Presentation Authors: Roy Mano*, Renzo Di Natale, Esther Drill, New York, NY, Alexander Sankin, Bronx, NY, Andrew Winer, Brooklyn, NY, Andrew Silagy, Julian Marcon, Kyle Blum, Ed Reznik, Jonathan Coleman, Irina Ostrovnaya, Paul Russo, A. Ari Hakimi, New York, NY
Introduction: Recent data suggests clear cell renal cell carcinoma (ccRCC) evolutionary subtypes, defined by specific driver mutations and mutation clonality, correlate with clinical phenotypes and outcome. Mutation clonality determination relies on multi-regional sequencing due to intra-tumoral heterogeneity (ITH). We evaluate whether it is possible to correctly identify mutational clonality by sequencing pooled multiregional tumor samples.
Methods: After obtaining informed consent, ex-vivo core needle biopsies were obtained from 6 geographically distinct regions of 5 primary ccRCC. DNA was extracted using standard techniques. Each tumor region was sequenced individually and as a pooled sample and profiled for genomic alterations using ultra-deep sequencing (MSK-IMPACT).Based on previous publications, clonal mutations were defined when cancer cell fraction (CCF) was >0.5 in all regions of multi-regional samples. We used a threshold of CCF >0.7 to identify clonal mutations in the pooled sample. ITH index, calculated as # subclonal driver mutations / # clonal driver mutations, and fraction of copy number altered genome (FCNA) were calculated for each tumor using the multiregional and pooled samples.
Results: Median tumor size was 6 cm (IQR 4.2, 7.5) and tumor stage was â‰¥pT3 in 4/5 patients. Samples were sequenced with an average coverage depth of 905x (range 509x-1920x). VHL mutations were identified in all samples and were defined as clonal in 4/5 tumors. Additional driver mutations identified were PBRM1, PIK3CA, KDM5C, BAP1, TP53 and TSC1. All mutations detected in regional samples were also identified in the pooled sample. Mutation clonality was correctly assigned in the pooled sample for 7/7 driver mutations and 14/16 of all mutations (Figure 1A). Median ITH index across all tumor samples was 0.67 (0.25, 1) for both the multiregional and pooled samples (Figure 1B). Median FCNA was 0.08 (0.07, 0.11) for the multiregional samples and 0.07 (0.04, 0.11) for the pooled samples (Figure 1C).
Conclusions: Our findings suggest that deep sequencing of pooled multiregional samples correctly identifies the type and clonality of all driver mutations within the tumor. Pooling of multi-regional tumor samples may serve as a simple and cost-effective technique to overcome limitations associated with ITH.
Source of Funding: This research was funded by the Stephen P. Hansen Family Fund Fellowship in Kidney Cancer Award._x000D_
This research was funded in part through the NIH/NCI Cancer Center Support Grant P30 CA008748.