<i>VHL</i> wild-type clinically advanced clear cell renal cell carcinoma (ccRCC): A genomic landscape study.

Abstract

533 Background: The most common alteration in both early and late stage ccRCC is of the von-Hippel Lindau ( VHL ) gene, which occurs in approximately 75% of cases. Here, we sought to identify differences in the genomic landscape between VHL mutated ( VHL mut) and VHL wild-type ( VHL wt) ccRCC to generate hypotheses for future targets and biomarkers. Methods: 2,137 cases of loco-regional or metastatic clinically advanced ccRCC underwent hybrid capture based comprehensive genomic profiling (CGP) using the FoundationOneCDx assay to identify all classes of genomic alterations (GA). Microsatellite instability status (MSI) and tumor mutation burden (TMB) were determined from the sequencing data. PD-L1 expression was determined by IHC using the Dako TPS score (0% = negative; 1-49% = low positive and ≥50% = high positive). All cases had relapsed either loco-regionally or with metastatic disease at the time of sequencing. Results: 1545 (72.3%) ccRCC cases were VHL mut and 592 (27.7%) were VHL wt. When comparing VHL wt vs VHL mut ccRCC, patients were more frequently of female gender (31.4% vs 26.0%; p = 0.013) and featured a similar age distribution (mean 61.3 vs 61.6 yr). When compared for biomarkers associated with potential immunotherapy response, results were overall similar: MSI-High status was extremely rare (0.0% vs &lt; 0.1%) while median TMB was low (2.6 vs 3.2 mutations/Mb). TMB &gt; 10 mutations/Mb was low in both groups but slightly higher in VHL mut (0.8% vs 2.3%; p = 0.022). Low positive PD-L1 expression status was relatively similar (29.6% vs 23.7%; not significant (NS)). The frequencies of GA in non- VHL tumor suppressor genes were similar for PBRM1 (35.3% vs 37.7%; NS), SETD2 (22.5% vs 24.9%; NS) and TP53 (15.4% vs 17.6%; NS), but there was a higher frequency of GA in BAP1 in the VHL mut group (14.4% vs 20.6%; p = 0.0008). GA in cell cycle regulatory genes were also more frequent in the VH Lmut ccRCC, including CDKN2A (16.4% vs 23.8%; p = 0.0002) and CDKN2B (12.5% vs 17.6%; p = 0.004). MTAP GA were similar (9.1% vs 10.2%; NS). GA in MTOR pathway genes including PTEN (10.0% vs 11.1%), PIK3CA (4.2% vs 4.9%), TSC1 (5.6% vs 6.7%), MTOR (5.6% vs 5.1%) and NF2 (6.3% vs 4.8%) were not significantly different between groups. Conclusions: Clinically advanced VHL wt ccRCC differ to some degree in their genomic landscape from VHL mut ccRCC. This may have implications for clinical trial design and future drug development. For example, the differences seen in frequency of CDKN2A/B alterations could impact combination strategies with cyclin dependent kinase 4/6 (CDK4/6) inhibitors that are currently being explored. Limitations include the retrospective nature, possible selection bias and lack of clinical data annotation. VHL mut ccRCC(1545 cases) VHL wt ccRCC(592 cases) p value PBRM1 37.7% 35.3% NS SETD2 24.9% 22.5% NS BAP1 20.6% 14.4% 0.0008 CDKN2A 23.8% 16.4% 0.0002 CDKN2B 17.6% 12.5% 0.004 MTAP 10.2% 9.1% NS PTEN 11.1% 10.0% NS TSC1 6.7% 5.6% NS NF2 4.8% 6.3% NS