Comparative genomic profiling from tumor tissue and circulating tumor DNA (ctDNA) in 111 patients (Pts) with metastatic clear cell renal cell carcinoma.

Abstract

4565 Background: Circulating tumor DNA (ctDNA) assessment is a non-invasive approach for genomic interrogation of solid tumors. As a novel tool, key benchmarks for applications in metastatic clear cell renal cell carcinoma (mccRCC) are yet to be determined. To understand the utility of ctDNA, we performed a large cohort analysis using a comparative genomics approach integrating matched primary tissue and ctDNA genomic data. Methods: Pts with prior tumor mutational profiles generated via next generation sequencing (NGS) from nephrectomy or metastatic specimens underwent single-time point plasma collection. Targeted NGS sequencing with MSK-IMPACT was performed on tumor and ctDNA with subsequent bi-directional cross genotyping using Waltz 2.0. All pts had matched germline comparison from peripheral blood; clinical data was extracted from medical record. Liberal (1-2 reads) and stringent (≥3 reads) filters were applied, with a cut-off of < 30% allele frequency to remove germline mutations. Results: 111 mccRCC pts, of whom available IMDC-risk was favorable (35%), intermediate (60%), and poor-risk (5%) were included for analysis. The median time between tissue and ctDNA collection was 23 months (R: 1-177), and 96% of patients had undergone nephrectomy prior to ctDNA collection. In primary tissue sequencing, 64/111 (58%) from nephrectomy and 42/111 (42%) from metastatic sites, 569 unique alterations were identified across the whole cohort, with a median of 4 mutations/pt (R:1-23). RCC-specific alterations included VHL (88%), PBRM1 (48%), SETD2 (34%), KDM5C (17%), TP53 (14%). Across the cohort, 176 alterations were identified in ctDNA. With cross genotyping, ctDNA alterations concordant with primary tumors were detected in 20% (22/111 pts, 28 unique alterations) using stringent criteria with a median of 1 mutation/pt (R:1-2). Using liberal criteria, concordance with primary tumors was 59% (66/111 pts, 142 unique alterations) with a median of 2 mutations/pt (R: 1-8). Conclusions: This large cohort study matching oncogenomics from tumor and ctDNA highlights complexities and challenges of applying liquid biopsy in biomarker development in mccRCC.