Quantification and Modeling of Quality Attributes of Therapeutic Monoclonal Antibodies
By determining and predicting the rate constant and proportion of post-translational modification of therapeutic monoclonal antibodies in vivo, using pharmacokinetic models, the problem of difficulty in monitoring and predicting changes in vivo modification in the prior art is solved, and the stability and safety of antibodies are achieved Effective assessment and improvement of sexuality.
Patent Information
- Application Number
- CN201980007440.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2018-02-01
- Filing Date
- 2019-01-31
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2039-06-27
AI Technical Summary
The prior art is difficult to effectively monitor and predict changes in post-translational modification of therapeutic monoclonal antibodies in vivo, especially in multi-dose regimens, affecting the stability and safety of the drug.
By determining the in vivo rate constant of target post-translational modification, predicting the serum concentration and post-translational modification ratio of the antibody, using a two-chamber pharmacokinetic model to describe the serum concentration changes of the antibody, and superimposing a single administration model under multiple administration conditions, predicting In vivo concentration of the antibody and subject's exposure to antibody variants.
Accurate prediction and evaluation of post-translational modification of therapeutic monoclonal antibodies is achieved, helping to design drugs with enhanced in vivo stability and improving the safety and efficacy of drugs.
Smart Images

Figure CN111902874B_ABST
Abstract
Description
[0001] Sequence Listing
[0002] This application incorporates by reference a sequence listing, which is submitted in computer readable form as 10421WO01-Sequence.txt, created on January 31, 2019, and containing 514 bytes. Field of the Invention
[0003] The present invention relates to biomedicine and to the prediction and modeling of in vivo post-translational modifications of therapeutic antibodies. Background of the Invention
[0005] Therapeutic monoclonal antibodies (mAbs) are heterogeneous molecules produced in mammalian cells with many product variants, including variants produced by post-translational modifications (PTMs). Variants produced via PTMs may occur during the production, purification, storage, and administration of mAbs throughout their life cycle. These variants or product-related modifications are also referred to as product quality attributes (PQAs). Controlling PQA within predefined acceptance criteria is critical to the biopharmaceutical industry because it ensures that product quality is consistent and reduces the potential impact on drug safety and efficacy. Modifications that occur during drug production and storage can usually be reliably monitored and controlled. However, due to the significant difference in the environment of blood flow and formulation buffers, additional modifications may occur after drug administration. The in vivo modification of mAbs is usually difficult to monitor and is usually less studied. The assessment and prediction of PTM changes in vivo will not only help to understand the criticality of quality attributes for product risk assessment, but will also help product design and R&D teams design candidate mAb drugs with enhanced in vivo stability.
[0006] The U.S. Food and Drug Administration's guidance to industry recommends that sponsors should evaluate the sensitivity of therapeutic proteins to modifications within the in vivo context (see Guidance for Industry, Immunogenicity Assessment for Therapeutic Protein Products. 2014). As a result, the in vivo behavior of many PQAs has been studied in animal or human samples, including deamidation (see, e.g., Huang et al., Analytical chemistry 2005;77:1432-9; Ouellette et al., mAbs 2013;5:432-44; Yin et al., Pharmaceutical research 2013;30:167-78; Li et al., mAbs 2016:0; Li et al., mAbs 2016:0), oxidation (see, e.g., Yin et al., Pharmaceutical research 2013;30:167-78; Li et al., mAbs 2016:0; Li et al., mAbs 2016:0), glycation (see, e.g., Goetze et al., Glycobiology 2012;22:221-34), glycosylation (see, e.g., Li et al., mAbs 2016:0; Li et al., mAbs 2016:0). 2016:0; Goetze et al., Glycobiology 2011;21:949-59; Alessandri et al., mAbs 2012;4:509-20.), disulfide (see, e.g., Li et al., mAbs 2016:0; Liu et al., The Journal of biological chemistry 2008;283:29266-72), N-terminal pyroglutamate (see, e.g., Yin et al., Pharmaceutical research 2013;30:167-78; Li et al., mAbs 2016:0; Li et al., mAbs 2016:0; Liu et al., The Journal of biological chemistry 2011;286:11211-7) and C-terminal lysine removal (see, e.g., Li et al., mAbs 2016:0; Cai et al., Biotechnology and bioengineering 2011; 108:404-12). Affinity purification is often used to extract therapeutic mAbs from serum samples to reduce interference from endogenous protein background in the analysis (see, e.g., Li et al., mAbs 2016:0; Li et al., mAbs 2016:0).For studies in animals, anti-human Fc antibodies or antigens can be easily used as capture agents; for human studies, antigens or antibodies specifically recognizing the unique complementary determining region (CDR) of therapeutic mAbs can be used as capture agents. After affinity purification, liquid chromatography tandem mass spectrometry (LC-MS / MS) peptide localization is usually used to quantify the relative abundance of PTM product variants. Recently, in vivo PTM quantification has been combined with pharmacokinetic (PK) models to evaluate the formation and elimination of PTM product variants, and quantitatively evaluate the exposure of subjects to PQA in single-dose regimens (see, for example, Li et al., mAbs 2016: 0; Li et al., mAbs 2016: 0; Goetze et al., mAbs 2010; 2: 500-7) to help establish the criticality of PQA for product risk assessment. However, no attempt has been made to simulate the in vivo progression of PTMs in multiple-dose regimens and the exposure of subjects to PQA to quantitatively evaluate the criticality of PQA modeling. Summary of the invention
[0007] In one aspect, the present invention provides a method for predicting the in vivo serum concentration of an antibody having a post-translational modification of interest after a single administration of the antibody, wherein the method comprises (a) using an in vivo rate constant determined for the post-translational modification to predict the percentage of the antibody having the post-translational modification of interest; and (b) multiplying the predicted percentage of the antibody having the post-translational modification of interest by the in vivo concentration of the antibody to determine the concentration of the antibody having the post-translational modification of interest. In some cases, the following equation is used in the calculation: C PTM (t) = C(t)·P(t), where C PTM (t) is the serum concentration of the antibody having the post-translational modification of interest; C(t) is the serum concentration of the antibody; and P(t) is the percentage of the post-translational modification of interest.
[0008] In some embodiments, the in vivo rate constant of the post-translational modification is determined by quantifying the percentage of the target post-translational modification over time and fitting the quantified percentage of the target post-translational modification to the equation Where P PTM (t) is the ratio of the post-translational modification over time; P0 is the initial post-translational modification percentage; and k PTM is the post-translational modification rate constant.
[0009] In some embodiments, the method further comprises determining the exposure of the subject to the antibody having the target post-translational modification. In some cases, determining the exposure of the subject to the antibody having the target post-translational modification comprises determining the concentration (C PTM(t)) of the area under the curve (AUC).
[0010] In some embodiments, wherein the serum concentration of the antibody is described by a two-compartment pharmacokinetic model equation: C(t) = Ae -αt +Be -βt , where α and β are first-order confounding constants, and A and B are confounding coefficients, respectively.
[0011] In some embodiments, the method further comprises predicting the in vivo serum concentration of the antibody after multiple administrations of the antibody having the target post-translational modification. In some cases, predicting the in vivo serum concentration of the antibody after multiple administrations of the antibody having the target post-translational modification comprises superimposing multiple single-administration models, wherein the serum concentration of the antibody at each administration is individually described by a two-compartment pharmacokinetic model equation as C(t)=Ae -αt +Be -βt , where α and β are first-order confounding constants, and A and B are confounding coefficients, respectively.
[0012] In various embodiments of the method, the antibody comprises a recombinant therapeutic monoclonal antibody. In some cases, the post-translational modification comprises a product quality attribute that can affect drug stability, safety and / or efficacy.
[0013] In some embodiments, the methods predictively monitor the post-translational modification profile of a monoclonal antibody.
[0014] In some embodiments, the method further comprises creating a predictive model of the in vivo progression of the post-translational modification and the subject's exposure to the product quality attribute.
[0015] In some embodiments, the method further comprises modeling the correlation between the PK concentration and the ratio of the post-translationally modified variant.
[0016] In some embodiments, the method further comprises quantitatively assessing and predicting the subject's exposure to a representative post-translational modification in both a single-dose regimen and a multiple-dose regimen.
[0017] In some embodiments, the method further comprises predicting the highest and lowest post-translational modification levels observed in the multiple-dose regimen.
[0018] In another aspect, the present invention provides a method for predicting exposure of a subject to a variant of a recombinant therapeutic monoclonal antibody after administration to a subject, wherein the method comprises (a) identifying a rate constant for a post-translational modification (PTM) process that results in formation of the variant; (b) calculating a predicted proportion of the variant over time; (c) based on the predicted proportion of the variant, calculating a predicted serum concentration of the variant in the subject over time; and (d) based on the predicted serum concentration of the variant, predicting exposure of the subject to the variant after administration of the antibody.
[0019] In some embodiments of the method, the calculation of the predicted ratio of the variants over time is performed via the following equation: PTM (t)=1–(1–P0)·e -k PTM t , where P PTM (t) is the ratio of the variant over time, P0 is the initial level of the variant at the time of administration; k PTM is the rate constant of the PTM process, and t is the time.
[0020] In some embodiments of the method, the calculation of the predicted serum concentration of the variant in the subject over time is performed via the following equation: PTM (t) = C(t)·P PTM (t), where C PTM (t) is the serum concentration of the variant, C(t) is the serum concentration of the antibody, and P PTM (t) is the proportion of the variant at time t.
[0021] In some embodiments, the method further comprises administering the antibody to the subject.
[0022] In some embodiments of the methods, the antibody is a mAb.
[0023] In some embodiments, the method further comprises determining the total in vivo concentration of the antibody in the subject.
[0024] In another aspect, the present invention provides a method for predicting exposure of a subject to a variant of a recombinant therapeutic monoclonal antibody after administering multiple doses of the antibody to a subject, the method comprising: (a) performing the method discussed immediately above; (b) calculating, based on the dosing interval, a predicted ratio of the variant before and after each subsequent dose of the antibody; (c) calculating a predicted serum concentration of the variant in the subject based on the predicted ratio of the variant at each dosing interval; and (d) predicting exposure of the subject to the variant over the course of two or more doses of the antibody based on the predicted serum concentration of the variant.
[0025] In various embodiments of the methods discussed above or herein, the post-translational modification includes one or more of deamidation, oxidation, glycation, disulfide formation, N-terminal pyroglutamate formation, C-terminal lysine removal, and high mannose glycosylation.
