Abstract Background: Antibody-drug conjugates (ADCs) have emerged as a compelling therapeutic modality in oncology with several recent regulatory approvals. However, on and off-target toxicities require further evaluation. Corneal microcyst-like epithelial changes (MECs) are a well-known off-target class effect of ADCs with microtubulin-acting payloads, with varying degrees of severity. Manifestation of MECs is generally unrelated to antibody target and occurs in the absence of corneal target expression. Methods: An ocular toxicology model in Dutch belted rabbits was established using a humanized monoclonal antibody not cross-reactive in rabbits conjugated through an SPDB linker to the maytansinoid drug, DM4. Radiolabeled ADC was administered once and tissues were collected for radioanalysis at multiple time points to evaluate the contribution of ADC exposure and ocular tissue distribution to corneal toxicity. Non-radiolabeled ADC was administered once or twice to evaluate onset and characterization of MECs using slit lamp microscopy and histopathology. Furthermore, potential prophylactic methods for reducing MECs were evaluated. Results: The drug radioactivity was measured over 14 days and was widely distributed with elimination from most ocular tissues (except cornea) similar to blood/plasma elimination (T1/2 of 4.5-7.5 days). In contrast, the corneal layers had an apparent absorption phase followed by a relatively stable concentration plateau through 14 days. Furthermore, preferential distribution to the corneal epithelium was evident by high tissue-to-plasma ratio (1.6 vs <0.5 for other ocular tissues, including other corneal layers). Generally, MECs increased in frequency (number of eyes affected) and decreased in time to onset with increasing dose. Ophthalmic observations of MECs began in the perilimbal cornea with axial migration over time and were reflected by the histopathologic findings in the corneal epithelium of single cell necrosis, increased mitosis, mononuclear cell infiltration, and/or atrophy. A topical ophthalmic vasoconstrictor (brimonidine tartrate 0.2%) delayed the onset of MECs compared to controls (8% vs. 67% of eyes at 7 days; 33% vs. 100% at 10 days) and mitigated severity of some histopathological findings (reduced or absent corneal epithelial atrophy, punctate erosion, pigment, and/or mononuclear cell infiltration). In contrast, corticosteroid administration (prednisolone acetate 1%) did not reduce MECs clinically but did reduce the histologic incidence and severity of corneal pigment and mononuclear cellular infiltration. Conclusion: Collectively, this rabbit model reproduces the MECs associated with maytansinoid ADCs, demonstrates selective long-term corneal epithelium exposure, and provides data to support clinical testing of vasoconstrictor drops as ocular prophylaxis. Citation Format: Ali Warbington, Olga Ab, Sribalaji Lakshmikanthan, Shana Dalton, Paul Miller, Edward Stevens, Patrick Zweidler-McKay, Grace Lytle. Nonclinical ocular toxicity of a maytansinoid payload-antibody drug conjugate: Ocular tissue distribution, lesion pathogenesis, and mitigation [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2591.
Read full abstract