Association of progression free survival with the timing of genomic testing in non-small cell lung cancer: Evidence from the GuardantInform clinic-genomic database.

Abstract

e21143 Background: The clinical utility of ctDNA analysis for biomarker identification in advanced non-small cell lung cancer (NSCLC) has been established, supporting clinical adoption of this testing modality. All FDA-approved, immune-checkpoint inhibitor (ICI) immunotherapies require the prerequisite of negativity for ALK rearrangement and EGFR mutation, or prior toxicity or progression on oral therapies targeting mutations in these genes. Despite these recommendations and support from the American Society of Clinical Oncology and National Comprehensive Cancer Network, broad-panel marker testing for targeted therapy options in NSCLC continues to be underperformed. Data are even limited on when those tests are typically conducted and the effects of timing on the treatment outcomes. Methods: The GuardantINFORM™ clinic-genomic database was interrogated for patients over the age of 18 having a Guardant360 test positive for EGFR mutations of L858R or exon 19 after 2016 and known to have a confirmed diagnosis of lung cancer one year prior to testing. Patients had to have at least three claims prior to, and 90 days of follow-up after, their test. Patients were stratified into three groups by pre- and post-test treatment: 1) patients treated with an EGFR tyrosine kinase inhibitor (TKI) as a first-line treatment after the test, with no other chemotherapy, ICI immunotherapy (IO), or targeted therapy before the test; 2) patients with other first-line treatments prior to the test and then EGFR TKI immediately following the test; and 3) patients treated with ICI or chemotherapy after the test and before EGFR TKI or patients not treated after the test. Real-world time to next treatment (rwTTN) was defined as the index date to the treatment different than the index therapy. Progression free survival (PFS) time was defined as rwTTN or time to death, whichever came earlier. Patients who experienced adverse events (AEs) associated with chemotherapy or IO were reported as a percentage (p-value<0.001). Results: Among the 3 cohorts of patients (384 in each group who were matched on age with difference < 3, gender, and follow-up time with difference < 4 months) identified, a statistically significant difference in PFS was discovered between group 1 ( EGFR TKI as first-line treatment) and the other two groups based on the Cox proportional hazard model [hazard ratio=1.8, p-value<0.001, median survival time for group 1= 26 months (95% CI 23-29) vs. 17 months (95% CI 14-19)]. No difference existed in PFS between groups 2 and 3. The proportion of patients experiencing treatment-associated AEs was lowest in group 1 (13.5% vs 15.9% in group 2 vs 30.2% in group 3 for ICI-related AEs, 16.9% vs 23.2% vs 38.5% for chemo-related AEs, p-value<0.001). Conclusions: Performing genomic testing sooner, as early as first-line treatment, may improve the treatment response for patients with NSCLC.