Induction of Immunogenicity of Monoclonal Antibodies by Conjugation with Drugs David A

(CANCER RESEARCH 51, 5774-5776, October 15, 1991] Advances in Brief

Induction of Immunogenicity of Monoclonal Antibodies by Conjugation with Drugs David A. Johnson,1 Russell L. Barton, Daniel V. Fix, William L. Scott, and Magda C. Gutowski

Lilly Research Laboratories, Indianapolis, Indiana 46285

Abstract Materials and Methods Human anti-mouse antibody has been a nearly consistent result of Antibody and Conjugates human clinical trials utilizing murine antibodies. It is generally antici pated that the problem of human anti-mouse antibody will be reduced as A murine monoclonal antibody directed against EGFr (9), 225 IgGl, genetically engineered, more human ("humanized") antibodies become was grown as ascites and purified using traditional protein A Sepharose available. It is not clear, however, what effect chemical modification of affinity Chromatographie techniques. The finca derivative DAVLB such "humanized" antibodies will have on their immunogenicity. The HYD was the generous gift of Mr. George Cullinan (Lilly Research present studies utilize a mouse antibody and rat host model to explore Laboratories, Indianapolis, IN). DAVLBHYD was conjugated to the aspects of this question. Rats injected with unmodified mouse monoclonal antibody as described utilizing a malonate linker (BAMME) (10) or a antibodies failed to mount anti-mouse immune responses, presumably BAP-based linker (11). due to their phylogenetic relatedness. In contrast, rats injected with a Rat Immunization Protocol Vinca immunoconjugate mounted strong anticonjugate antibody re sponses that were directed primarily against the linker portion of the conjugate. The in vivo serum pharmacokinetics of I2!l-labeled antibody Groups of young adult HarÃanASI brown rats were injected i.p. with unmodified 225 IgGl or 225 IgGl-BAP-DA VLBHYD on a twice- and conjugates were evaluated in rats with existing anticonjugate anti weekly x 2 protocol (4 doses total). Individual animals were bled, and body. The peak serum level attained was inversely correlated with the serum was stored frozen. level of reactivity of the anticonjugate antibody with the injected com pound. This model provides a potentially useful tool for exploration of Immunoassays the immunogenicity of drug, toxin, or radionuclide monoclonal antibody Indirect Solid-Phase Antigen-binding Assay. Target antigens consist conjugates. ing of unmodified 225 IgGl or 225 IgGl-BAP-DA VLBHYD were dried onto Falcon Microtest II flexible assay plates (Becton Dickinson, Introduction Oxnard, CA). Following washing, bound rat antibodies were detected using '"I-mouse anti-rat IgG (Jackson ImmunoResearch, West Grove, Monoclonal antibody-based therapies continue to be investi PA). This reagent had no detectable cross-reactivity to mouse antibod gated both preclinically and clinically. Efforts are under way to ies. The total bound radioactivity in sample wells was determined. explore the efficacy of unmodified antibodies and conjugates of Fluid Phase Competition Assay. Dilutions of rat sera with demon antibodies with drugs, toxins, or radionuclides. To date, the strated anticonjugate activity were mixed with excess quantities of 225 majority of human clinical trials have utilized murine monoclo IgGl-DA