[0026] In some embodiments, the method further comprises administering the antibody to the subject.
[0027] In some embodiments of the methods, the antibody is a mAb.
[0028] In some embodiments, the method further comprises determining the total in vivo concentration of the antibody in the subject.
[0029] Any of the various aspects or embodiments discussed above or herein may be combined in alternative embodiments, each of which is encompassed within the scope of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0030] Figure 1 Set of traces showing recovery of MAB1 protein by affinity purification. Recovery of affinity purification of MAB1 was >99.5%, with <0.5% total MAB1 detected in the flow-through. Calculation of percent recovery is described in the Materials and Methods section of Example 1. Each sample was run with a known amount of a high molecular weight standard (shown as "Upper Marker") as an internal normalization control. The ratio of the MAB1 peak area in each sample to the Upper Marker peak area was calculated to correct for run-to-run variability.
[0031] Figure 2A and Figure 2B is an exemplary MS / MS spectrum for peptide identification ( Figure 2A ) and exemplary extracted ion chromatograms for peptide peak integration and PTM quantification ( Figure 2B ). Figure 2A) MS / MS spectrum of an exemplary Met-oxidized peptide, DTLMISR (SEQ ID NO: 1) (upper panel); and MS / MS spectrum of an exemplary wild-type peptide, DTLMISR (SEQ ID NO: 1) (lower panel). ( Figure 2B ) Extracted ion chromatogram of Met-oxidized peptide, DTLMISR (SEQ ID NO: 1) (upper panel); and extracted ion chromatogram of wild-type peptide, DTLMISR (SEQ ID NO: 1) (lower panel).
[0032] Figure 3A and Figure 3B It is shown that from a single-dose PK study ( Figure 3A ) and multiple dose PK studies ( Figure 3B ) is a graph showing the relative abundance of deamidation at each of the three Asn sites in the Fc region of MAB1. ( Figure 3A ) In the single-dose study, the relative abundance of deamidation at each of the three Asn sites increased over time at different rates following the following first-order kinetic equation: 脱酰胺化 (t) = 1-(1-P0)·e -k脱酰胺化t Using nonlinear regression, the in vivo deamidation rate constants at Asn sites 1, 2, and 3 were determined to be 0.003523% / day, 0.5394% / day, and 0.1546% / day, respectively. Figure 3B ) In the multiple dose study, the relative abundance of deamidation at each of the three Asn sites increased over time following a first-order kinetic equation during each dosing interval, but decreased sharply after each subsequent dose due to dilution of the newly administered unmodified MAB1, showing an upward-trending sawtooth pattern. Each dosing time is indicated by an arrow "↑".
[0033] Figure 4A and 4B It is shown that from a single-dose PK study ( Figure 4A ) and multiple dose PK studies ( Figure 4B ) is a graph showing the relative abundance of oxidation at each of the three Met sites in the MAB1 Fc region. ( Figure 4A ) In the single-dose study, the relative abundance of Met oxidation fluctuated slightly but remained nearly constant at all three sites. ( Figure 4B ) In the multiple dose study, the relative abundance of oxidation at Met site 1 decreased slightly during each dosing interval and increased slightly after each dose. The relative abundance of oxidation at Met sites 2 and 3 remained stable. Each dosing time is indicated by an arrow "↑".
[0034] Figure 5A and Figure 5B It is shown that from a single-dose PK study ( Figure 5A) and multiple dose PK studies ( Figure 5B ) of the relative abundance of N-terminal pyroglutamate. Figure 5A ) In the single-dose study, the relative abundance of N-terminal pyroglutamate increased over time. ( Figure 5B ) In the multiple dose study, the relative abundance of N-terminal pyroglutamate increased during each dosing interval, but decreased sharply after each subsequent dose of MAB1 due to dilution of the newly administered unmodified MAB1, showing an upward trending sawtooth pattern. Each dosing time is indicated by an arrow "↑".
[0035] Fig. 6A and Figure 6B It is shown that from a single-dose PK study ( Fig. 6A ) and multiple dose PK studies ( Figure 6B ) with a heavy chain C-terminal lysine. In both the single-dose and multiple-dose studies, the C-terminal lysine was rapidly removed within one day after each dose. Each dosing time is indicated by an arrow "↑".
[0036] Fig. 7A and Figure 7B It is shown that from a single-dose PK study ( Fig. 7A ) and multiple dose PK studies ( Figure 7B ) is a graph showing the relative abundance of mannose 5 glycoforms. Fig. 7A ) In a single-dose study, the relative abundance of mannose 5 decreased from 0.5% to undetectable within 6 weeks. ( Figure 7B ) In the multiple dose study, the relative abundance of mannose 5 decreased during each dosing interval, but increased dramatically at each subsequent new dose due to the newly administered MAB1 with higher levels of mannose 5, showing a downward trending sawtooth pattern. Each dosing time is indicated by an arrow "↑".
[0037] Fig. 8A is a graph showing the exposure of subjects to total MAB1 from a single-dose PK study. The two-compartment pharmacokinetic model equation is described as C(t) = Ae -αt +Be -βt The MAB1 serum concentration-time curve (black line) was fitted to the MAB1 serum concentration measured by ELISA (black dots). The subject's exposure to total MAB1 over the course of 56 days, represented by the AUC of C(t), was determined to be 4302.0 μg / mL·day.
[0038] Figure 8B is a graph showing exposure of subjects from a single dose study to MAB1 with PQA. The serum concentration-time profile of MAB1 with deamidation is depicted as C PQA (t) = C(t)·P PQA (t), CPQA (t) The AUC of the curve corresponds to the subject's exposure to MAB1 with PQA. The subject's exposure to MAB1 with deamidation at Asn site 2 was determined to be 426.7 μg / mL·day over 56 days (solid line). The subject's exposure to MAB1 with deamidation at Asn site 3 was determined to be 152.5 μg / mL·day over 56 days (dashed line). The subject's exposure to MAB1 with N-terminal pyroglutamate was determined to be 191.5 μg / mL·day over 56 days (dotted line).
[0039] Fig.9A is a graph showing model predictions and experimental measurements of serum concentrations of total MAB1 and MAB1 with deamidation at Asn Site 2 or Asn Site 3 from a multiple dose PK study. The predicted pre-dose and post-dose serum concentrations of total MAB1 and MAB1 with deamidation at Asn Site 2 or Asn Site 3 were in good agreement with the experimental measurements. The pre-dose and post-dose concentrations of total MAB1 and MAB1 with deamidation at Asn Site 2 or 3 approached steady-state levels after an extended dosing period.
[0040] Fig. 9B It is a figure showing the model prediction and experimental measurement results of the relative abundance of deamidation at Asn site 2 or Asn site 3. The predicted level is fully consistent with the experimental value. The relative abundance before and after the dose of deamidation at Asn site 2 or Asn site 3 is close to the steady-state level after the extended administration period. Each administration time is indicated by an arrow "↑".
[0041] Fig. 10A is a graph showing model predictions and experimental measurements of serum concentrations of MAB1 with N-terminal pyroglutamate from a multiple dose PK study. The predicted pre- and post-dose serum concentrations of MAB1 with N-terminal pyroglutamate were in good agreement with the experimental measurements. The pre- and post-dose concentrations of MAB1 with N-terminal pyroglutamate approached steady-state levels after an extended dosing period.
[0042] Fig. 10B : is a graph showing the model prediction and experimental measurement results for the relative abundance of N-terminal pyroglutamate. The predicted levels are fully consistent with the experimental values. The relative abundance of N-terminal pyroglutamate before and after the dose approaches the steady-state level after the extended administration period. Each administration time is indicated by an arrow "↑".
[0043] Fig.11A and Fig. 11B is to show that single-dose studies ( Fig.11A ) and multiple dose studies ( Fig. 11B) to a hypothetical CDR deamidation with an in vivo deamidation rate of 2.5% / day and initial deamidation levels of 0%, 10%, and 20%. Fig.11A ) Subjects in the single-dose study were exposed to 1385 μg / mL·day, 1653 μg / mL·day, and 1921 μg / mL·day over 56 days to hypothetical CDR deamidated variants with 0%, 10%, and 20% initial deamidation, respectively, which constituted 32.5%, 38.8%, and 45.1% of the subject's exposure to the total mAb, respectively. ( Fig. 11B ) The exposures of subjects in the multiple-dose study over 5 doses (56 days) to the hypothetical CDR deamidated variants with 0%, 10%, and 20% initial deamidation were 1086 μg / mL·day, 1478 μg / mL·day, and 1871 μg / mL·day, respectively, which constituted 21.7%, 29.5%, and 37.3% of the subjects' exposure to the total mAb, respectively, with each dosing time indicated by an arrow "↑".
[0044] Fig.12 is a block diagram of a computer system that can be used to implement aspects of the present disclosure.
[0045] Fig.13 is an illustration of a distributed computing environment in which aspects of the present disclosure may be implemented. Specific implementation plan
[0046] Before describing the inventive method, it should be understood that the present invention is not limited to the specific methods and experimental conditions described, because such methods and conditions may vary. It should also be understood that the terminology used herein is only for the purpose of describing specific embodiments, and is not intended to be limiting, because the scope of the present invention will be limited only by the appended claims.
[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the art to which the invention belongs. Although any methods and materials similar or equivalent to those described herein can be used in practicing or testing the present invention, preferred methods and materials are now described. All patents, applications, and non-patent publications mentioned in this specification are incorporated herein by reference in their entirety.
