Feasibility of whole-genome sequencing in routine clinical practice.

3013 Background: In the next few years numerous drugs will be approved for defined genomic targets, most of these in a tumor agnostic manner. Identifying patients who can benefit from this is critical for the future success of precision medicine, ideally using a single comprehensive test to detect all possible biomarkers. The WIDE study (WGS Implementation in standard cancer Diagnostics for Every cancer patient) aimed to evaluate the feasibility, clinical validity (primary endpoints) and added value (secondary endpoint) of clinical grade Whole Genome Sequencing (cWGS) in routine clinical practice. Methods: cWGS was prospectively performed on 1,200 consecutive patients with (suspected) metastatic cancer. Tumor material was obtained during routine clinical procedures for both Standard-Of-Care (SOC) and cWGS. Next to securing a high quality specimen for SOC diagnostics, multiple frozen sections per patient were evaluated to identify the sample most suitable for WGS. cWGS was conducted independently of, but in parallel with SOC diagnostics, which included SOC molecular diagnostics (Moldx) for 48% of patients. cWGS and MolDx results were compared and discussed in a dedicated tumor board. Additional tests for resolving discordances were applied when needed. Results: cWGS was successfully performed in 69% (841/1217) of samples with a technical success rate of 97% (841/871). An insufficient amount of tumor cells ( < 20%) was the main reason for not completing cWGS (25%, 310/1217). cWGS turn-around-time (TAT), due to continuous improvements to the clinical procedure and cWGS pipeline over the course of the study, decreased to 10 working days. A total of 856 genomic biomarkers identified by SOC MolDx could be compared to cWGS results. Initial analyses of discordances revealed an error rate of 2.1% (18/856) for cWGS compared to a 1.0% (8/856) error rate for SOC Moldx. After optimizing cWGS and SOC pipelines based on these findings, error rates dropped to 0.6% (5/856) and 0.7% (6/856) for cWGS and SOC MolDx, respectively. Overall, cWGS identified clinically actionable (routine practice and experimental) biomarkers in 71% of all patients tested. Compared to SOC MolDx, cWGS identified one or more additional clinically actionable biomarkers in 54% (446/832) of patients. Interestingly, in patients who were not tested by SOC MolDx, actionable variants were identified by cWGS in 54% (153/282). Conclusions: The WIDE study has shown that cWGS is feasible in routine clinical practice in a comprehensive cancer center setting, using tumor material obtained during routine procedures. Furthermore, cWGS showed added value by identifying one or more additional clinically actionable biomarkers in 54% of patients including patients who had not received SOC Moldx. These outcomes have led to the successful adoption of cWGS at the Netherlands Cancer Institute as part of routine care, which will further facilitate precision medicine for cancer patients.