Although EGFR-Tyrosine Kinase Inhibitors (TKIs) have improved disease management in EGFR mutation-positive NSCLC patients, all patients experience disease progression (PD) and ultimately die. In current practice, EGFR-TKIs are often continued beyond PD (RECIST1.1), however optimal management at PD remains ambiguous. We set out to study the survival differences and patient characteristics associated with continuation vs discontinuation of EGFR-TKI at PD retrospectively at a single institution: the Tom Baker Cancer Centre, AB Canada (2010-2014).
EGFRmut+ NSCLC patients treated with an EGFR-TKI were retrospectively analyzed. Demographic and clinical data, EGFR mutation type, treatment-variations at PD (chemotherapy, radiation, continuation with EGFR-TKI) were collected. SPSS-v19 was used to generate Kaplan-Meier survival curves. Progression-free survival (PFS) (time from initiating EGFR-TKI to PD) and time-to-complete EGFR-TKI-cessation (time from initiating EGFR-TKI to cessation of TKI completely, if extended beyond PD), were assessed.
127 EGFRmut+ NSCLC patients received EGFR-TKIs. 106/127 (83%) developed PD (RECIST 1.1) (64.8% F/35.2% M; 95% adenocarcinoma; smoking history: 47%; ex 19 del/L858R/other: 45.7%/ 35.4%/ 18.9 %). Median PFS was 7.9 months, while median overall survival (mOS) was 17.3 months. A log-rank test showed a statistical significant difference in mOS between exon 19 deletion and L858R (22.6months vs 15.0months; 95% CI, 14.8m to 21.9m; P-value = 0.014). At PD, 54.7% continued EGFR-TKI treatment; 45.2% received other forms of treatments. Female gender, PD with solitary-lesions and younger age are associated with EGFR-TKI continuation at PD.
This study confirms the difference in survival of exon 19 del vs L858R patients when treated with an EGFR-TKI. Additionally, it is common practice for oncologists to continue TKI treatment beyond PD. Further analysis, by time-to-complete EGFR-TKI cessation, will provide a more stratified base for future clinical-trials exploring treatment at EGFR-TKI-PD.
Clinical trial identification
Legal entity responsible for the study
University of Calgary
All authors have declared no conflicts of interest.