[0048] Abbreviations used in this article
[0049] Asn: Asparagine
[0050] AUC: Area Under the Curve
[0051] CDR: complementarity determining region
[0052] CV: Coefficient of variation
[0053] EIC: Extracted ion chromatography
[0054] ELISA: Enzyme-linked immunosorbent assay
[0055] FDA: U.S. Food and Drug Administration
[0056] HC: Heavy chain
[0057] IgG: Immunoglobulin G
[0058] LC: light chain
[0059] mAb: monoclonal antibody
[0060] MAB1: Therapeutic human monoclonal antibody used in Example 1
[0061] Met: Methionine
[0062] PK: Pharmacokinetics
[0063] PQA: Product Quality Attributes
[0064] PTM: post-translational modification
[0065] RP-LC-MS / MS: Reversed-phase liquid chromatography tandem mass spectrometry
[0066] definition
[0067] As used herein, the term "antibody" is intended to refer to an immunoglobulin molecule composed of four polypeptide chains, two heavy (H) chains and two light (L) chains, interconnected by disulfide bonds (i.e., a "complete antibody molecule"), as well as multimers thereof (e.g., IgM) or antigen-binding fragments thereof. Each heavy chain consists of a heavy chain variable region ("HCVR" or "V H ”) and the heavy chain constant region (composed of domain C H 1. C H 2 and C H 3). In various embodiments, the heavy chain can be of IgG isotype. In some cases, the heavy chain is selected from IgG1, IgG2, IgG3, or IgG4. In some embodiments, the heavy chain has an isotype of IgG1 or IgG4, optionally comprising a chimeric hinge region of isotype IgG1 / IgG2 or IgG4 / IgG2. Each light chain consists of a light chain variable region ("LCVR" or "V L ”) and the light chain constant region (C L ) H and V L The V region can be further subdivided into regions of hypervariability, termed complementarity determining regions (CDRs), interspersed with regions that are more conserved, termed framework regions (FRs). H and V LIt is composed of three CDRs and four FRs arranged in the following order from amino terminus to carboxyl terminus: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. The term "antibody" includes reference to both glycosylated and unglycosylated immunoglobulins of any isotype or subclass. The term "antibody" includes antibody molecules prepared, expressed, created or isolated by recombinant means, such as antibodies isolated from host cells transfected to express the antibody. For a review of antibody structure, see IMGT unique numbering for immunoglobulin and T cell receptor variable domains and Ig superfamily V-like domains, 27 (1) Dev. Comp. Immunol. 55-77 (2003); and M. Potter, Structural correlates of immunoglobulin diversity, 2 (1) Surv. Immunol. Res. 27-42 (1983).
[0068] The term antibody also encompasses "bispecific antibodies," which include heterotetrameric immunoglobulins that can bind to more than one different epitope. One half of a bispecific antibody comprising a single heavy chain and a single light chain and six CDRs binds to one antigen or epitope, and the other half of the antibody binds to a different antigen or epitope. In some cases, bispecific antibodies can bind to the same antigen, but bind to different epitopes or non-overlapping epitopes. In some cases, both halves of a bispecific antibody have the same light chain while retaining dual specificity. Bispecific antibodies are generally described in U.S. Patent Application Publication No. 2010 / 0331527 (December 30, 2010).
[0069] As used herein, the term "antigen-binding portion" (or "antibody fragment") of an antibody refers to one or more fragments of an antibody that retain the ability to specifically bind to an antigen. Examples of binding fragments encompassed within the term "antigen-binding portion" of an antibody include (i) a Fab fragment, i.e., a monovalent fragment consisting of VL, VH, CL, and CH1 domains; (ii) a F(ab')2 fragment, i.e., a bivalent fragment comprising two Fab fragments linked by a disulfide bond at the hinge region; (iii) a Fd fragment consisting of the VH and CH1 domains; (iv) a Fv fragment consisting of the VL and VH domains of a single arm of an antibody; (v) a dAb fragment consisting of the VH domain (Ward et al. (1989) Nature 241:544-546); (vi) isolated CDRs; and (vii) a scFv consisting of the two domains of the Fv fragment, VL, and VH, joined by a synthetic linker to form a single protein chain in which the VL and VH regions are paired to form a monovalent molecule. Other forms of single chain antibodies such as diabodies are also encompassed under the term "antibody" (see, e.g., Holliger et al. (1993) 90 PNAS USA 6444-6448; and Poljak et al. (1994) 2 Structure 1121-1123).
[0070] Furthermore, antibodies and antigen-binding fragments thereof can be obtained using standard recombinant DNA techniques generally known in the art (see Sambrook et al., 1989).
[0071] "Fc fusion proteins" comprise part or all of two or more proteins that do not occur together in nature, one of which is the Fc portion of an immunoglobulin molecule. The preparation of fusion proteins comprising certain heterologous polypeptides fused to various portions of an antibody-derived polypeptide (comprising an Fc domain) has been described, for example, by Ashkenazi et al., (1991) 88 Proc. Natl. Acad. Sci. USA 10535; Byrn et al., (1990) 344 Nature 677; and Hollenbaugh et al., (1992) "Construction of Immunoglobulin Fusion Proteins", in Current Protocols in Immunology, Supplement 4, pp. 10.19.1-10.19.11. "Receptor Fc fusion proteins" comprise one or more extracellular domains of a receptor coupled to an Fc portion, the one or more extracellular domains comprising, in some embodiments, a hinge region, followed by the CH2 and CH3 domains of an immunoglobulin. In some embodiments, the Fc fusion protein contains two or more different receptor chains that bind to one or more ligands. For example, the Fc fusion protein is a trap, such as, for example, an IL-1 trap (e.g., rilonacept, which contains the IL-1RAcP ligand binding region fused to the extracellular region of IL-1R1, which is fused to the Fc of hIgG1; see U.S. Pat. No. 6,927,004), or a VEGF trap (e.g., aflibercept, which contains the Ig domain 2 of the VEGF receptor Flt1 fused to the Ig domain 3 of the VEGF receptor Flk1, which is fused to the Fc of hIgG1; see U.S. Pat. No. 7,087,411 (issued Aug. 8, 2006) and U.S. Pat. No. 7,279,159 (issued Oct. 9, 2007)).
[0072] The term "human antibody" is intended to include antibodies with variable regions and constant regions derived from human germline immunoglobulin sequences. The human mAb of the present invention may include amino acid residues not encoded by human germline immunoglobulin sequences (e.g., mutations introduced by random or site-specific mutagenesis in vitro or by somatic mutations in vivo), such as in CDR and particularly in CDR3. However, as used herein, the term "human antibody" is not intended to include mAbs in which CDR sequences derived from the germline of another mammalian species (e.g., mice) have been transplanted to human FR sequences. The term includes antibodies produced recombinantly in non-human mammals or in cells of non-human mammals. The term is not intended to include antibodies separated from or produced in human subjects.
[0073] As used herein, the term "subject" refers to an animal, preferably a mammal, and more preferably a human, for example, in need of improvement, prevention and / or treatment of a disease or disorder.
[0074] As used herein, the term "treat" or "treating" refers to the reduction or improvement of the severity of at least one symptom or indication caused by administering a therapeutic agent such as an antibody of the present invention to a subject in need thereof. The term includes inhibition of disease progression. The therapeutic agent may be administered to a subject at a therapeutic dose.
[0075] "Post-translational modification" (PTM) refers to the covalent modification of a protein after protein biosynthesis. Post-translational modifications may occur on amino acid side chains or at the C-terminus or N-terminus of a protein. Exemplary post-translational modifications of antibodies include deamidation, oxidation, glycation, disulfide formation, N-terminal pyroglutamate formation, C-terminal lysine removal, and glycosylation.
[0076] General Description
[0077] Therapeutic monoclonal antibodies (mAbs) produced in mammalian cells are heterogeneous due to post-translational modifications (PTMs). PTMs of therapeutic monoclonal antibodies (mAbs) are important product quality attributes (PQAs) that can potentially affect drug stability, safety, and efficacy (Xu et al., Journal of Applied Bioanalysis 2017; 3: 21-5). PTMs can occur during mAb production, purification, storage, and administration (Kozlowski et al., Advanced drug delivery reviews 2006; 58: 707-22; Liu et al., Journal of pharmaceutical sciences 2008; 97: 2426-47; Goetze et al., mAbs 2010; 2: 500-7; Wang et al., Journal of pharmaceutical sciences 2007; 96: 1-26). Modifications that occur during drug production and storage can be reliably monitored and controlled. However, since mAbs encounter different environments in vivo and in vitro, additional modifications may occur after drug administration. Compared to in vitro conditions, the PTMs of mAbs change significantly after administration into the bloodstream. Therefore, monitoring the changes in PTMs of mAbs in vivo helps to evaluate the PQA criticality during product risk assessment. In addition, quantitative assessment of the exposure of subjects to PTM variants helps to assess the safety impact of PTMs and the efficacy of therapeutic mAbs. The evaluation and prediction of PTMs that may occur in vivo will not only help to understand the criticality of quality attributes for product risk assessment, but will also help product development teams design candidate mAb drugs with enhanced in vivo stability.
[0078] As disclosed herein, the inventors have conducted detailed analysis and determination of product quality attributes (PQAs) present in therapeutic monoclonal antibodies. Using the therapeutic monoclonal antibody MAB1 as a model, the inventors have evaluated multiple post-translational modifications, including, for example, deamidation, oxidation, N-terminal pyroglutamate formation, and C-terminal lysine removal. As disclosed herein, the inventors determined the PTMs (e.g., deamidation, oxidation, N-terminal pyroglutamate formation, C-terminal lysine removal, and high mannose glycosylation) of therapeutic mAb (MAB1) in single-dose and multiple-dose monkey pharmacokinetic (PK) studies. Using the resulting data, the inventors used three common PTMs as examples to construct modeling equations to calculate the in vivo serum concentrations of PQAs in single-dose and multiple-dose regimens, the exposure of subjects to PQAs, and the relative abundance of PQAs. The model predictions show consistency with experimental results. Exemplary applications of the model are illustrated using hypothetical PTMs. The model provides a simplified modeling tool for bioanalytical chemists to quantitatively evaluate the criticality of PQAs when advanced PK models are not available. As detailed in the Examples below, monkey serum samples from single-dose and multiple-dose PK studies were analyzed using affinity purification followed by LC-MS / MS to determine multiple post-translational attributes of MAB1. Using the data obtained from this analysis, the inventors were able to construct mathematical models and / or simulations of the in vivo progression of antibody post-translational modifications to predict subject exposure to product quality attributes of therapeutic antibodies. The simulations provide a quantitative method to assess the criticality of PQAs.
[0079] Based on the single-dose PK study of PQA of MAB1, the inventors were able to model the PK concentrations and ratios of PTM variants, and were able to quantitatively assess and predict the exposure of subjects to representative PTM attributes (e.g., asparagine deamidation) in both single-dose and multiple-dose regimens. The model also predicts the highest and lowest PTM levels observed in multiple-dose regimens. This modeling approach provides a new tool for bioanalytical chemists to quantitatively assess PQA criticality in therapeutic mAbs. The model can be used to evaluate the impact of PQA caused by process changes or variability of different batches by adjusting the initial PQA level (P0). If the modeling shows that PQA progresses or the subject's exposure to PQA is insensitive to the initial PQA level, the acceptable range of PQA can be broadened to adopt process changes.
