Effects of DrugâAntibody Ratio on Pharmacokinetics, Biodistribution

Subscriber access provided by University of Newcastle, Australia

Article

Effects of Drug-Antibody Ratio (DAR) on Pharmacokinetics, Biodistribution, Efficacy and Tolerability of Antibody-Maytansinoid Conjugates Xiuxia Sun, Jose F. Ponte, Nicholas Yoder, Rassol Laleau, Jennifer Coccia, Leanne Lanieri, qifeng qiu, rui wu, Erica Hong, Megan Bogalhas, Lintao Wang, Ling Dong, Yulius Setiady, Erin Maloney, Olga Ab, xiaoyan zhang, Jan Pinkas, Thomas Keating, Ravi V.J. Chari, Hans erickson, and John M Lambert Bioconjugate Chem., Just Accepted Manuscript • DOI: 10.1021/acs.bioconjchem.7b00062 • Publication Date (Web): 07 Apr 2017 Downloaded from http://pubs.acs.org on April 9, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Bioconjugate Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bioconjugate Chemistry

Effects of Drug-Antibody Ratio (DAR) on Pharmacokinetics, Biodistribution, Efficacy and Tolerability of Antibody-Maytansinoid Conjugates Xiuxia Sun, Jose F. Ponte, Nicholas C. Yoder, Rassol Laleau, Jennifer Coccia, Leanne Lanieri, Qifeng Qiu, Rui Wu, Erica Hong, Megan Bogalhas, Lintao Wang, Ling Dong, Yulius Setiady, Erin K. Maloney, Olga Ab, Xiaoyan Zhang, Jan Pinkas, Thomas A. Keating, Ravi Chari, Hans K. Erickson, John M. Lambert

Note: following authors have same address Jose F. Ponte, Nicholas C. Yoder, Rassol Laleau, Leanne Lanieri, Qifeng Qiu, Rui Wu, Megan Bogalhas, Lintao Wang, Ling Dong, Yulius Setiady, Erin K. Maloney, Olga Ab, Jan Pinkas, Thomas A. Keating, Ravi Chari, John M. Lambert

ImmunoGen, Inc., 830 Winter Street, Waltham, Massachusetts 02451-1477, United States

Present Addresses Xiuxia Sun, Shire, 300 Shire Way, Lexington, MA 02421 Jennifer Coccia, Regeneron Pharmaceuticals Inc, 777 Old Saw Mill River Road, Tarrytown, New

York 10591, United States Hans Erickson, Genentech, 1 DNA Way, South San Francisco, California 94080, United States

Erica Hong, Finnegan, Henderson, Farabow, Garrett&Dunner, LLP, 2 Seaport Lane, Boston, MA, 02210, United States Xiaoyan Zhang, Sanofi, 55 Corporate Drive, Bridgewater, NJ 08807, United States

Author Contributions Co-first authors: Xiuxia Sun and Jose F. Ponte

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Corresponding Author: Jose F. Ponte, ImmunoGen Inc 830 Winter Street, Waltham, MA 02451. Tel: 781895-0852. Fax 781-895-0611. E-mail: [email protected]


Page 2 of 31

Page 3 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Abstract Antibody-drug conjugates (ADCs) are being actively pursued as a treatment option for cancer following the regulatory approval of brentuximab vedotin (Adcetris®) and ado-trastuzumab emtansine (Kadcyla®). ADCs consist of a cytotoxic agent conjugated to a targeting antibody through a linker. The two approved ADCs (and most ADCs now in the clinic that use a microtubule disrupting agent as the payload) are heterogeneous conjugates with an average drug to antibody ratio (DAR) of 3-4 (potentially ranging from 0-8 for individual species). Adotrastuzumab emtansine employs DM1, a semisynthetic cytotoxic payload of the maytansinoid class, which is conjugated via lysine residues of the antibody to an average DAR of 3.5. To understand the effect of DAR on the preclinical properties of ADCs using maytansinoid cytotoxic agents, we prepared a series of conjugates with a cleavable linker (M9346A-sulfoSPDB-DM4 targeting folate receptor alpha (FR)) or an uncleavable linker (J2898A-SMCC-DM1 targeting the epidermal growth factor receptor (EGFR)) with varying DAR, and evaluated their biochemical characteristics, in vivo stability, efficacy, and tolerability. For both formats, a series of ADCs with DARs ranging from low (average ~2, range 0-4) to very high (average 10, range 7-14) were prepared in good yield with high monomer content and low levels of free cytotoxic agent. The in vitro potency consistently increased with increasing DAR, at constant antibody concentration. We then characterized the in vivo disposition of these ADCs. Pharmacokinetic analysis showed that conjugates with an average DAR below ~6 had comparable clearance rates, but for those with an average DAR around 9-10, rapid clearance was observed. Biodistribution studies in mice showed that these 9-10 DAR ADCs rapidly accumulate in the liver, with maximum localization for this organ at 24-28% percentage injected dose per gram (%ID/g) compared with 7-10% for lower DAR conjugates (all at 2-6 h post-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