VLBHYD conjugates, free DAVLBHYD, irrelevant conju nal antibodies with the nearly universal finding that human gate, or irrelevant protein. These mixtures were added to wells of assay anti-mouse antibody is produced by the host (1-3). The field is plates as described above, to which 225-BAP-DAVLBHYD had been clearly moving toward utilization of genetically engineered, attached. After incubation and washing, target-bound rat antibody was more human ("humanized") antibodies with the goal of reduc detected using the '-'I-mouse anti-rat IgG. ing human anti-mouse antibody and its negative impact on Data are reported as averages of values from individual sera, with therapy (4-7). Such antibodies have undergone initial clinical SE indicated. evaluation with a favorable reduction in human anti-mouse Serum Clearance antibody, suggesting that "humanization" may be able to reduce Groups of five 225 IgGl-BAP-DA VLBHYD immunized rats were the significance of the immunogenicity issue (4, 6, 8). With injected i.p. with '"1-225 IgGl-BAP-DA VLBHYD, '"1-225 IgGl- regard to antibody conjugates, however, an additional problem BAMME-DAVLBHYD, or '"1-225 IgGl. Two million cpm of labeled might be predicted to arise which has not yet received significant attention. That is, chemical modification of "humanized" an antibody or conjugate were mixed with 500 fig of unlabeled carrier antibody or conjugate and injected in a 0.5 nil volume. Blood samples tibodies may render them once again more immunogenic. The were taken from the retroorbital plexus at the indicated time points, present study describes a preclinical model which allows explo serum was separated, and an aliquot was evaluated for radioactivity. ration of this concept. The model used mouse monoclonal antibody injected into rats under conditions in which little or no anti-mouse antibody response occurs with unmodified anti Results and Discussion body. A drug conjugate with the mouse antibody, however, Mouse monoclonal antibody (225 IgGl) was injected i.p. into induced ACA,2 which was directed primarily against the linker rats on a twice-weekly x 2 protocol at 25 mg/kg. This protocol and drug moiety of the conjugate. The effect of ACA on serum was chosen since it approximates an efficacious protocol with pharmacokinetics was explored. antibody drug conjugates (12). Fig. 1 shows that no detectable Received 8/14/91; accepted 9/5/91. rat antimouse antibody responses to unmodified antibody oc The costs of publication of this article were defrayed in part by the payment curred during the 45 days of the experiment. This may be due of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. to the close phylogenetic relationship between rats and mice, 1To whom request for reprints should be addressed, at Code 0444, Lilly such that, under this protocol, rats do not recognize mouse Research Laboratories, Lilly Corporate Center, Indianapolis, IN 46285. 2The abbreviations used are: ACA, anticonjugate antibody; DAVLBHYD, 4- antibody as being sufficiently foreign to be immunogenic. It is desacetylvinblastine-3-carboxhydrazide; BAP, /3-alanine pyrrole; BAMME, ÃŸ- clear, however, that under different conditions and protocols, alanine mÃ©thylÃ¨nemalonateethyl ester; EGFr, epidermal growth factor receptor. rats are capable of responding to mouse Ig. This is evidenced 5774