[0080] The model can also be used to simulate the exposure of the subject by extending the dosing period (t) or changing the dosing interval (τ), thereby providing a quantitative assessment of the exposure of the subject to PQA in preclinical and clinical studies. The quantitative assessment of the exposure of the subject to PQA can help evaluate the risks associated with PQA. PQAs with known adverse immune responses that have been determined to have high exposure levels during PK studies will have a high potential risk. In such cases, certain strategies (such as protein redesign to enhance in vivo stability) can be implemented to eliminate or reduce the level of PQA.
[0081] Methods for predicting exposure to antibody product quality attributes
[0082] Disclosed herein is a method for predicting the in vivo serum concentration of an antibody (such as a therapeutic monoclonal antibody) having a target post-translational modification after administration to a subject. The disclosed method includes using an in vivo rate constant determined for the target post-translational modification to predict the percentage of the antibody having the target post-translational modification. The percentage of the antibody having the post-translational modification can be multiplied by the total in vivo concentration of the antibody at any time point to determine the concentration of the antibody having the target post-translational modification at that time point. For example, using Equation C PTM (t) = C(t)·P(t), where C(t) is the serum concentration of the antibody; and P(t) is the percentage of the post-translational modification of interest. The serum concentration of the antibody with the post-translational modification of interest (C PTM (t)). In other words, the serum concentration of the antibody with the post-translational modification of interest is a percentage of the total antibody with that post-translational modification. As disclosed herein, the serum concentration of the antibody can be described by the two-compartment pharmacokinetic model equation C(t) = Ae -αt +Be -βt , where α and β are first-order confounding constants, and A and B are confounding coefficients, respectively.
[0083] In embodiments, the in vivo rate constant of a target post-translational modification is determined by quantifying the percentage of the target post-translational modification over time after a single dose and fitting the quantified percentage of the target post-translational modification to the equation Where P PTM (t) is the ratio of the target post-translational modification over time; P0 is the initial target post-translational modification percentage; and k PTM is the PTM rate constant for the PTM of interest. By fitting a curve to the measured percentage of the PTM of interest, the rate constant k can be determined. PTM , for example using nonlinear regression or other mathematical fitting algorithms.
[0084] In an embodiment, the method may further comprise determining the subject's exposure to the amount of an antibody having a post-translational modification of interest. For example, using an area under the curve (AUC) analysis of the concentration of an antibody having a post-translational modification of interest over a period of time (e.g., days, weeks, or even months) PTM (t)) generates a total exposure value for the subject to the amount of the antibody having the post-translational modification of interest.
[0085] Simultaneous observation at multiple sites is also possible. For example, the proportion of antibodies with post-translational modifications at a particular site can be described by first-order kinetics as Where P PTM_位点_i (t) is the ratio of the post-translational modification at site i over time; P0 is the initial post-translational modification level at site i; k PTM_位点_i V is the post-translational modification rate constant at site i. Therefore, the PK concentration of an antibody at a site i within a post-translationally modified antibody can be described as V PTM_位点_i (t) = C(t)·P PTM_位点_i (t) C PTM_位点_i (t) The area under the curve (AUC) of the graph presents the subject's exposure to the quality attribute. The approximate concentration of each class of antibody variants (e.g., with 1, 2, 3 or even more different post-translational modifications) can be further calculated.
[0086] As discussed below in the example of deamidation, the methods disclosed herein can be used to predict the in vivo serum concentration of an antibody having a post-translational modification of interest after multiple administrations by extending the model to multiple administrations. For example, predicting the in vivo serum concentration of an antibody having a post-translational modification of interest after multiple administrations may include superimposing multiple single-administration models, wherein the serum concentration of the antibody at each administration is individually described by a two-compartment pharmacokinetic model equation of C(t)=Ae -αt +Be -βt , where α and β are first-order confounding constants, and A and B are confounding coefficients, respectively.
[0087] The serum concentration of the antibody with the post-translational modification of interest immediately before the mth dose can be described as follows:
[0088] C PTM,m,剂量前 (m·τ)=P PTM [(m-1)·τ]C[(m-1)·τ]+…+P PTM (1·τ)C(1·τ)+P PTM (0)C(0)
[0089] The serum concentration of the antibody with the post-translational modification of interest immediately following the mth dose can be described as follows:
[0090] C PTM,m,剂量后 (m·τ)=P PTM [(m-1)·τ]C[(m-1)·τ]+…+P PTM (1·τ)C(1·τ)+P PTM (0)C(0)
[0091] Where P PTM (t) is the first order reaction equation as described for a single administration.
[0092] Therefore, the proportion of antibodies with a post-translational modification of interest immediately before the mth dose can be described as follows:
[0093]
[0094] The proportion of antibodies with the post-translational modification of interest immediately following the mth dose can be described as follows:
[0095]
[0096] When m approaches infinity, P 脱酰胺化,m,剂量前 and P PTM,m,剂量后 Arriving at the platform, it can be described as follows:
[0097]
[0098] therefore:
[0099]
[0100]
[0101] The equation constants A, B, α, β, and k can be solved by fitting the equation to the experimental results obtained in the first dosing interval.
[0102] In various embodiments, post-translational modifications include product quality attributes that can potentially affect drug stability, safety and / or efficacy. In certain embodiments, the method is used to monitor the in vivo post-translational modification profile of a monoclonal antibody (such as a recombinant therapeutic monoclonal antibody). In some cases, the method also includes creating a predictive model of the in vivo progress of post-translational modifications and the exposure of subjects to product quality attributes. In some embodiments, post-translational modifications include one or more of deamidation, oxidation, saccharification, disulfide formation, N-terminal pyroglutamate formation, C-terminal lysine removal and / or high mannose glycosylation. In some embodiments, the method also includes modeling the correlation between the PK concentration and ratio of post-translational modified variants. In some embodiments, the method also includes quantitatively assessing and predicting the exposure of subjects to representative post-translational modifications in both single-dose regimens and multiple-dose regimens. In some embodiments, the method also includes predicting the highest and lowest post-translational modification levels observed in multiple-dose regimens.
[0103] Exemplary Computer System
[0104] refer to Fig.12 An exemplary computer system for implementing the disclosed method includes a computer 120 (such as a personal computer, a laptop computer, a palmtop computer, a machine-top computer, a server, a mainframe computer, a handheld device, and a variety of other computers), the computer 120 includes a processing unit 121, a system memory 122, and a system bus 123 that couples various system components including the system memory to the processing unit 121. The processing unit can be any of a variety of commercially available processors, including processors from etc. (including Cyrix, AMD and Nexgen) x86, and compatible microprocessors; Alpha from Digital; Intel processors from MIPS Technology, NEC, MIPS from Siemens et al.; and Dual microprocessor and other multi-processor architectures may also be used as processing unit 121.
[0105] The system bus may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of conventional bus architectures such as PCI, VESA, AGP, MicroChannel, ISA, and EISA, to name a few. A basic input / output system (BIOS) containing basic routines for transferring information between elements within the computer 120, such as during startup, is stored in ROM 124. The system memory includes read-only memory (ROM) 124 and random access memory (RAM) 125. The computer 120 may also include a hard disk drive 127, a magnetic disk drive 128 (e.g., to read from or write to a removable magnetic disk 129), and an optical disk drive 130 (e.g., to read from or write to a CD-ROM disk 131 or to read from or write to other optical media). The hard disk drive 127, magnetic disk drive 128, and optical disk drive 130 are connected to the system bus 123 by a hard disk drive interface 132, a magnetic disk drive interface 133, and an optical disk drive interface 134, respectively. The drives and their associated computer-readable media provide non-volatile storage of data, data structures (databases), computer-executable instructions, etc. for the computer 120. Although the above description of computer-readable media refers to hard disks, removable disks, and CDs, those skilled in the art will appreciate that other types of media that can be read by the computer, such as magnetic tape cassettes, flash memory cards, digital video disks, Bernoulli cassettes, etc., may also be used in the exemplary operating environment.
[0106] Data may be stored in the drives and RAM 125, which include an operating system 135, one or more application programs 136, other program modules 137, and program data 138. A user may enter commands and information into the computer 120 using various input devices, such as a keyboard 140 and a pointing device, such as a mouse 142. Other input devices (not shown) may include a microphone, a circular satellite dish, a scanner, and the like. These and other input devices are typically connected to the processing unit 121 via a serial port interface 146 coupled to the system bus, but may be connected via other interfaces, such as a parallel port, a game port, or a universal serial bus (USB). A monitor 147 or other type of display device is also connected to the system bus 123 via an interface, such as a video adapter 148. In addition to the monitor, the computer typically includes other peripheral output devices (not shown), such as a printer. Alternatively, the display medium is a printout or other tangible medium. The output information may also be retained in a computer-readable medium for storage and / or subsequent use or display.
[0107] The computer 120 may operate in a networked environment using logical connections to one or more other computer systems, such as the computer 102. The other computer systems may be servers, routers, peer devices, or other common network nodes, and typically include most or all of the elements described with respect to the computer 120, except that only the memory storage device 149 is present. Fig.12 Shown in. Fig.12 The logical connections depicted in 1 include a local area network (LAN) 151 and a wide area network (WAN) 152. Such networking environments are commonplace in offices, enterprise-wide computer networks, intranets and the Internet.
[0108] When used in a LAN networking environment, the computer 120 is connected to the local area network 151 via a network interface or adapter 153. When used in a WAN networking environment, the computer 120 typically includes a modem 154 or other means for establishing communications within a wide area network 152 (such as the Internet) (e.g., via the LAN 151 and a gateway or proxy server 155). The modem 154 can be internal or external and is connected to the system bus 123 via the serial port interface 146. In a networking environment, program modules or portions thereof depicted with respect to the computer 120 may be stored in a remote memory storage device. It should be appreciated that the network connections shown are exemplary and other means for establishing a communications link between computer systems (including Ethernet cards, ISDN terminal adapters, ADSL modems, 10BaseT adapters, 100BaseT adapters, ATM adapters, etc.) may be used.