injection). Our preclinical findings on tolerability and efficacy suggest that maytansinoid conjugates with DAR ranging from 2-6 have a better therapeutic index than conjugates with very high DAR (~9-10). These very high DAR ADCs suffer from decreased efficacy likely due to faster clearance. These results support the use of DAR 3-4 for maytansinoid ADCs, but suggest the exploration of lower or higher DAR may be warranted depending on the biology of the target antigen.


Page 4 of 31

Page 5 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Antibody-drug conjugates (ADCs) represent a promising approach for cancer therapy that combine the antigen targeting specificity of monoclonal antibodies with the high potency of cytotoxic agents. ADCs have become a widely investigated modality for cancer therapy recently, in part due to the success of ado-trastuzumab emtansine (T-DM1, Kadcyla®) 1 and brentuximab vedotin (Adcetris®).2 The development of ADCs with therapeutic potential involves the optimization of several critical parameters including the choice of target antigen, antibody selection, linker chemistry, method of conjugation, the cytotoxic agent, and the drug to antibody ratio (DAR), where “drug” refers to the cytotoxic agent. The two approved ADCs (and most ADCs currently in the clinic that use a tubulin agent as the payload) are heterogeneous, with an average DAR of 3-4 (ranging from 0-8 for individual species). Heterogeneity results from conjugation to a subset of the ~90 antibody lysine residues (as for T-DM1) or of the 8 cysteine residues derived by partial reduction of the 4 interchain disulfide bonds (as for brentuximab vedotin). It has been shown previously that higher DAR correlates with increased in vitro potency3, 4, but it can also adversely affect biophysical and pharmacological properties.3, 5 For example, increased loading of monomethyl auristatin E (MMAE) has been reported to negatively impact the pharmacokinetic properties of the resulting ADCs, apparently due to its high hydrophobicity.3 Similarly, ADCs of the cytotoxic agent doxorubicin conjugated to high DAR were reported to be prone to aggregation.5 Recently it has been found that the detrimental effects of payload hydrophobicity can be modulated through linker-payload design. Thus, ADCs with high DAR showing good in vivo properties have been synthesized by using hydrophilic linkers.6 Similarly, a recently reported polymer-based ADC platform permits even higher cytotoxic agent loading (DAR of ~20) without compromising the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

physicochemical and pharmacokinetic properties of the ADCs, due to the high hydrophilicity and polyvalency of the polymer.7 For maytansinoid conjugates, the properties of ADCs with high values for DAR have not been explored in detail, and the majority of ADCs in clinical and preclinical evaluation use an average DAR of 3-4. To understand the effects of different average DAR ranges on the preclinical properties of maytansinoid ADCs, and especially to evaluate whether DAR 3-4 is the optimal range for these ADCs, we prepared a series of maytansinoid-ADCs containing a cleavable disulfide linkage (M9346A-sulfo-SPDB-DM4, targeting folate receptor alpha (FR), currently being evaluated in clinical trials as a 3-4 DAR ADC)8 or a non-cleavable thioether-containing linkage (J2898A-SMCC-DM1, targeting the epidermal growth factor receptor (EGFR); this SMCC-DM1 linker-cytotoxic agent format was utilized in ado-trastuzumab emtansine) with different DARs, and evaluated their biochemical characteristics, in vivo stability, efficacy, and tolerability. ADCs with high DAR (average 10, range 7-14) were prepared for both formats, however, these high DAR ADCs suffered from decreased efficacy in vivo likely due to faster distribution and clearance via the liver. ADCs with an average DAR below ~6 had similar clearance rates, superior in vivo efficacy, and comparable tolerability based on the linked maytansinoid dose.