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1991 American Association for Cancer Research. IMMUNOCONJUGATE IMMUNOGENICITY tions had similar serum half-lives (35-40 h). It is notable, igCI* FREE 225 TARGET: 225 IgGl 8AP DAVLBHYD 1500400035003000250020001500 however, that the peak serum concentrations achieved differed, with the unmodified antibody being the highest, followed by the BAMME and BAP conjugates, respectively. This order is inversely related to the reactivity of the preexisting ACA with the free antibody, BAMME and BAP conjugates. That is, the BAP conjugate-induced ACA reacted most strongly with the POSITIVE CONTROLSERUMFREE BAP conjugate, which achieved the lowest peak serum level, 1000500 and least strongly with the unmodified 225 IgGl, which IgClSERUMo 225 o o v â€”¿o achieved the highest peak serum level. It is tempting to specu 0TARGET: 14 21 28 35 42 14 21 28 35 42 late that BAP conjugate was rapidly cleared by the preexisting

DAYS ACA. The similar half-lives suggest that the ACA may have Fig. 1. Indirect solid-phase radiometrie assay. Unmodified 225 IgGl (left) or been neutralized by the relatively high doses of antibody or 225 IgGl-BAP-DAVLBHYD (right) antigen targets were dried onto 96-well conjugates that were administered. plates and reacted with sera from rats injected with unmodified 225 IgGl (left) It is interesting to note that the peak serum concentration of or 225 IgGl-BAP-DAVLBHYD (right). Sera were collected at the indicated time points, and target-bound rat antibody was detected with '"I-mouse anti-rat IgG. the BAMME conjugate was only slightly below that of the unmodified antibody and significantly higher than that achieved by the positive control serum shown in Fig. 1, a rat anti-mouse with the BAP conjugate. This is also consistent with the inter mediate reactivity of the BAP conjugate-induced ACA with the serum provided by Dr. James J. Starling (Lilly Research Lab BAMME conjugate and suggests that it may be possible to oratories) that was produced in an unrelated study. avoid a particular ACA by switching to a similar but immuno- Fig. 1 also shows a parallel arm of the experiment in which 225-IgGl-BAP-DAVLBHYD Vinca immunoconjugate was in logically distinct immunoconjugate. jected into rats under an identical protocol. There was a very clear ACA response by day 21 that persisted throughout the remainder of the experiment. This result allows the conclusion that conjugation of DAVLBHYD to 225 IgGl via the BAP linker caused the otherwise nonimmunogenic antibody to be come significantly immunogenic under these conditions. CL The next question asked was to what portion of the molecule O was the rat ACA directed? To address this, a competition radioimmune assay was run with the ACA-positive sera. The target on the plate was 225-BAP-DAVLBHYD. As is shown on Fig. 2, there was no significant difference in the ability of diluent, bovine serum albumin, or unmodified 225 IgGl to compete with the anticonjugate sera. This indicates that the ACA was directed to a portion of the conjugate other than antibody. In contrast, bovine serum albumin-BAP-DAVLB- HYD and 225-BAP-DA VLBHYD both competed strongly with the anticonjugate sera, indicating that the response was directed toward the linker and/or drug portion of the conjugate. Free DAVLBHYD showed only slight competition, suggesting that the response is directed primarily toward the BAP linker. This Fig. 2. Fluid-phase competition assay. Dilutions of rat sera with demonstrated anticonjugate activity were mixed with the indicated compounds and reacted with conclusion is partially supported by the observation that a 225 225 IgGl-BAP-DAVLBHYD antigen target plates. Target-bound rat antibody IgGl DAVLBHYD conjugate made with a different linker was detected with '"I-mouse anti-rat IgG. (BAMME) showed reduced competition relative to the conju gate made with the BAP linker. Together, these data suggest that modification of an other wise nonimmunogenic mouse antibody with a drug and linker renders it immunogenic, and the ACA response, in this case, was directed primarily toward the linker portion of the mole cule. The reduced competition observed with a similar Vinca conjugate made with a different linker suggests that it may be possible to produce other drug conjugates using dissimilar UNMODIFIED 225 I linkers which have even less cross-reactivity. Indeed, in prelim inary studies with a conjugate made utilizing a novel linker 225 IgCI-SAHME-DAVieHYD under development, we have observed no competition with the

ACA generated in the present study. 225 IqGI-BAP-DAVLBHVD We next wished to determine whether the presence of ACA would significantly alter the serum pharmacokinetics of conju 0 12 24 36 48 60 72 gate versus unmodified antibody. Rats from the above studies, HOURS POST INJECTION with BAP conjugate-induced ACA, were injected with ' "1-225 Fig. 3. Serum clearance of antibody or conjugates. Groups of 225 IgGl-BAP- IgGl, '"1-225 IgGl-BAMME-DAVLBHYD, or '"1-225- DAVLBHYD-immunized rats were injected with the indicated compounds, which had been radiolabeled. Serum samples were taken at the indicated time points IgGl-BAP-DAVLBHYD. Fig. 3 shows that all three prepara and evaluated for radioactivity. 5775

