Ocular toxicities observed in polypeptide vaccine trials for melanoma.

Abstract

e14643 Background: Peptide vaccines are a promising approach for melanoma treatment. They are generally well-tolerated but a subset of patients developed an ocular treatment-related adverse event (TRAE) during University of Virginia melanoma vaccine trials. We observed that some patients with severe ocular TRAEs had good clinical outcomes. We aimed to examine the clinical and immunogenic outcomes of patients who had an ocular TRAE on vaccine trial. We hypothesized that there is a correlation between vaccine immunogenicity and development of an ocular TRAE. Methods: Patients were identified by participation in an experimental peptide vaccine trial led at UVA from 2000 to 2023, where TRAE data was available. Information on duration of vaccine therapy, type and grade of ocular toxicity, time to resolution of toxicity (if resolved), and treatment of toxicity (if any) were collected. All events were reviewed by an ophthalmologist to ascertain the potential immunologic etiology of the toxicity. For patients with measurable disease on trial, clinical response was reported. Immune response, or PBMC T cell response, was determined using an interferon gamma ELIspot assay or proliferation assay. Results: Thirty five patients experienced ocular toxicity (5.3% of all trial patients), including 18 males and 17 females, with median age at enrollment of 56 years. Two patients had ocular melanoma and 33 had cutaneous melanoma. Ocular toxicities reported varied and included, but were not limited to, blurred vision, conjunctivitis, and uveitis. Of 20 patients with available toxicity grading, 15 had grade 1, 2 had grade 2, and 3 had grade 3 toxicities. Twenty four patients had complete resolution of toxicity and one had partial resolution. Per review with ophthalmology, uveitis, orbital edema, and exudative retinal detachment were possibly true TRAEs, and these five patients required ocular steroid injections or oral steroids for treatment. Of the 35 patients, twenty had favorable clinical response (i.e., no disease recurrence or progression) and nineteen (54%) had immune response, compared to 63% immune response in those without ocular TRAE. The majority of those with immune response had response at more than one time point. Of the twenty patients with favorable clinical response, fourteen had immune response (p = 0.044). Of the 5 patients with possibly true TRAE per ophthalmology review, only two had immune response (p = 0.642). Conclusions: Ocular TRAEs were reported during UVA melanoma vaccine trials, though only a subset were thought to be immunologic in origin upon retrospective analysis by ophthalmology. Of the patients with suspected immune-related ocular toxicity from the vaccine, few had immune response to vaccine. Also, immune response in those with ocular TRAE did not differ greatly from those without ocular TRAE. Therefore, our results confirm the expected clinical safety of vaccine therapy in melanoma. Clinical trial information: 00003222;00928902;00938223;00089219;00089193;00118274;00118313;01585350;02382549;02385669;02515227 .