[0109] The methods described above, including the acts and operations included therein, may be performed by a computer 120 or by an instrument or other device that is specially programmed or dedicated to perform the disclosed methods. Thus, the methods may be performed on a specific machine, such as a device other than a general purpose computer. Such acts and operations are sometimes referred to as computer-implemented. It should be understood that the acts and symbolically represented operations include manipulation of the electrical signal processing unit 121 representing data bits, the manipulation causing the conversion or reduction of the electrical signal representation, and the maintenance of data bits at memory locations in the memory system (including system memory 122, hard disk drive 127, floppy disk 129, and CD-ROM 131) to thereby reconfigure or otherwise change the operation of the computer system, as well as other processing of signals. The memory location where the data bits are maintained is a physical location having specific electrical, magnetic, or optical properties corresponding to the data bits.
[0110] Exemplary Distributed Computing Environment
[0111] Fig.13A distributed computing environment in which software and / or database elements for implementing the method of the present disclosure can reside is shown. Although the distributed computing environment 200 includes two computer systems 202, 204 connected by a connection medium 206, the disclosed method is equally applicable to any large number of computer systems connected by the connection medium 206. Computer systems 202, 204 can be any of several types of computer system configurations, including personal computers, multiprocessor systems, handheld devices, etc. In terms of logical relationships with other computer systems, the computer system can be a client, server, router, peer device or other common network node. Additional computer systems 202 or 204 can be connected by any number of connection media 206. Connection media 206 can include any local area network (LAN), wide area network (WAN) or other computer networks, including but not limited to Ethernet, enterprise-wide computer networks, enterprise intranets and the Internet.
[0112] Portions of the software and databases for storing data may be implemented in a single computer system 202 or 204, where the application is later distributed to other computer systems 202, 204 in the distributed computing environment 200. Portions of the software may also be practiced in a distributed computing environment 200 where tasks are performed by a single computer system 202 or 204 acting as a remote processing device accessed via a communications network, where the distributed application is later distributed to other computer systems in the distributed computing environment 200. In a networked environment, program modules and databases for storing data may be located on more than one computer system 202 or 204. Communications between computer systems in a distributed computing network may advantageously include encryption of the communicated data.
[0113] Example
[0114] The following examples are presented in order to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the methods and compositions of the present invention, and are not intended to limit the scope of what the inventors regard as their invention. Every effort has been made to ensure accuracy with respect to the numbers used (e.g., amounts, temperatures, etc.), but some experimental errors and deviations should be taken into account. Unless otherwise indicated, parts are parts by weight, molecular weight is average molecular weight, temperature is in degrees Celsius, room temperature is about 25° C., and pressure is at or near atmospheric pressure.
[0115] Example 1
[0116] Quantification and Modeling of Product Quality Attributes of a Therapeutic Monoclonal Antibody in Single- and Multiple-Dose Monkey Pharmacokinetic Studies
[0117] This example describes a study of the number of PQAs for a therapeutic monoclonal antibody (MAB1) using monkey serum samples from single-dose and multiple-dose PK studies. Several PTM attributes of MAB1 were evaluated using affinity purification followed by LC-MS / MS, including deamidation, oxidation, N-terminal pyroglutamate, C-terminal lysine removal, and high mannose glycosylation. In addition, a model was constructed to simulate the in vivo progression of PTMs and subject exposure to PQAs, thereby providing a quantitative method to assess the criticality of PQAs. Asparagine deamidation was used as a representative PTM attribute to evaluate the model. Subject exposure to deamidated MAB1 simulated by the disclosed model was well consistent with the experimental results of both single-dose and multiple-dose studies.
[0118] Materials and methods
[0119] Material
[0120] The human IgG4 monoclonal antibody (MAB1) and anti-human antibodies used in this study were produced at Regeneron Pharmaceuticals, Inc. (Tarrytown, NY). The average half-life of MAB is approximately 11.5 days. Unless otherwise specified, all reagents were obtained from Sigma-Aldrich (St. Louis, MO) or Thermo Fisher Scientific (Waltham, MA).
[0121] Preclinical sample information
[0122] Preclinical serum samples were obtained from either single-dose or multiple-dose cynomolgus monkey PK studies. MAB1 was administered intravenously (IV) to subjects. For single-dose studies, animals were dosed at 10 mg / kg and serum samples were collected at designated time points (pre-dose, 5 minutes, 4 hours, 12 hours, 1 day, 3 days, 7 days, 14 days, 18 days, 30 days, 42 days, and 56 days) over the course of 8 weeks (56 days). For multiple-dose studies, animals were dosed at 3 mg / kg every 2 weeks (14 days) and serum samples were collected at designated time points (pre-dose, 1 hour, 4 hours, 1 day, 3 days, 7 days) within the first dosing interval over the course of 8 weeks (56 days) and at designated time points before and after each new dose (14 days, 28 days, 42 days, 56 days). Serum samples were stored at -80°C until analysis. Enzyme-linked immunosorbent assay (ELISA) was used to measure the MAB1 serum concentration at each collection time point. Briefly, MAB1 was captured on a microtiter plate coated with the drug target. MAB1 captured on the plate was detected using a biotinylated mouse anti-human IgG4 monoclonal antibody followed by neutraavidin conjugated to horseradish peroxidase (NeutraAvidin-HRP). A luminol substrate specific for peroxidase was then added to achieve a signal intensity proportional to the concentration of MAB1.
[0123] Affinity purification of MAB1 from serum samples
[0124] MAB1 was purified from the collected monkey serum samples by affinity purification. In brief, biotin-labeled anti-human antibodies were conjugated to Dynabeads MyOne Streptavidin T1 magnetic beads (Invitrogen, Carlsbad, CA) at room temperature for 10 minutes. The conjugated beads were incubated with serum samples at room temperature for 30 minutes. The beads were washed with HBS-EP buffer (GE Healthcare, Pittsburgh, PA) and then eluted with 0.1% formic acid and 50% acetonitrile. A bioanalyzer (Agilent Technologies, Santa Clara, CA) was used to assess the recovery of MAB1 during affinity purification. Each sample was performed with a known amount of 230 kDa high molecular weight standard (shown as "upper marker") from Agilent as an internal normalization control. The ratio of the MAB1 peak area in each sample to the upper marker peak area was calculated to correct for round variability. The recovery of MAB1 during affinity purification was calculated using the following equation:
[0125] Equation 1:
[0126] Trypsin digestion
[0127] Purified MAB1 samples were dried using a vacuum concentrator (LABCONCO, Kansas City, MO). The dried samples were resuspended in 8M urea and 10mM TCEP and incubated at 37°C for 30 minutes. The reduced cysteine residues were alkylated with 10mM iodoacetamide at room temperature in the dark for 30 minutes. After alkylation, the urea concentration was diluted to 1.25M before digestion. Trypsin (Promega, Sunnyvale, CA) was added to the sample at an enzyme: substrate ratio of 1:10 and incubated at 37°C for 4 hours. Digestion was terminated by adding 20% formic acid (FA; Thermo Scientific, San Jose, CA). The digested samples were stored at -80°C until analysis.
[0128] LC-MS / MS and data analysis
[0129] Peptides generated by trypsin digestion were separated using an Acquity UPLC CSH C18 1.7 μm, 2.1 mm x 150 mm column (Waters, Milford, MA) coupled to a Q Exactive plus mass spectrometer (Thermo Fisher Scientific, San Jose, CA). Mobile phase A was 0.1% FA in water and mobile phase B was 0.1% FA in acetonitrile. A gradient from 2% mobile phase B to 30% mobile phase B for 56 minutes at a flow rate of 0.25 mL / min was used for peptide separation. MS acquisition consisted of a full mass scan followed by a tandem mass (MS / MS) scan of the top 5 highest intensity ions of each full scan. Peptide and PTM identification was determined by Byonic (version 2.16.11, ProteinMetrics Inc., San Carlos, CA) and manual verification was performed. In order to quantify the relative abundance of PTM, the extracted ion chromatograms based on the m / z of the first isotope peak of both modified peptides and native peptides were generated, and the extracted peak areas were integrated using Skyline-daily (version 4.1.1.18151) using a mass window of 5 ppm. The percentage of each PTM variant was calculated using the extracted ion chromatogram (EIC) peak area of the modified peptide relative to the sum of the peak areas of the modified peptide and native peptide.
[0130] Calculation of Subject Exposure to Quality Attributes in Single-Dose PK Studies
[0131] Quantitative assessment of subject exposure to quality attributes in a single-dose PK study was previously described by Flynn et al. (Goetze et al., mAbs 2010; 2: 500-7). The MAB1 serum concentration-time equation can be described by a two-compartment model (Rowland and Tozer Wolters Kluwer Health / Lippincott William & Wilkins, 2011):
[0132] Equation 2:
[0133] , where C(t) is the serum concentration of MAB1; C1, C2, λ1 and λ2 are confounding constants.
[0134] To simplify readability, the following format is used:
[0135] Equation 3: C(t) = Ae -αt +Be -βt
[0136] , where A and B are zero time intercepts (confounding coefficients); α and β are confounding first order constants, respectively. The equation was fitted to the MAB1 serum concentrations measured by ELISA at the collection time points to find the best fit parameters. The fitting constraints were: C(0) must be greater than the MAB1 serum concentration measured by ELISA after the first dose; the C(infinity) constant equals 0; A and B must be greater than 0. R squared was used to quantify the goodness of fit and an R square greater than 0.95 was considered a good fit.
[0137] The relative abundance of MAB1 with site-specific deamidation at Asn can be described by the following first-order kinetic equation:
[0138] Equation 4:
[0139] , where P 脱酰胺化 (t) is the percentage of deamidation; P0 is the initial deamidation level at day 0; k 脱酰胺化 is the deamidation rate constant.
[0140] The serum concentration of MAB1 with deamidation at a particular Asn site can be described as:
[0141] Equation 5: C 脱酰胺化 (t) = C(t)·P 脱酰胺化 (t)
[0142] , where C 脱酰胺化 (t) is the serum concentration of MAB1 with deamidation at a specific Asn site. 脱酰胺化The area under the curve (AUC) of C(t) represents the subject's exposure to the quality attribute (i.e., MAB1 with deamidation at a specific Asn site). The relationship describing C(t) and P was solved by nonlinear regression using JMP (version 13.2.1, SAS, Cary, NC). 脱酰胺化 C was calculated using Excel (Microsoft, Redmond, WA). 脱酰胺化 (t) and plotted using JMP. Using JMP to calculate the 脱酰胺化 The AUC of (t) represents the exposure of the subject.
