Huang, Minqi

247 Background: Approximately 10% to 15% of men with metastatic castration-resistant prostate cancer (mCRPC) harbor germline mutations. Circulating tumor-derived DNA (ctDNA) as non-invasive approach to characterizing somatic alterations. In this study, we report longitudinal ctDNA assessments in mCRPC patients who have undergone germline testing. Methods: A total of 238 patients, who had undergone both germline testing and ctDNA assessment from Tulane Cancer Center were included in this study. Germline testing was conducted using a multi-gene cancer panel from Invitae, covering 50-86 genes, while somatic alterations in ctDNA were tested through Guardant 360, including 70-83 genes. Somatic alterations with <0.1%, synonymous, and variants of unknown significance were excluded from analyses. Statistical analyses were performed using Pearson Chi-Square Test or Fisher Exact Test. Results: From 2015-2024, 17.2% (41/238) had pathogenic or likely-pathogenic (P/LP) germline mutations. The most common pathogenic germline mutations were in BRCA2 (17.1%; 7/41), BRCA1 (9.8%; 4/41), and CHEK2 (9.8%; 4/41). In ctDNA, germline positive patients were more likely to have frameshift mutations (OR= 1.75, 95% C.I. [1.09, 2.81], p=0.02) and less likely to have copy number amplifications (OR= 0.58, 95% C.I. [0.36, 0.93], p= 0.03) compared to germline negative patients. Among mCRPC patients with germline HRR P/LP mutations, frameshift alterations were significantly more frequent (OR=5.03, 95% CI [2.03, 13.23], p-value=0.0008). mCRPC patients with germline P/LP mutations were more likely to have somatic mutations in BRCA2 (OR=2.85; 95% CI: [1.62, 4.93]; p=0.0002) and GATA3 (OR=3.37, 95% CI: [1.26, 8.51]; p=0.016) detected ctDNA and less likely to have EGFR alterations (mutations and/or amplifications) (OR=0.46; 95%CI: [0.25, 0.80]; p=0.0094). BRCA2 somatic mutations are also more likely to be detected in patients with germline P/LP mutations in HRR genes (OR=0.11, 95% CI [0.03, 0.33], p=0.00004). Conclusions: mCRPC patients with pathogenic germline mutations are more likely to have frameshift and less likely to have copy number amplifications compared to germline negative mCRPC patients. Frameshift mutations are more likely to be detected in patients with germline P/LP in HRR genes. Patients with germline P/LP mutations are more likely to have somatic mutations BRCA2 and GATA3 and less likely to have mutations and/or amplifications in EGFR.

245 Background: Circulating tumor DNA (ctDNA) can be used as a prognostic biomarker in mCRPC patients, with changes in ctDNA providing insight into patient survival. In this study, we evaluate the ctDNA landscape of patients with mCRPC prior to their death. Methods: A retrospective clinical data review, 2015-2024, was conducted at Tulane Cancer Center to identify mCRPC patients who underwent ctDNA testing prior to death. All ctDNA data were obtained through a multi-gene cancer panel via the Guardant 360 assay (70-83 genes). Lethal mCRPC was defined as patients who underwent ctDNA testing ≤3 months prior to death. A matched control group (CTRL) was defined as mCRPC patients with ctDNA collection ≥2 years prior to death or most recent follow-up. Statistical analyses were performed using Pearson Chi-Square Test or Fisher Exact Test whenever appropriate. Furthermore, survival analysis was conducted using the Kaplan-Meier method. Results: There were 70 patients in lethal mCRPC group and 157 patients in CTRL group. Compared to the CTRL group, in lethal mCRPC, somatic mutations and/or amplifications in BRAF(OR=2.1549, 95% C.I. [1.1136,4.1523], p=0.0298), CCND2 (OR=3.7925, 95% C.I.[1.1937,13.3380], p=0.03326), MYC (OR=2.6332, 95%C.I. [1.4158, 4.9120], p=0.0029), PDGFRA (OR=2.5235, 95% C.I. [1.1438, 5.5793], p=0.0291), and PIK3CA (OR=2.0328, 95% C.I. [1.1098, 3.7166], p=0.02846) were significantly more frequent. Additionally, lethal mCRPC patients were more likely to have missense (OR=4.3194, 95% C.I. [1.443456,19.3338], p=0.01687) and nonsense (OR=1.9364, 95% C.I. [1.06091,3.5250], p=0.04112) mutations at the end-of-life. Patients with a higher allelic fraction (≥0.6%) of TP53 alterations had a shorter median survival (p-value= 0.034). Conclusions: ctDNA mutations and/or amplifications in BRAF, CCND2, MYC, PDGFRA, and PIK3CA were more frequently detected in lethal mCRPC compared to matched CTRLs. Additionally, mCRPC patients with a lethal phenotype were more likely to have missense and nonsense mutations compared to the CTRL group. CtDNA assessment of copy number alterations was limited to amplifications only; additionally, with this assay, distinguishing between gene level and large scale chromosomal amplification is not assessable. Though this study is limited to a single-institution with retrospective analyses of clinical testing, understanding the somatic landscape of lethal mCRPC is critical for development of new biomarkers, strategies and ultimately druggable targets.

212 Background: Patients with prostate cancer with liver metastases are known to have a worse prognosis. In this study, we evaluated ctDNA in mCRPC (metastatic castration-resistant prostate cancer) patients with and without liver metastases. Methods: A retrospective chart review identified mCRPC patients with liver metastases at Tulane Cancer Center and Huntsman Cancer Institute. Liver metastases were confirmed via molecular or conventional imaging. All patients had ctDNA assessed with a multi-gene cancer panel via the Guardant 360 assay. Alterations with <0.1% allelic fraction, synonymous alterations, and variants of unknown significance were excluded from analyses. Statistical analyses were performed using Fisher’s Exact and Wilcoxon Rank Sum Tests. Correction for multiple testing was not performed, hence these alterations may be subject to a type 1 error. Results: 53 mCRPC patients with liver metastases and 144 mCRPC patients without liver metastases were included in the study. We found seven genes that are more likely to be pathologically altered in ctDNA in mCRPC patients with liver metastases (See Tables 1-3): AR (OR=2.38, 95% CI: [1.25, 4.52], p=0.01), BRAF (OR=2.74, 95% CI:[1.05, 7.18], p=0.05), BRCA2 (OR=5.33, 95% CI:[1.49, 19.02], p=0.01), CDK6 (OR=4.09, 95% CI:[1.35, 12.42], p=0.01), EGFR (OR=3.12, 95% CI:[1.22, 7.99], p=0.02), FGFR1 (OR=5.69, 95% CI:[1.81, 17.86], p-value=0.003), MYC (OR=4.55, 95% CI:[1.63, 12.68], p=0.004), and PIK3CA (OR=4.34, 95% CI:[1.87, 10.06], p=0.01). Additionally, mCRPC patients with liver metastases are more likely to have copy number alterations (CNA) (p= 0.01) compared to mCRPC patients without liver involvement. There were no other significant differences in the remaining 70 genes interrogated. Germline pathogenic alterations did not vary between the two groups. Conclusions: In mCRPC patients with liver metastases, these exploratory analyses suggest ctDNA testing is more likely to detect pathologic alterations in AR, BRAF, BRCA2, CDK6, EGFR, FGFR1, MYC, and PIK3CA. Additionally, mCRPC patients were more likely to have copy number alterations (CNA). Limitations of the Guardant 360 assays are notable in this setting given the incomplete ascertainment of gene deletions and the incomplete number of prostate-relevant genes assessed. [Table: see text]