Successful completion of EGFR/ALK testing in non-squamous non-small cell lung cancer (non-sq NSCLC) with the implementation of reflex testing (RT) by pathologists.

93 Background: EGFR mutation and ALK rearrangement testing is standard in the management of advanced non-sq NSCLC patients (pts). Previously at our centre, EGFR/ALK biomarker testing was requested by medical oncologists (MO). In June 2013 we implemented biomarker RT, defined as pathologists requesting EGFR/ALK at time of diagnosis of non sq-NSCLC irrespective of stage. The objective of this study was to test the hypothesis that if pathologists requested biomarker testing, appropriate tissue would be preserved and selected for testing, which would improve success rates of biomarker testing.

METHODS: Retrospective review of advanced non-sq NSCLC pts seen by MO at Sunnybrook Odette Cancer Centre from March 2010 to April 2014. Pt and EGFR/ALK test characteristics were compared before and after RT using Chi-square tests of association. Time outcomes were compared using Mann-Whitney U tests.

RESULTS: Of the 310 pts included, median age was 68, 44% female, 47% Caucasian, 93% adenocarcinoma, 22% EGFR+, 1% ALK+ and 84% either presented with or developed stage IV. Samples tested for EGFR and ALK were respectively: 53%, 51% core biopsies; 25%, 32% surgical resections; 20%, 16% cytology. The number of biomarker tests across all stages increased with RT (EGFR 70% vs 95%, p < 0.001 / ALK 44% vs 83%, p < 0.001). RT improved the rate of successfully completed tests (EGFR 86% vs 96%, p = 0.04 / ALK 83% vs 97%, p = 0.04). The remainder of tests were unsuccessful due to inconclusive results (EGFR 9% vs 4%, p = 0.25 / ALK 7% vs 2%, p = 0.25), insufficient tissue (EGFR 3% vs 0%, p = 0.33 / ALK 10% vs 0%, p = 0.03) or cancellation due to appropriate tissue not being sent from holding lab to testing lab (EGFR 2% vs 0, p = NS / ALK 0% vs 2%, p = NS). From core biopsies, there was trend to improved success of EGFR testing with RT (82% vs 97%, p = 0.06) and significant improvement of ALK testing (82% vs 100%, p = 0.04), with no impact on success from cytology samples. Rebiopsy rate for biomarker testing was low in both cohorts. Turnaround time for EGFR testing decreased [19 days (IQR 15-25) vs 17 days (IQR 12-21), p = 0.02]; ALK was unchanged.

CONCLUSIONS: Implementation of RT improved successful completion of EGFR/ALK testing.