[0143] Calculation of Subject Exposure to Quality Attributes in Multiple-Dose PK Studies The modeling of subject exposure to quality attributes in multiple-dose PK studies was mathematically simplified using the superposition principle as a linear accumulation of multiple single doses. The serum concentration of MAB1 after each dose was described by the two-compartment model discussed above in the single-dose PK study. The serum concentration of MAB1 immediately before the mth dose can be described as follows:
[0144] Equation 6: C m,剂量前 (m·τ)=C[(m-1)·τ]+…+C(1·τ)
[0145] , where m is the number of doses and τ is the time interval between doses.
[0146] The serum concentration of MAB1 immediately after the mth dose can be described as follows:
[0147] Equation 7: C m,剂量后 (m·τ)=C[(m-1)·τ]+…+C(1·τ)+C(0)
[0148] When m approaches infinity, C m,剂量前 and C m,剂量后 Approaching a steady-state concentration, described as follows:
[0149] Equation 8:
[0150] Equation 9:
[0151] The serum concentration of MAB1 with deamidation at a particular Asn site immediately before the mth dose can be described as follows:
[0152] Equation 10:
[0153] The serum concentration of MAB1 with deamidation at a particular Asn site following the mth dose determination can be described as follows:
[0154] Equation 11: , where P 脱酰胺化 (t) is the first order reaction equation discussed in the single dose PK study above.
[0155] Therefore, the relative abundance of MAB1 with deamidation at a particular Asn site immediately before the mth dose can be described as follows:
[0156] Equation 12:
[0157] The relative abundance of MAB1 with deamidation at a particular Asn site following the mth dose can be described as follows:
[0158] Equation 13:
[0159] When m approaches infinity, P 脱酰胺化,m,剂量前 and P 脱酰胺化,m,剂量后 Approaching steady-state levels, described as follows:
[0160] Equation 14:
[0161] Equation 15:
[0162] Since the calculation of the summation term converges, the pre-dose level of deamidation at steady state can be simplified as follows:
[0163] Equation 16:
[0164] Therefore, the post-dose level of deamidation at steady state can be simplified as follows:
[0165] Equation 17:
[0166] The equation constants A, B, α, β and k can be solved by fitting the equation to the experimental results within the first dosing interval. Simulated and predicted values are calculated using Excel and then plotted using JMP. Exposure (i.e., AUC) for subjects is calculated using JMP.
[0167] result
[0168] Quantification of post-translational modifications of MAB1 in vivo
[0169] The in vivo kinetics of MAB1 variants with various PTMs were evaluated in both single-dose and multiple-dose monkey PK studies.
[0170] A biotinylated anti-human Fc antibody conjugated to streptavidin magnetic beads was used to extract MAB1 from monkey serum samples. The affinity purification recovery of MAB1 was >99.5% ( Figure 1), indicating that the affinity purification method can completely extract both native MAB1 and its PTM variants, thereby ensuring accurate quantification of the relative PTM abundance of each PTM.
[0171] Figure 2A Exemplary tandem MS spectra showing identification of the wild-type Fc tryptic peptide "DTLMISR" (SEQ ID NO: 1) (lower panel) and the corresponding methionine (Met) oxidized peptide (upper panel). Figure 2B Extracted ion chromatogram peak areas are shown for the wild-type peptide (lower panel) and the corresponding Met-oxidized peptide (upper panel).
[0172] Deamidation of asparagine (Asn) is a common PTM in the highly conserved fragment crystallizable (Fc) region of mAbs (Chelius et al., Analytical Chemistry 2005; 77: 6004-11; Sinha et al., Protein Science: a publication of the Protein Society 2009; 18: 1573-84). Three common deamidation sites were monitored in this study: "NG" (Asn site 1) in the CH2 domain, "NG" (Asn site 2) and "NN" (Asn site 3) in the PENNY peptide of the CH3 domain. In this work, three Asn deamidation sites were monitored: "NG" (Asn site 1) in the CH2 domain, "NG" (Asn site 2) and "NN" (Asn site 3) in the CH3 domain. In a single-dose PK study, the relative abundance of deamidation at Asn site 1 remained almost unchanged at a low level of about 0.2% over 56 days ( Figure 3A The relative abundance of deamidation at Asn-2 and Asn-3 increased from approximately 1% to 25.5% and 9.1%, respectively, over 56 days ( Figure 3A). The deamidation rate depends on the primary amino acid sequence, protein structure, pH, temperature and buffer composition (Chelius et al., Analytical Chemistry 2005; 77: 6004-11; Sinha et al., Protein Science: a publication of the Protein Society 2009; 18: 1573-84; Kossiakoff, Science 1988; 240: 191-4; Athmer et al., The Journal of Biological Chemistry 2002; 277: 30502-7; Xie et al., Journal of Pharmaceutical Sciences 1999; 88: 8-13; Wearne et al., Proteins 1989; 5: 8-12; Capasso et al., The Journal of Peptide Research: official journal of the American Peptide Society 1999; 54: 377-82; Song Y et al., Journal of Pharmaceutical Sciences 2001;90:141-56; Wakankar et al., Journal of Pharmaceutical Sciences 2006;95:2321-36; Zheng et al., International Journal of Pharmaceutics 2006;308:46-51; Pace et al., Journal of Pharmaceutical Sciences 2013;102:1712-23).Asn residues followed by glycine (Gly) or serine (Ser) within the primary amino acid sequence are particularly susceptible to deamidation (Patel et al., Pharmaceutical Research 1990; 7:787-93; Stephenson et al., The Journal of Biological Chemistry 1989; 264:6164-70; et al., Proceedings of the National Academy of Sciences of the United States of America 2001; 98:944-9; Robinson et al., The Journal of Peptide Research: official journal of the American Peptide Society 2004; 63:426-36). The "NG" and "NN" positions are fully exposed to the solvent and are therefore more susceptible to deamidation than the "NG" position in the CH2 domain (DeLano et al., Science 2000; 287: 1279-83).
[0173] The deamidation rate can be described using a first order rate equation as previously established (Geiger et al., The Journal of Biological Chemistry 1987; 262: 785-94). The rate equation can be expressed as: As described in Materials and Methods, P 脱酰胺化 (t) is the ratio of deamidation over time; P0 is the initial deamidation level; k 脱酰胺化 is the deamidation rate constant. The equation was fitted nonlinearly to the deamidation levels measured by LC-MS in the single-dose study to determine the best fit parameters ( Figure 3A and Table 1). The in vivo deamidation rate constants of Asn sites 1, 2, and 3 were fitted in a nonlinear manner to be 0.003278% / day, 0.5384% / day, and 0.1566% / day, respectively, where R 2 They are 0.9946, 0.9866 and 0.8941 respectively (Table 1).
[0174] Table 1: Best fit parameter values in the modified rate equation from a single-dose PK study
[0175]
[0176]
[0177] where P0 is the initial deamidation level at day 0; k 脱酰胺化 is the deamidation rate constant.
[0178] In the multiple-dose study, deamidation of Asn site 1 remained almost constant at a low level of approximately 0.5% after 5 biweekly doses over 56 days ( Figure 3B The relative abundance of deamidation at Asn positions 2 and 3 accumulated during each dosing interval but dropped dramatically after each new dose due to dilution of the newly administered non-deamidated MAB1 ( Figure 3B Due to the accumulation of deamidated MAB1, the pre-dose and post-dose levels of deamidation gradually increased after each dose. The pre-dose level of deamidation at Asn site 2 increased to 14.2% before the 5th dose on day 56, while the post-dose level reached 7.9% ( Figure 3B ). Pre-dose levels of deamidation at Asn site 3 increased to 6.2% before the fifth dose on day 56, while post-dose levels reached 3.6% ( Figure 3B As a result, the relative abundance of deamidation at Asn sites 2 and 3 exhibited a sawtooth pattern with an increasing trend over time ( Figure 3B ). Figure 3B It is also shown that both the pre-dose and post-dose levels of deamidation will approach steady state after a limited number of doses. In order to predict the steady-state level of deamination, a two-compartment model for MAB1 concentration was constructed, which will be discussed in the following modeling section. Met oxidation is another common PTM in mAb (Yin et al., Pharmaceutical research 2013; 30: 167-78; Li et al., mAbs 2016: 0; Li et al., mAbs 2016: 0; Liu et al., The Journal of biological chemistry 2011; 286: 11211-718). Met oxidation in the Fc region has been shown to affect Fc receptor binding (Pan et al., Protein Science: a publication of the Protein Society 2009; 18: 424-33). In this work, three common Met oxidation sites in the Fc region were evaluated: Met site 1 in the CH2 domain and Met sites 2 and 3 in the CH3 domain. In the single-dose study, the relative abundance of oxidation at Met sites 1, 2, and 3 was relatively unchanged over 56 days at levels of approximately 8.0%, approximately 2.0%, and approximately 5.0%, respectively ( Figure 4A). In the multiple-dose study, the level of oxidation at Met site 1 decreased slightly during each dosing interval and increased slightly after each dose, showing a sawtooth pattern that fluctuated between about 7.0% and about 9.0%. In the multiple-dose study, the level of oxidation at Met sites 2 and 3 remained stable at levels of about 1.0% and 4.0%, respectively ( Figure 4B ).
[0179] N-terminal glutamine and glutamic acid at the N-terminus of mAb tend to form pyroglutamate via chemical or enzymatic cyclization. The conversion of glutamine occurs much faster than that of glutamic acid (Yin et al., Pharmaceutical research 2013; 30: 167-78; Li et al., mAbs 2016: 0; Li et al., mAbs 2016: 0; Liu et al., The Journal of biological chemistry 2011; 286: 11211-718). N-terminal cyclized proteins are resistant to digestion by aminopeptidases, thereby preventing degradation. Although glutaminyl cyclase, the enzyme that catalyzes this conversion, is found in human blood (Schilling et al., Biological Chemistry 2008;389:983-91), the formation of pyroglutamate from glutamate of mAb in vivo is primarily a pH-dependent non-enzymatic reaction, as it has the same rate constant in PBS in vitro (Yin et al., Pharmaceutical Research 2013;30:167-78; Liu et al., The Journal of Biological Chemistry 2011;286:11211-7). In a single-dose study, N-terminal pyroglutamate increased from 0.7% to 11.7% over 56 days ( Figure 5A ). By nonlinear fitting to N-terminal pyroglutamate levels measured by LC-MS, the in vivo N-terminal pyroglutamate rate was determined to be 0.2201% / day, where R 2 is 0.9970. In multiple dose studies, N-terminal pyroglutamate forms a sawtooth pattern similar to Asn deamidation ( Figure 5B ). Pyroglutamate levels increased during each dosing interval but dropped dramatically after each new dose due to dilution of the newly administered unmodified MAB1. Pyroglutamate levels rose to 7.6% before the fifth dose on day 56 and then dropped to 4.2% after the fifth dose ( Figure 5B ).