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1991 American Association for Cancer Research. IMMUNOCONJUGATE IMMUNOGENICITY It is well understood that anti-hapten immune responses can Acad. Sci. USA, 86: 4420-4224, 1986. 5. Hutzell, P., Kashmiri, S., Colcher, D., Primus, F. J., Hand, P. H., Roseli!, be stimulated by conjugating small molecules (haptens) to larger M., Finch, M., Yarranton, G., Bodmer, M., Wittle, N., King, D., Loullis, C. carriers. Typically, highly immunogenic carriers are intention C., McCoy, D. W., Callahan, R., and Schlom, J. Generation and character ization of a recombinant/chimeric B72.3 (human 7 1). Cancer Res., 51: 181- ally chosen for this purpose. It is less clear that modifying 189, 1991. weakly immunogenic soluble proteins such as humanized 6. Hamblin, T. J., Caftan, A. R., Glennie, M. J., Mackenzie, M. R., Stevenson, monoclonal antibodies will render them more immunogenic in F. K., Watts, H. F., and Stevenson, G. T. Initial experience in treating human lymphoma with a chimeric univalent derivative of monoclonal anti-idiotype humans. The mouse-rat model described here may provide a antibody. Blood, 69: 790-797, 1987. tool for exploring this question preclinically. It may be possible 7. Beidler, C. B., Ludwig, J. R., Cardenas, J., Phelps, J., Papworth, C. G., to develop linker and chelate chemistry with low immunogenic Melcher, E., Sierzega, M., Myers, L. J., Unger, B. W., Fisher, M., David, G. S., and Johnson, M. J. Cloning and high level expression of a chimeric potential. Similar studies may lead to a better understanding of antibody with specificity for human carcinoembryonic antigen. J. Imimniol., the relative immunogenicities of drug, toxin, or radioconjugates 141: 4053-4060, 1988. 8. Saleh, M. N., Wheeler, R. H., Khazaeli, M. B., Liu, T. P., Allen, L., Grizzle, and the influence of linkers on this property. With this infor W., Tilden, A. B., and LoBuglio, A. F. A phase I trial of chimeric ami ( .1)2 mation in hand, it may be possible to avoid the most highly monoclonal antibody C14.18 in patients with metastatic melanoma. Antibody immunogenic configurations or, alternatively, to sidestep de Immunoconj. Radiopharmacol., 4: 207, 1991. 9. Sobol, R. E., Astarita, R. W., Hofeditz, C., Masui, H., Farishter, R., Royston, veloping immune responses by switching to other effective I., and Mendelsohn, J. Epidermal growth factor receptor expression in human conjugates possessing minimal cross-reactivity. lung carcinomas defined by a monoclonal antibody. J. Nati. Cancer Inst., 79: 403-407, 1987. 10. Barton, R. L., Guttman-Carlisle, D., Armour, K. H., Starling, J. J., Bumol, References T. F., Apelgren, L. D., and Koppel, G. A. Novel highly active lysine-based immunoconjugates: synthesis of a series of novel malonate linkers and their 1. Dillman, R. O. Monoclonal antibodies for treating cancer. Ann. Intern. utilization in preparing Vinca and adriamycin immunoconjugates. Antibody Med.,///: 592-603, 1989. Immunoconj. Radiopharmacol., 3: 65 (Abstract 108), 1990. 2. Larson, S. M. Clinical radioimmunodetection, 1978-1988: overview and 11. Scott, W. L., Horvath, S., Nichols, C. L., Johnson, D. A., Starling, J. J., suggestions for standardization of clinical trials. Cancer Res., 50 (Suppl. 3): Vorndran-Jones, C., Baker, A. L., Owens, R., Cullinan, G., Laguzza, B. C., 892s-898s. 1990. and Bumol, T. Drug releasability as a critical factor of active immunoconju 3. Murray, J. L., and Unger, M. W. Radioimmunodetection of cancer with gates: an evolution to new broadly applicable, lysine based, drug-releasing monoclonal antibodies: current status, problems, and future directions. Crii. linkers yielding active conjugates of antibodies and their fragments. Antibody Rev. Oncol. Hematol., 8: 227-253, 1988. Immunoconj. Radiopharmacol., 3: 65 (Abstract 109), 1990. 4. LoBuglio, A. F., Wheeler, R. H., Trang, J., Haynes, A., Rogers, K., Harvey, 12. Johnson, D. A., Baker, A. L., Laguzza, B. C., Fix, D. V., and Gutowski, M. E. B., Sun, L., Ghrayeb, J., and Khazaeli, M. B. Mouse/human chimeric C. Antitumor activity of L/lC2-4-desacetylvinblastine-3-carboxhydrazide monoclonal antibody in man: kinetics and immune response. Proc. Nati. immunoconjugate in xenografts. Cancer Res., 50: 1790-1794, 1990.

5776

David A. Johnson, Russell L. Barton, Daniel V. Fix, et al.

Cancer Res 1991;51:5774-5776.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/51/20/5774

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/51/20/5774. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.