Abstract 3218: The Secretome: a multigene biomarker for the prognosis and selection of solid tumor patients evaluated for treatment with immune checkpoint inhibitors

Abstract Introduction: The advent of immune checkpoint inhibitors (ICIs) has been revolutionary for cancer treatment. However, only a minority of solid tumor patients significantly benefit from them as a monotherapy. Approximately 3 out of 4 patients receiving ICIs experience a broad range of immune-related (ir) adverse events (AEs), including many AEs of grades 3 and 4 as well as fatalities (0.3 - 1.3%). Accurate criteria to select the best candidates, sparing other patients from unnecessary physical and financial burden, are urgently needed. Methods: We developed a Secretome signature, based on 50 genes, utilizing The Cancer Genome Atlas (TCGA) lung squamous cancer (LUSC) RNA-Seq data, to identify patients whose tumor is characterized by abundant immune cell infiltration. This signature was generated through gene selection and generalized linear models. An algorithm was created to make this signature portable across multiple solid tumor types and sequencing platforms. We used this algorithm to assign a Secretory-like (SL) status to a cohort of 78 patients of the University of North Carolina (UNC) - Chapel Hill, mostly diagnosed with lung and head and neck cancer and having received ICIs. A retrospective chart review collected their socioeconomic and racial/ethnicity demographics, smoking and alcohol use, and characterized cancer outcome and irAEs. We used categorical data analysis to investigate salient features of the SL group. After dichotomizing patients according to their SL status, we performed Gene Set Enrichment Analysis (GSEA) of their RNA-Seq data. Using Kaplan-Meier Plotter - Immunotherapy, we assessed the role played by Secretome genes in overall survival (OS). Results: GSEA shows that main features found in TCGA LUSC cohort (e.g. enrichment of the ‘Hallmark TNFA signaling via NFKB’ gene set) are also present in the UNC SL cohort. These results reveal that our model, which strongly constraints the number of genes, can still retrieve patients with key biological features displayed by the group of TCGA LUSC Secretory. SL lung cancers are more prevalent in Black and Hispanic/Latinx (BHL) than in White (P = 0.02), which generally points towards a worse prognosis for BHL patients. SL patients have > 3 times post-ICIs dermatologic AEs than non-SL (P = 0.01). Remarkably, the hazard ratios of high vs. low gene expression levels show that > 80% of testable Secretome genes are individually predictive of OS in patients receiving anti-PD1 ICIs, supporting the combined use of Secretome genes for optimal performance. Conclusions: The Secretome can be used to identify patients, who are most likely to be associated with prolonged survival after ICI therapy. The Secretome can also characterize patients (i) expected to have a worse prognosis and (ii) who should be more closely monitored when using ICIs, since they have an increased risk of developing dermatologic AEs. Citation Format: Alessandro Porrello, Marjory Charlot, Michael M. Cunningham, Shruti Saxena-Beem, Leon G. Coleman, Rumey G. Ishizawar, Benjamin G. Vincent, Chad V. Pecot. The Secretome: a multigene biomarker for the prognosis and selection of solid tumor patients evaluated for treatment with immune checkpoint inhibitors. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3218.