[0180] The C-terminal lysine residue of the mAb heavy chain is easily removed by alkaline carboxypeptidases during protein expression (Dick et al., Biotechnology and Bioengineering 2008; 100: 1132-43). Partial removal of the C-terminal lysine leads to charge heterogeneity. The percentage of MAB1 drug product containing the C-terminal lysine before administration was approximately 2.0%. In both the single-dose and multiple-dose studies, the C-terminal lysine was rapidly removed within one day after each dose (Figure 6). These observations are consistent with previously reported studies.
[0181] N-linked glycosylation at the Fc region is a commonly monitored PQA for mAbs. The major structure of N-linked glycoforms on Chinese hamster ovary (CHO) expressing mAbs is generally composed of a core biantennary pentasaccharide structure with one or more additional monosaccharides attached. In addition to complex biantennary oligosaccharide structures, high mannose glycoforms (e.g., Mannose 5) can also be observed on CHO expressing antibodies. In a single dose study, the relative abundance of the major glycoform (fucosylated complex biantennary) was unchanged over 56 days. However, the relative abundance of the Mannose 5 glycoform decreased from 0.5% to undetectable ( 0.1% ) within 6 weeks. Fig. 6A ). This finding is consistent with the faster serum clearance of high mannose glycoforms due to a high mannose receptor-mediated clearance pathway (Goetze et al., Glycobiology 2011; 21: 949-59; Alessandri et al., mAbs 2012; 4: 509-20). In the multiple dose study, the level of mannose 5 decreased during each dosing interval, but upon evaluation, it showed a sawtooth pattern immediately following each new dose due to the admixture of newly administered MAB1 with higher levels of mannose 5 ( Figure 6B ). Before the fifth dose, the level of mannose 5 decreased to 0.1%. After the fifth dose, the level of mannose 5 increased to 0.4%.
[0182] Modeling Subject Exposure to MAB1 PQA in Single-Dose Studies
[0183] Quantitative assessment of subject exposure to mAb quality attributes helps to assess the impact of quality attributes on drug safety and efficacy. As described in the Materials and Methods section, the in vivo serum concentration of mAbs with certain PTMs can be described as C PTM C(t) = C(t)·P(t). PTMThe area under the curve (AUC) of (t) corresponds to the exposure of the subject to the mAb with PTM (Goetze et al., mAbs 2010; 2: 500-7). In this work, deamidation at Asn sites 2 and 3 and N-terminal pyroglutamate formation are used as examples to demonstrate modeling of subject exposure to PQA in a single-dose study. Other PTMs discussed in this paper are not modeled because Met oxidation is not altered in vivo (Figure 4); C-terminal lysine is rapidly removed within hours after administration (Figure 6); the levels of mannose 5 and deamidation at Asn site 1 are extremely low (less than 0.5%, Figures 3 and 7), making modeling of these PTMs likely not of practical significance.
[0184] The MAB1 serum concentration-time curve is an equation, which is described by the two-compartment pharmacokinetic model equation as C(t) = Ae -αt +Be -βt , where A and B are confounding coefficients when α and β are confounding first-order constants. The experimental MAB1 serum concentrations from the single-dose PK study were determined using an ELISA assay. The modeling equation was solved by nonlinear fitting to the MAB1 serum concentrations measured by ELISA, where R 2 is 0.9916( Fig. 8A The best fit values for the parameters in the modeling equation are shown in Table 2. The subject's exposure to total MAB1 over 56 days, represented by the AUC of C(t), was 4302.0 μg / mL·day ( Fig. 8A ).
[0185] The rate equation for deamidation at Asn position 2 was determined in the previous section above to be P 脱酸胺化-Asn2 (t) = 1 - (1 - 0.01143) e -0.005394t Serum concentration-time equation C for MAB1 with deamidation at Asn site 2 脱酰胺化-Asn2 (t) = C(t)·P 脱酰胺化-Asn2 (t) can therefore be obtained through C(t) and P 脱酰胺化-Asn2 (t) by multiplying them. 脱酰胺化-Asn2 The curve of (t) is plotted on Figure 8B In. By C 脱酰胺化 The subject exposure to MAB1 with deamidation at Asn 2 over 56 days, represented by the AUC of (t), which constituted 9.9% of the subject exposure to total MAB1 over 56 days, was calculated to be 426.7 μg / mL·day ( Figure 8B ).
[0186] Similarly, the rate equation for deamidation at Asn position 3 was determined in the previous section above to be 脱酰胺化-Asn3(t) = 1 - (1 - 0.008848) e -0.001546t Serum concentration-time equation for MAB1 with deamidation at Asn site 3 C 脱酰胺化-Asn3 (t) = C(t)·P 脱酰胺化-Asn3 (t) can therefore be obtained through C(t) and P 脱酰胺化-Asn3 (t) by multiplying them. 脱酰胺化-Asn3 The curve of (t) is plotted on Figure 8B In. By C 脱酰胺化-Asn3 The subject exposure to MAB1 with deamidation at Asn 3 over 56 days, represented by the AUC of (t), which consists of 3.5% of the subject exposure to total MAB1 over 56 days, was calculated to be 152.5 μg / mL·day ( Figure 8B ).
[0187] The rate equation for N-terminal pyroglutamate formation was determined in the previous section above to be P PyroE (t) = 1 - (1 - 0.006868) e -0.002201t Serum concentration-time equation for MAB1 with N-terminal pyroglutamate C PyroE (t) = C(t)·P PyroE (t) can therefore be obtained through C(t) and P PyroE (t) by multiplying them. PyroE The curve of (t) is plotted on Figure 8B In. By C PyroE The subject's exposure to MAB1 with N-terminal pyroglutamate over 56 days, as represented by the AUC of (t), was 191.5 μg / mL·day ( Figure 8B ), which constituted 4.5% of the subjects' total MAB1 exposure over 56 days.
[0188] Table 2: Best-fit parameter values in the MAB1 serum concentration-time equation for the first dose interval from the single-dose PK study and the multiple-dose PK study
[0189] Research Single-dose studies First dose in multiple-dose study A 139.0925 57.8676 B 180.8075 25.3133 α 1.1227 0.0413 β 0.0389 0.9286 The standard error of the fitted A 15.4245 9.9061 Standard error of the fitted B 17.3316 9.7499 Standard error of the fitted α 0.3262 0.0185 Standard error of the fitted beta 0.0058 0.7592 <![CDATA[R 2 ]]> 0.9916 0.9818
[0190] Modeling Subject Exposure to MAB1 PQA in Multiple-Dose Studies
[0191] Using the area under the concentration-time curve, the subject's exposure to both total MAB1 and MAB1 with a specific attribute (e.g., Asn deamidation) can be calculated for any given time interval. The AUC of the predicted and experimental total MAB1 serum concentration-time curves represents the subject's exposure to total MAB1 over a defined time interval. Based on the predicted and experimental MAB1 serum concentration-time curves ( Fig.9A), the predicted and experimentally determined exposures of the subjects to total MAB1 over the course of 5 doses (56 days) were 5013 μg / mL·day and 4789 μg / mL·day, respectively. The difference between the predicted and experimental values was 4.7%. Similarly, the predicted and experimentally determined AUCs of the concentration-time curves for MAB1 with deamidation at Asn site 2 represent the exposure of the subjects to MAB1 with deamidation at Asn site 2 over a defined time interval. The predicted and experimentally determined exposures of the subjects to Asn site 2-deamidated MAB1 over the course of 5 doses (56 days) were 381.5 μg / mL·day and 352.6 μg / mL·day, respectively ( Fig.9A The predicted and experimentally determined exposure of the subjects to MAB1 2-deamidated at Asn as a fraction of the total MAB1 exposure over 56 days was 7.0% and 7.9%, respectively, confirming that the model accurately predicted the experimental results.
[0192] The predicted and experimentally determined exposures of subjects to MAB1 3-deamidated at Asn over the course of 5 doses (56 days) were 172.8 μg / mL·day and 158.8 μg / mL·day, respectively. Fig.9A ). The predicted and experimentally determined subject exposure to MAB1 3-deamidated at Asn as a fraction of the subject's total MAB1 exposure over the 56-day span was 3.4% and 3.3%, respectively.
[0193] Predicted and experimentally determined subject exposures to MAB1 with N-terminal pyroglutamate over the course of 5 doses (56 days) were 218.5 μg / mL·day and 231.8 μg / mL·day, respectively ( Fig. 10A ). The predicted and experimentally determined subject exposure to MAB1 with N-terminal pyroglutamate as a fraction of the subject's exposure to total MAB1 over the 56-day span was 4.3% and 4.8%, respectively.
[0194] Modeling the in vivo relative abundance of MAB1 PQAs in multiple-dose studies
[0195] The model can also be used to predict the pre-dose and post-dose levels of PQA for any given dose of mAb during a multiple dose study using Equation 12 and Equation 13, respectively. For example, the predicted pre-dose and post-dose levels of deamidation at Asn site 2 were 14.2% and 7.9% on day 56, respectively. The predicted values were consistent with the pre-dose and post-dose values determined by LC / MS on day 56, which were 15.0% and 7.9%, respectively. All other predicted deamidation levels ( Fig. 9B , hollow circles) are consistent with the experimental measurement results ( Fig. 9B , solid dots) are well consistent. The pre-dose and post-dose PTM levels in a multiple-dose study will reach a steady state because the pre-dose and post-dose mAb serum concentrations reach a steady-state concentration ( Fig. 9B ). The predicted pre-dose and post-dose steady-state levels (i.e., plateau levels) of PQA correspond to the highest and lowest levels of this PQA in vivo in the multiple dose study. Using Equation 16 and Equation 17, the pre-dose and post-dose steady-state levels of MAB1 with deamidation at Asn position 2 were determined to be 17.9% and 9.7%, respectively.
