Genomic Characterization Of Pulmonary-Metastatic Prostate Cancer: A Unique Molecular Subtype.

210 Background: Metastatic prostate cancer (mPC) rarely spreads to the lungs and lung-only metastatic disease has not been clearly defined. This study investigated the clinical and molecular features of lung-only mPC. Methods: This is a retrospective single-institution study. Medical records of 40 metastatic prostate patients with lung-only (N=25) or lung-predominant (N=15) disease were analyzed. Germline and/or somatic DNA sequencing results, when available, were recorded. Results: 25 lung-only mPC patients were identified, of which 14 underwent germline/somatic DNA analysis. Nine (64%) had a driver mutation in either: the WNT pathway (29%), PI3K/PTEN pathway (21%), homologous recombination DNA repair (HRD) (36%) or mismatch repair (MMR) (29%). 15 lung-predominant mPC patients were also analyzed, of which 10 underwent genomic analysis. Seven (70%) had a driver mutation in either: WNT (10%), PI3K/PTEN (40%), HRD (40%), or MMR (21%). Representative driver mutations for both cohorts summarized in Table. In comparison, in our genomic database, 52.1% had a driver mutation (in WNT 15%, PI3K/PTEN 33%, HRD 22% or MMR 4%). The lung-only and lung-predominant cohorts did not differ by age (p=0.41), race, PSA (p=0.16), Gleason sum (p=0.77), intraductal/ductal histology (p=0.71), lymphovascular invasion (p=0.64), perineural invasion (p=0.73), or having relatives with cancer (p=0.22). Our results are limited by the small sample size and retrospective nature of the analysis. Conclusions: In this exploratory study, the overall prevalence of DNA repair gene alterations was higher than anticipated (35%), suggesting a potential enrichment (esp. in MMR mutations) in men with lung-only or lung-predominant mPC. This finding may have important therapeutic implications. These results require prospective validation of the hypothesis that pulmonary-tropic mPC biology may be fundamentally different from non-pulmonary mPC, and that these patients may benefit from PARP inhibitor or anti–PD-1 treatments. Clinical outcomes data will also be presented. [Table: see text]