that may affect aggregation), produc-
`tion process (that may affect both aggregation and post-
`translational modifications), impurities, subject character-
`istic (disease population, inflammation status, concomitant
`medications), as well as drug pharmacology (specifically
`
`related to immunosuppression). All of the above factors
`are thought to contribute to variability in ADA responses
`observed across the biologic modalities, species, and sub-
`jects.
`ADA may affect both the PK and PD profiles of
`therapeutic proteins by introducing additional (IC-depen-
`dent) clearance and distribution pathways and by modu-
`lating biological activity, including neutralization of the
`test article. In the case of replacement proteins, the ADA
`can result in neutralization of the endogenous protein as
`well, as has been described with erythropoietin[56,57] and
`factor Ⅷ[58] replacement factors.
`When a drug/ADA immune complex is formed, the
`clearance of a therapeutic protein within the IC may be
`much faster compared to unbound drug, explaining a
`rapid concentration drop in PK profiles. It is believed
`that the clearance of IgG-containing ICs (which would
`include a drug bound to ADA) occurs primarily in the
`liver[59-63]. This can be facilitated by red blood cells, which
`can bind ICs in the circulation (via the complement re-
`ceptor 1) and deliver them to the tissue macrophages
`of the mononuclear phagocyte system (such as Kupffer
`cells) in the liver[60]. Because the extent and rate of IC
`formation varies among human subjects, the IC-related
`clearance could be considered as a major contributor to
`the inter-subject variability in clinical and nonclinical PK
`profiles for therapeutic proteins.
`Under some circumstances, ICs (including ADA-bound
`therapeutic proteins) might not be transported to the liver
`and cleared properly[59]. Factors that could influence this
`phenomenon include the IC characteristics (such as nature
`and quantity of the antigen and the antibody response, in-
`cluding antibody isotype and antigen/antibody stoichiom-
`etry) and the state of the systems involved in IC clearance
`and transport (for example expression of complement
`components, complement receptors, liver phagocytic
`system, red blood cells). In these cases, the deposition of
`circulating complement-fixing IC in various organs (such
`as the kidney) is observed, with important consequences
`for safety assessments of biotherapeutics. The impact
`of ADA on toxicology and PK-PD of therapeutic pro-
`teins is further discussed below in the “WHAT IS NOT
`KNOWN” section.
`
`Glycosylation
`Glycosylation, most frequently at asparagine residues (“N-
`linked”) and at serine or threonine residues (“O-linked”),
`is the most common, complex, and heterogeneic post-
`translational modification that occurs on endogenous and
`therapeutic proteins. Recent reviews by Sola et al[64] and
`Li et al[65] summarize the current knowledge in this field.
`The inter- and intra-product heterogeneity in glycosylation
`profile can arise from the variability in glycan type and
`structure (including degree of branching), the site of at-
`tachment, and the degree of occupancy and can, in part,
`be controlled by the production system and conditions
`(such cell-type, cell culture media, and purification pro-
`cess). The glycosylation of proteins is important from the
`
`WJBC|www.wjgnet.com
`
`77
`
`April 26, 2012|Volume 3|Issue 4|
`
`Page 5 of 20
`
`

`

`Vugmeyster Y et al. Pharmacokinetics and toxicology of therapeutic proteins
`
`ADME and efficacy standpoint, because improperly gly-
`cosylated proteins, whether endogenous or exogenously
`produced biotherapeutics, may be rapidly cleared from the
`circulation by specific receptor-based mechanisms, such as
`high mannose receptor or asialoglycoprotein receptor, and
`because glycosylation may directly affect biological activ-
`ity of a biotherapeutic. For many approved protein drugs,
`clinical efficacy depends on proper glycosylation[64,65]. The
`ongoing research in the field is discussed below in the
`“WHAT IS NOT KNOWN” section.