[0196] Similarly, the predicted pre- and post-dose levels of deamidation at Asn site 3 ( Fig. 9B , hollow triangle) and experimental measurement results ( Fig. 9B , solid triangles) are in good agreement. The pre-dose and post-dose steady-state levels of MAB1 with deamidation at Asn position 3 were determined to be 8.4% and 4.6%, respectively.
[0197] Predicted pre- and post-dose levels of N-terminal pyroglutamate ( Fig. 10B , hollow circles) and experimental measurement results ( Fig. 10B , solid dots) were well consistent. The pre- and post-dose plateau levels of MAB1 with N-terminal pyroglutamate were determined to be 8.1% and 4.5%, respectively.
[0198] Example Application of the Model
[0199] The quantitative assessment of PQA of therapeutic proteins in vivo is important because it helps to confirm potential critical quality attributes (CQA) that cannot be confirmed in vitro, thereby affecting product risk assessment and control strategies. The models described in this paper can be used to quantitatively evaluate PQA in many applications. First, these models can be used to calculate the in vivo progress of PQA in single-dose and multi-dose regimens and the exposure of subjects to PQA. The full consistency between model predictions and experimental measurements has been shown in the above section (Figure 8, Figure 9 and Figure 10). Second, the model can also be used to predict the exposure of subjects by extending the dosing period (t) and determine the pre-dose and post-dose steady-state concentrations / levels of PQA (Figure 9 and Figure 10), thereby providing insights for preclinical and clinical studies. Third, the model can be used to evaluate the impact of changes in the initial PQA level caused by process changes or round variability by adjusting the initial PQA level (P0) in the modeling equation. If PQA is CDR modified, this is of particular significance. Since the MAB1 used in this work does not have CDR modifications suitable for this modeling application, a hypothetical CDR deamidation with an in vivo deamidation rate of 2.5% / day is used as an example of using the model to assess the effect of the initial level of CDR deamidation on the subject's exposure to PQA. If there are three batches of drug product with initial levels of hypothetical CDR deamidation of 0%, 10%, and 20%, respectively. The hypothetical CDR deamidation of these batches in vivo over time can be described as P 脱酰胺化 (t) = 1 - (1 - 0) e -0.025t , P 脱酰胺化 (t) = 1 - (1 - 0.1) e -0.025t and P 脱酰胺化 (t) = 1 - (1 - 0.2) e -0.025t Using the model described above, the exposures of subjects in the single-dose study to the hypothetical CDR deamidated variants with 0%, 10%, and 20% initial deamidation over 56 days were 1385 μg / mL·day, 1653 μg / mL·day, and 1921 μg / mL·day, respectively, which constituted 32.5%, 38.8%, and 45.1% of the subject's exposure to the total mAb, respectively ( Fig.11A For the multiple-dose study, subject exposures to the hypothetical CDR deamidated variants with 0%, 10%, and 20% initial deamidation over 5 doses (56 days) were 1086 μg / mL·day, 1478 μg / mL·day, and 1871 μg / mL·day, respectively, which constituted 21.7%, 29.5%, and 37.3% of the subject's exposure to the total mAb, respectively ( Fig. 11B). Therefore, if process control produces an initial level of this hypothetical CDR deamidation of ±10%, the exposure of subjects in single-dose or multiple-dose studies to this CDR-deamidated mAb will vary by approximately 6%-8% over 56 days. Because the modeling indicates that subject exposure to this hypothetical CDR deamidation is insensitive to the initial level of the PQA, the acceptable range of the PQA can be broadened to allow for adjustments in process risk assessment. Combined with the effects of CDR modifications on potency measured by potency assays, the modeling results can be used to estimate subject exposure to active drug when initial CDR modifications vary due to process control or batch variations, thereby providing key information for product risk assessment.
[0200] Discussion
[0201] Compared with the static incubation of mAb in PBS or serum in vitro, the dynamic in vivo environment of circulating blood flow that leads to mAb clearance, elimination and degradation is more relevant to assessing the PQA changes of patients (Yin et al., Pharmaceutical research 2013; 30: 167-78). Therefore, in vivo PQA quantification and modeling provide key information for assessing the criticality of PQA that cannot be obtained from in vitro studies. For example, for example, these results presented herein (Fig. 6) and previous studies (Li et al., mAbs 2016: 0; Cai et al., Biotechnology and Bioengineering 2011; 108: 404-12) have shown that the C-terminal lysine in vivo is rapidly removed within one day. Similarly, it has been shown that trisulfide bonds are rapidly converted to disulfide bonds in vivo (Li et al., mAbs 2016: 0). The results presented herein ( FIG. 4 ) and previous studies ( Li et al., mAbs 2016: 0; Li et al., mAbs 2016: 0) indicate that the level of methionine oxidation is generally unaffected after in vivo administration. Therefore, C-terminal lysine removal, trisulfide bonds, and methionine oxidation are unlikely to be considered CQAs. In contrast, deamidation was shown to accumulate in vivo in single and multiple dosing regimens ( FIG. 3 ) and previous studies ( Li et al., mAbs 2016: 0; Li et al., mAbs 2016: 0). Deamidation that occurs and accumulates in the CDR region can potentially affect drug efficacy ( Yan et al., Journal of pharmaceutical sciences 2009; 98: 3509-21; Haberger et al., mAbs 2014; 6: 327-39) or cause off-target binding. Therefore, as demonstrated above, modeling CDR deamidation in vivo using the equations described in this paper can quantitatively assess the effect of CDR deamidation on exposure in a subject ( FIG. 11 ).
[0202] In this work, asparagine deamidation and N-terminal pyroglutamate formation were used as representative PTMs to demonstrate the validity of our model. Our model can also be applied to other PTMs that exhibit increasing or decreasing trends. For example, because of the enzymatic removal of mannose 5, the model was applied to evaluate mannose 5 clearance by adopting third-order dynamics. Although the relative abundance of mannose 5 decreased to less than 1.0% over time, the predicted values fully matched the experimental data (data not shown). Our model can be used to calculate other parameters to evaluate a given attribute or PTM in a multiple dose study. These parameters include the accumulation rate to reach steady state, the average level at steady state, and the degree of change of PTM at steady state.
[0203] In summary, the inventors characterized the common in vivo PTM changes of therapeutic mAb (MAB1), and modeled the in vivo PTM behavior of both single-dose and multiple-dose study regimens to evaluate the impact of in vivo PTMs. The three asparagine residues located in the Fc region of MAB1 exhibited different deamidation rates. The oxidation levels at the three methionine residues in the Fc region of MAB1 did not show changes in vivo over time. N-terminal pyroglutamate was formed rapidly in vivo. C-terminal lysine was completely removed within one day. MAB1 with high mannose glycosylation revealed accelerated clearance. We used two Asn deamidation and N-terminal pyroglutamate formation as representative PQAs and constructed modeling equations to calculate the serum concentration of PQAs, the exposure of subjects to PQAs, and the relative abundance of PQAs in vivo for both single-dose and multiple-dose regimens. The model predictions were verified by experimental measurement results. Therefore, the model can be used to simulate the in vivo PQA progression, exposure of subjects to PQA in both single-dose and multiple-dose regimens, thereby providing a quantitative method for PQA criticality assessment in therapeutic mAbs.
[0204] The scope of the present invention is not limited by the specific embodiments described herein. In fact, various modifications of the present invention, in addition to the modifications described herein, will become apparent to those skilled in the art from the foregoing description and the accompanying drawings. Such modifications are intended to fall within the scope of the appended claims. Sequence Listing <110> Regeneron Pharmaceuticals, Inc. Xu Xiaobin Huang Yu <120> Quantification and Modeling of Quality Attributes of Therapeutic Monoclonal Antibodies <130> 10421WO01 <150> 62 / 625,219 <151> 2018-02-01 <160> 1 <170> PatentIn version 3.5 <210> 1 <211> 7 <212> PRT <213> Artificial Sequence <220> <223> Met oxidized peptide <400> 1 Asp Thr Leu Met Ile Ser Arg 1 5
Claims
1. A method for determining the criticality of a product quality attribute (PQA) of an antibody after multiple administrations of the antibody, the method comprising: (a) Predicting the in vivo serum concentration of an antibody having PQA after a single administration of the antibody by: (i) using the in vivo rate constant determined for the PQA to predict the percentage of the antibodies having the PQA; and (ii) multiplying the predicted percentage of the antibody having the PQA by the total in vivo concentration of the antibody to determine the concentration of the antibody having the PQA, thereby generating a single administration model; (b) repeating step (a) at least once, thereby generating one or more additional single administration models; (c) superimposing the single administration models of steps (a) and (b) to generate a multiple administration model; and (d) determining the criticality of the PQA using the multiple administration model, wherein the PQA includes a target post-translational modification.
2. The method of claim 1, wherein the in vivo rate constant of the post-translational modification is determined by quantifying the percentage of the target post-translational modification over time and fitting the quantified percentage of the target post-translational modification to the equation Where P PTM (t) is the ratio of the post-translational modification over time; P0 is the initial post-translational modification percentage; and k PTM is the post-translational modification rate constant.
3. The method of claim 1, further comprising determining exposure of the subject to the antibody having the target post-translational modification, wherein determining exposure of the subject to the antibody having the target post-translational modification comprises determining an area under the curve (AUC) of the concentration of the antibody having the target post-translational modification.
4. The method of claim 1, wherein the serum concentration of the antibody is described by a two-compartment pharmacokinetic model equation: C(t) = Ae -αt +Be -βt , where α and β are first-order confounding constants, and A and B are confounding coefficients, respectively.
5. The method of claim 1, wherein the antibody comprises a recombinant therapeutic monoclonal antibody.
6. The method of claim 1, wherein the PQA is capable of affecting drug stability, safety and / or efficacy.
7. The method of claim 1, wherein the method is used to predict or monitor the post-translational modification profile of an antibody.
8. The method of claim 1, wherein the post-translational modification comprises one or more of deamidation, oxidation, glycation, disulfide formation, N-terminal pyroglutamate formation, C-terminal lysine removal, and high mannose glycosylation.
9. The method of claim 1, further comprising: (i) Modeling the correlation between pharmacokinetic (PK) concentration and ratio of post-translationally modified variants; (ii) quantitatively assess and predict subject exposure to representative post-translational modifications in a single-dose regimen or a multiple-dose regimen; (iii) predicting the maximum and minimum post-translational modification levels observed in the multiple-dose regimen; (iv) determining the total in vivo concentration of the antibody in the subject.
