Introduction: Advancements in genomic profiling now allow for routine comprehensive somatic genomic alteration testing in all patients with advanced cancer. A subset of patients will have targetable genomic alterations, though the frequency of these alterations and the efficacy of the matched treatments have varied amongst published data. Several commercially available platforms exist, but the ideal method to appropriately interpret and apply this data across various clinical tumor types and disease stages is still unclear.
Methods: We obtained a list of all the next generation sequencing (NGS) panels submitted from our center to the National Precision Oncology Program (NPOP). A total of 53 patients were included in the analysis. We analyzed the most frequently altered genes, the tumor types most frequently profiled, the frequency of cases with targetable alterations, and the efficacy of the matched treatments in individual patients. We also compared the number and types of alterations reported as well as the length of reports generated by the three different commercial NGS platforms used in our cohort.
Results: A total of 19/53 (35.8%) patients had targetable alterations. Five out of 21 (23.8%) received a targeted therapy. Non-small cell lung cancer [NSCLC] (n = 14; 26%) and prostate cancer (n=9; 17%) were the most frequently profiled tumors. In the NSCLC cohort, 7/14 (50%) had targetable alterations, including two patients in whom a prior single gene test for the specific alteration [EGFR, BRAF] was negative. NGS panels produced on average 6.6-13.0 alterations per patient, and average report length ranged from 8.3-19.0 pages.
Conclusions: NGS testing has been implemented by providers across a variety of tumor types at our institution, though the number of patients receiving matched treatments is low. Reflexive serial single-gene testing in NSCLC for EGFR, ALK, ROS1, and BRAF is likely reducing the number of NGS panels sent in these patients. Two false-negative single gene tests in our small cohort suggests we are underdiagnosing driver alterations in these patients with this approach. We would suggest exploring decision support tools and provider education in order to encourage judicious and clinically meaningful use of this valuable resource.