BackgroundTumor-specific antibodies have been reported in patients with cancers responding to immune checkpoint inhibitors (ICI), and there is an increasing appreciation for the potential role of B cells in mediating ICI responses. However, the humoral immune response to melanoma remains incompletely defined. We hypothesized that screening sera for antibodies by immunoprecipitation with lysates of cultured melanoma cells would increase the likelihood of detecting circulating antibodies in melanoma patients receiving ICI, and potentially identify novel antibody targets associated with treatment response and/or immune-related adverse events (IRAEs).MethodsPre-and on/post-treatment sera or plasma from 12 clinically-annotated patients with advanced metastatic melanoma receiving ICI were assayed for tumor-specific antibodies with an established immunoprecipitation platform. 35S-methionine-labeled lysates from cultured 624Mel cells were used for immunoprecipitation. 624Mel expresses several shared non-mutated melanoma antigens (e.g., MAGEA3, tyrosinase, MART-1/Melan-A, gp75, and gp100). Antigen identity was determined using on-bead digests followed by mass spectrometry, and was confirmed by immunoprecipitation with in vitro transcription/translation (IVTT) products.ResultsAntibodies reactive against 624Mel proteins were detected in 4 of 12 (33%) patients (table 1). Mass spectrometric sequencing performed on proteins captured with sera from 3 of 4 patients identified several putative antigens. Immunoprecipitation with IVTT candidate proteins confirmed antibodies against melanoma-associated and cancer testis antigens NY-ESO-1, SSX2 and MAGEA10. Antibodies were observed in 1 of 1 (100%) patient with a complete response, 2 of 4 (50%) with a partial response, 1 of 1 (100%) with stable disease, and 0 of 6 (0%) with progressive disease. Antibody levels varied over the course of therapy, with previously undetectable specificities arising during treatment response in patients #1–3. Patient #1 with a complete tumor regression developed antibodies to SSX2 and MAGEA10 that were absent before treatment. Further, detection of these antibodies coincided with diagnosis of IRAEs (anti-SSX2 with pancreatitis and anti-MAGEA10 with dermatitis). In contrast, patient #3, initially with a partial tumor regression, demonstrated a loss of detectable anti-NY-ESO-1 antibodies upon disease progression, and subsequent metastasectomy demonstrated loss of NY-ESO-1 protein expression in the progressing tumor. Testing sera from all 12 patients with IVTT products for NY-ESO-1, SSX2 and MAGEA10 did not reveal additional humoral responses.Abstract 231 Table 1Antibodies detected in the serum or plasma of patients with metastatic melanoma treated with ICI therapy. Treatment response indicates best overall response according to RECIST v1.1. Post-treatment blood collections were drawn during or after ICI therapy.ConclusionsOur comprehensive screening platform detected circulating antibodies specific to multiple melanoma-associated and cancer testis antigens in patients deriving clinical benefit from ICI. Expanded investigations of the evolution of antibody production over the course of ICI therapy, associated with tumor response to treatment and development of IRAEs, are warranted.AcknowledgementsThis study was supported by the Johns Hopkins Bloomberg-Kimmel Institute for Cancer Immunotherapy, and NIH P30-AR070254.Ethics ApprovalThis study was approved by the Johns Hopkins Institutional Review Board, approval #NA_00090257.
24P Using real-world evidence data and machine learning to identify molecular biomarkers for patient response to immune checkpoint inhibitors in metastatic melanoma
