Background – PRAME (Preferentially Expressed Antigen in Melanoma) is an intracellular cancer-testis antigen over-expressed across solid tumor types. Elevated PRAME expression is associated with poor prognosis, particularly in lung cancer. Further, its selective expression in tumor cells and ability to induce robust T-cell-mediated immune responses make it a promising pan-cancer immunotherapy target. This study investigates the influence of histological subtypes on PRAME expression and its impact on outcome to immunotherapy in first line (1L) lung cancer patients.

Methods – We analyzed 4,519 non-small cell lung cancer patients from the Tempus de-identified real-world database. PRAME expression levels (quantified by the Tempus xR RNA-seq assay), were classified into high and low based on median expression. We included pre-treatment biopsy samples receiving immunotherapy (IO) and chemotherapy (chemo), as individual monotherapies or in combination in the 1L setting. We evaluated PRAME expression by histology, tumor purity and biopsy site using a linear model and compared real-world overall survival and progression-free survival for high versus low expressors in each histological subtype.

Results – PRAME expression displayed a bimodal distribution in both adenocarcinoma (n=3,213) and squamous (n=1,306) histologies, with a significant enrichment of high expressors in the squamous subtype (Chi square p < 2.22 x 10^-16). The squamous group exhibited inferior rwOS and rwPFS compared to the adenocarcinoma group across all treatment groups (rwOS: IO+chemo: HR: 1.25, p<0.01, IO: HR: 1.63, p<0.01, chemo: HR: 1.31, p=0.02, rwPFS: IO+chemo: HR: 1.04, p=0.471, IO: HR: 1.13, p<0.01, chemo: HR: 0.88, p=0.297). Notably, high PRAME expressors in the squamous group correlated with improved overall survival only within the IO+chemo treatment group (HR: 1.28, p=0.027), while low PRAME expressors in the adenocarcinoma cohort demonstrated better survival outcomes compared to high expressors (HR: 0.79, p<0.01). PRAME expression was not significantly associated with outcome in the IO mono or chemo mono treatment groups. Similarly for rwPFS, high PRAME expression was only significantly associated with benefit in the IO+chemo group (HR: 0.81, p<0.01, IO: HR: 0.91, p=0.425, chemo: HR: 0.85, p=0.232).

Discussion – Our findings identify PRAME expression as a prognostic biomarker in lung cancer, with distinct implications based on histological context. High PRAME expressors may benefit from PRAME-targeted therapies combined with IO and chemo, highlighting the potential for personalized treatment strategies. These results emphasize the utility of PRAME expression in clinical treatment decisions and demonstrate the value of real-world data in identifying patients with unmet need.

VIEW THE PUBLICATION

VIEW THE POSTER