Page 6 of 31

Page 7 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Results and Discussion To evaluate the impact of DAR on the preclinical properties of maytansinoid ADCs, a series of ADCs containing a disulfide-cleavable linkage (M9346A-sulfo-SPDB-DM4, targeting FR) or an uncleavable link (J2898A-SMCC-DM1, targeting EGFR) with different DARs were prepared. Although ADC disposition may vary based on the properties of the antibody, we believe that the general trends identified will inform the design of future ADCs. The linker-maytansinoid moieties were conjugated to the antibody via lysine residues through amide bonds (Figure 1A). These conjugation reactions proceeded in good yields (unoptimized yield of ≥52% when DAR was lower than 8.5, 33-49% at DAR around 10), high monomer content (>95%) and low free drug levels (98%) in plasma is associated with ADCs as free maytansinoids are rapidly cleared.14 PK studies were conducted by administering a single i.v. injection of 10 mg/kg radiolabeled ADCs to CD-1 mice (Figure 3). There was a minimal impact on PK for ADCs with DAR range from 1.8 to 5.1. However, at DAR 5.8-6.8, while there was only a slight deviation of the full clearance curves towards more rapid clearance, the values for clearance (CL) and steadystate volume (Vss) increased by about 50 to 80%, rising to about 5-fold higher values at DAR 9.7- 10.4 as calculated from these clearance curves. Remarkably, the findings are consistent



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

between maytansinoid-ADCs with non-cleavable and cleavable (hindered disulfide) linkages, suggesting that the biophysical properties of these two classes of ADCs are similar. In comparison, endogenous cysteine-linked MMAE ADCs with DAR 8 were reported to clear 3fold faster than a purified fraction with DAR 4, and 5-fold faster than a purified fraction with DAR 2.3

Figure 3. Pharmacokinetic studies evaluating ADC concentrations using radiolabeled ADCs. Clearance of M9346A-sulfo-SPDB-[3H]DM4 (A) and J2898A-SMCC-[3H]DM1 (B) conjugates from plasma of CD-1 mice that were injected IV as a single 10 mg/kg dose were measured by counting the radioactivity in plasma arising from the tritium label on the maytansinoid. The heterogeneity of the lysine-linked maytansinoid-ADCs means that each ADC contains a significant amount of material above the nominal (average) DAR. For example, a DAR 6.2


Page 14 of 31

Page 15 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


conjugate contains species up to DAR 10 (Figure 1B). The similar PK profiles of DAR 6.8 J2898A-SMCC-DM1 and DAR 6.0 M9346A-sulfo-SPDB-DM4 compared with lower DAR ADCs suggests that conjugation of maytansinoid imposes a lower hydrophobicity penalty on a molar basis than MMAE conjugation. Thus, ADCs with a high load of MMAE may be cleared more rapidly than maytansinoid-ADCs due to this higher hydrophobicity. Of note, the selfimmolative aromatic ring in the linker of MMAE-ADCs is not present in the final metabolite (MMAE)15 but it does contribute to the overall hydrophobicity of these ADCs. Another factor that may contribute to the hydrophobic characteristics of an ADC, besides the hydrophobicity of the payload/linker system itself, may be the site of linkage of the payload to the antibody. It is possible that the conjugation of hydrophobic payloads in close proximity to each other at the endogenous paired cysteine residues of the hinge region of the antibody (as for the MMAEADCs) results in greater effective hydrophobicity of the final ADC via creation of a hydrophobic “patch”.16 It is also conceivable that maytansinoid-ADCs can bear higher drug loading than endogenous cysteine conjugates before incurring a significant pharmacokinetic liability because conjugation of the hydrophobic maytansinoid to lysine residues results in the payload moieties being widely dispersed over the surface of the antibody. For example, in the case of the maytansinoid conjugates lorvotuzumab mertansine17 and trastuzumab emtansine18, peptide mapping identified 40 or 70 lysine sites conjugated out of ~90 total lysine residues, respectively. To assess the effect of maytansinoid loading on tissue distribution, biodistribution studies were performed with the two radiolabeled ADCs, M9346A-sulfo-SPDB-[3H]DM4 and J2898ASMCC-[3H]DM1, each with varying DAR. As in the PK studies, ADCs were injected in CD-1 mice at single dose of 10 mg/kg. There was a similar tissue distribution pattern for ADCs with DAR lower than 5.8 for the cleavable linker format, and lower than 6.4 for the uncleavable linker