`
`Toxicology
`In the past three decades of the development of bio-
`therapeutics, the toxicity of the molecules and the meth-
`ods and studies by which to measure such toxicity have
`been refined. In some cases (for example, for mAbs and
`fusion proteins that block cytokine pathways), no effects
`may be seen. When effects are seen in toxicity studies
`with biotherapeutics, in almost all cases the findings have
`been linked to target-mediated effects. In some cases
`these target-mediated effects may be undesirable, and are
`considered to be a result of exaggerated pharmacology.
`In this regard, they may not be considered to represent
`primary toxicity.
`There are many examples of on-target pharmacologic
`effects that can be undesirable. For example, a variety of
`tumor necrosis factor-α (TNF) inhibitors are used to treat
`inflammatory and autoimmune diseases, including rheu-
`matoid arthritis, psoriasis, inflammatory bowel disease,
`and multiple sclerosis. B cell depletion therapies are used
`for the treatment of B cell tumors, and for inflammatory
`and autoimmune diseases. These therapies have proven to
`provide life-altering benefits to many patients. However,
`infections related to immunosuppression, which can be
`considered exaggerated pharmacology, have occurred in a
`small number of patients[66-70], although not all studies have
`demonstrated such a risk relative to treatment with non-
`biologic regimens[71]. When they occur, these infections
`may be associated with latent viral or bacterial infections
`that recrudesce following the immunosuppression, or in-
`fections by organisms that are normally not pathogens in
`humans, and include Mycobactrium tuberculosis, atypical my-
`cobacterial infections, hepatitis B, and John Cunningham
`virus [JCV, which causes progressive multifocal leucoen-
`cephalopathy (PML)]. It should be noted that because pa-
`tients often receive multiple immunosuppressive therapies
`as well as have various diseases, identifying clear casual as-
`sociations between infections and specific biologics can be
`challenging[67].
`Another example of exaggerated pharmacology comes
`from the erythropoiesis stimulating agents such as erythro-
`poietins[72]. At higher doses, such as those used in toxicity
`studies, the animals develop polycythemia, chronic blood
`hyperviscosity, vascular stasis, thromboses, increased pe-
`ripheral resistance and hypertension, which can be fatal.
`Similar adverse effects have been suspected in athletes
`who are seeking supraphysiological hematocrits[73]. How-
`ever, in an anemic person or animal, the increased red cell
`
`mass can be beneficial. The concept that one scientist’s
`pharmacology is another scientist’s toxicity is indeed well
`represented in the field of biotherapeutics. That said,
`very recently there has been some concern raised regard-
`ing off-target effects with biotherapeutics, and this is cur-
`rently a topic of discussion within the biopharmaceutical
`industry (a recent case study is reported by Everds et al
`at the Toxicologic Pathology Annual Symposium, 2011;
`Abstract 04).
`
`Species selection
`Regulations require the use of one rodent and one non-
`rodent animal species in general toxicity studies to assess
`the toxicity of biotherapeutics, as long as the species are
`relevant[74,75]. The selection of species for most biothera-
`peutics should primarily be based by the presence of
`pharmacological activity. The specificity and biological
`activity of the biotherapeutic is typically first evaluated
`in vitro. This can be done using binding assays and cell-
`based assays. Ideally the biotherapeutic will be specific
`and bind only to the intended target. However, from a
`practical standpoint, only a limited number of targets can
`be evaluated, and there is always a chance for unintended
`binding to untested targets to occur. The in vitro activ-
`ity of the biotherapeutic on the human target should be
`compared with the activity in commonly used toxicity
`species. Ideally, the activity in the animal species is simi-
`lar to that observed in humans. If so, it suggests these
`species may be relevant for toxicity studies. However, in
`many cases the activity in animals is lower, and sometimes
`absent, especially in rodents. When pharmacologic activ-
`ity is not present in a species, they should generally not
`be used for toxicity studies (although they may still have
`value for PK studies). Whenever possible,