Representative Sequencing: Unbiased Sampling of Solid Tumor Tissue

While hundreds of thousands of solid tumors have been sequenced to date, a fundamental under-sampling bias is inherent in current methodologies. This is caused by a tissue sample input of fixed dimensions (e.g. 6mm punch) from a single spatial location, which becomes grossly under-powered as tumor volume scales. Indeed our analysis of current clinical and research practice shows that existing protocols sample from only 0.0005% to 2.0% of the total tumor mass, raising the prospect of considerable sampling bias. Failure to address this bias risks undermining the clinical utility of genomic medicine in cancer; through lack of sensitivity to detect actionable mutations, miss-assignment of subclonal variants as clonal, and unreliable estimate of tumor mutational burden (TMB). Here we demonstrate Representative Sequencing (Rep-Seq), as a novel method to achieve unbiased sampling of solid tumor tissue. The Rep-Seq protocol comprises homogenization of all residual tumor material not taken for pathology into a well-mixed solution, coupled with next generation sequencing. Rep-Seq was implemented on a proof of concept basis, and benchmarked against current methods across multiple solid tumor types including renal cell carcinoma, lung cancer, melanoma, breast and colorectal cancer. Analysis of intra-tumor TMB variability showed a high level of misclassification with current single biopsy methods, with 20% of lung tumors, and 52% of bladder tumors, having ≥1 biopsy with high-TMB, but low clonal TMB overall (based on 10.0 mutations/Mb threshold for immune checkpoint inhibitor therapy (CPI)). Misclassification rates by contrast were reduced to 2% (lung) and 4% (bladder) when a more representative sampling frame was used. Clonal clustering analysis revealed rapid convergence of cancer cell fraction estimates in Rep-Seq towards true values, as validated in >60 biopsies, and >5ctDNA samples, taken from a single tumor. As a consequence 100% of variants were correctly classified as clonal by Rep-Seq, compared to ~85% in single biopsy sequencing. In terms of subclonal mutation detection, Rep-Seq achieved a greater sensitivity to detect variants, as compared to single-biopsy sequencing at equivalent depth (p=6.6x10-11). Finally, in a rapid autopsy setting Rep-Seq was able to accurately reconstruct the clonal phylogeny of advanced stage disease, correctly identifying polyclonal metastases, whereas single biopsy sequencing predicted monoclonal disease. Clinically this case showed lack of response to three lines of immune CPI therapy, which was observed in the context of a heterogeneous subclonal neoantigen repertoire. In conclusion, Rep-Seq effectively implements an unbiased tumor sampling approach, drawing DNA molecules from a well-mixed solution of the entire tumor mass, hence removing spatial bias inherent in current approaches. As a result Rep-Seq detects more mutations, achieves greater accuracy in determining clonal from subclonal variants. Rep-Seq offers an improved sampling protocol for tumor profling, with the same equivalent sequencing costs as current methods (single biopsy sequencing), but with significant potential for improved clinical utility.