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(Figures 4A and 4B). As observed in the PK study, the blood concentration for ADCs with very high DAR (9.5 for M9346A-sulfo-[3H]DM4 and 9.8 for J2898A-SMCC-[3H]DM1) rapidly decreased over time. At the same time, these very high DAR ADCs rapidly distributed to the liver, with the maximum %ID/g for this organ reaching 24-28% at 4-6 h post-injection as compared to 7-10% for ADCs with DAR 3.2-3.7 (at 0.5-2 h post-injection), which is consistent with the faster clearance of the very high DAR conjugate. The biodistribution of J2898ASMCC-DM1 with a very low DAR (0.7), which could mimic the biodistribution of naked antibody, was similar to the conjugates with DAR 3.7 and 5.8. The finding is consistent with a previous report by Xie et al. that maytansinoid conjugation (at DAR 4.1) does not alter tissue biodistribution compared to the unconjugated antibody.14


Page 16 of 31

Page 17 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Tissue distribution of (A) M9346A-sulfo-SPDB-[3H]DM4 and (B) J2898ASMCC-[3H]DM1 in CD-1 mice after intravenous administration at a dose of 10 mg/kg. At each time point, three animals were sacrificed, bled, and their organs collected, weighted, and assayed for tritium radioactivity. Data are presented as %ID/g of tissue and values are the mean±SD derived from three different animals.

The single-dose tolerability of ADCs with varying DAR was evaluated in CD-1 mice. For M9346A-sulfo-SPDB-DM4 conjugate, the body weight loss in mice dosed i.v. with ADCs of varying DAR at total administered maytansinoid doses of 1.25 mg DM4/kg were similar, at ~ 49% body weight loss, suggesting that the maximum tolerated doses (MTDs) for these ADCs were similar with respect to maytansinoid dose. Close inspection of the curves in Figure 5A suggests that the weight loss of mice dosed with ADCs with lower DAR (2 and 3.5) trends



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

towards less severe (nadir ~ 4% body weight loss) compared to ADCs with DAR 5.5, 6.8 and 9.8 (nadir 7-9% weight loss). The body weight loss in mice dosed with J2898A-SMCC-DM1 ADCs having varying DAR at the same total maytansinoid doses, were comparable (9-11% body weight loss, close to MTD), thus suggesting that the MTD for these ADCs was close to 3.0 mg DM1/kg irrespective of DAR values in this range (2.3-10.1) (Figure 5B).

Figure 5. The tolerability study of ADCs with varying DAR in CD-1 mice. CD-1 mice (8 mice/group) were dosed intravenously with either M9346A-sulfo-SPDB-DM4 (A) or J2898ASMCC-DM1 (B) conjugates with varying DAR at a single dose based on maytansinoid respectively. Mice were weighed daily, and their body weight loss was expressed as a percentage of their initial weight before administration of the ADCs.

The effect of drug loading on in vivo antitumor activity was evaluated in severe combined immune-deficient (SCID) mice bearing subcutaneous human tumor xenografts. In the efficacy


Page 18 of 31

Page 19 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


studies, conjugates were dosed at constant maytansinoid dose, given the above finding that the MTDs were determined by the total amount of maytansinoid administered irrespective of DAR, and the fact that in practice, the recommended phase II doses for ADCs in clinical evaluation are the highest achievable, limited by dose-limiting toxicities.19 We evaluated the effect of DARs in the range of about 2 to 10 with J2898A-SMCC-DM1 in the EGFR-positive H1703 xenograft model (Figure 6A). The lack of bystander activity of the SMCC-DM1 format ensured that the observed tumor cell kill was entirely the result of antibody-mediated delivery of the payload. 13 The J2898A-SMCC-DM1 at DAR 9.6 was far less efficacious than the lower DAR ADCs (6.3, 3.9 and 2.3) at the same DM1 dose (Figure 6A). The relatively lower activity of the ADC with very high DAR is consistent with its faster clearance (see Figure 3), suggesting that even though more payload can be delivered to a tumor cell per individual binding event with a high-DAR ADC (see Figure 2), the poor PK behavior of a high DAR ADC results in less overall delivery of payload into tumor in vivo. The comparable activity of J2898A-SMCC-DM1 with DAR from 2.3 to 6.3 at the same DM1 dose is consistent with their similar pharmacokinetic behavior (see Figure 3). The efficacy of ADCs with disulfide linkers, M9346A-sulfo-SPDB-DM4 with DAR values of 1.6, 3.3 and 5.2, dosed at a maytansinoid dose of 40 µg/kg (single dose), was evaluated in the FRα–positive KB xenograft model (Figure 6B). The ADCs in this range of DAR were all highly active (T/C values

Effects of DrugâAntibody Ratio on Pharmacokinetics, Biodistribution

Recommend Documents

Effects of DrugâAntibody Ratio on Pharmacokinetics, Biodistribution