Among NSCLC patients who received third-line therapy, we examined the association of PD-L1 expression and EGFR mutations with survival.
Third-line therapy NSCLC patients diagnosed during 2001-2012 were selected from the Danish Lung Cancer Group Registry. We retrieved patient data from population-based medical registries, and paraffin-embedded tumor tissue from pathology archives. We assessed PD-L1 expression using the Ventana IHC (SP263) validated assay (using 25% cutoff for high versus low), and genotyped EGFR to identify mutations at Exon 18 (G719X (G719A, G719C, and G719S), Exon 19 (deletions and complex mutations), Exon 20 (S768I, T790M, and insertions), and Exon 21 (L858R) via a PCR-based kit. Follow-up was from third-line therapy start to the first of death, emigration, or 31/12/2014. We used Cox regression to compute hazard ratios (HR) and associated 95% confidence intervals (95% CI) for PD-L1 and EGFR.
Among 344 third-line therapy patients, 185 (54%) were men, 165 (48%) were aged >60 years at diagnosis, and the majority were ever-smokers. 200 (58%) had adenocarcinoma histology, 85 (25%) had high PD-L1 expression in their tumors and 27 (8%) had EGFR mutations. Compared to patients with wildtype EGFR tumors, those with EGFR mutations less often showed high PD-L1 expression (21%, 95%CI = (4%, 38%) versus 25%, 95%CI = (19%, 31%); Fisher's exact p-value = 0.81). Median overall survival differed little by PD-L1 status, but was longer in patients with mutant compared with wildtype EGFR. Neither PD-L1 expression nor EGFR mutations were significantly associated with survival.
|N||No. of deaths||Median Survival (months) (95%CI)||Adjusted HR* (95%CI)|
|PD-L1 low||249||208||8.0 (6.6, 9.1)||ref.|
|PD-L1 high||85||72||8.3 (6.1, 11.6)||0.89 (0.68, 1.17)|
|Wildtype||260||213||7,7 (6.5, 8.8)||ref.|
|Mutant||27||22||11.0 (6.9, 22.1)||0.79 (0.51, 1.24)|
*Adjusted for age, sex, histology (adenocarcinoma versus other)
Our findings suggest no association of PD-L1 expression or EGFR mutations with survival in third-line therapy NSCLC patients.
Clinical trial identification
Legal entity responsible for the study
T. Dalvi: TD is an employee and shareholder of MedImmune/AstraZeneca. E. Hedgeman: Funding to support this work from AstraZeneca Inc., Gaithersburg, MD 20878. A. Midta, N. Shire, R. Brody, D. Lawrence, D. Potter, J. Walker: Employee and shareholder of AstraZeneca. J. Fryzek: JPF was employed by AstraZeneca, 2009-2011. JPF received funding to support this work from AstraZeneca Inc., Gaithersburg, MD 20878. J. Rigas: Consultant physician to AstraZeneca. A. Mellemgaard: Honoraria and/or travel expenses from Lilly, Boehringer Ingelheim, BMS, MSD. S. Hamilton-Dutoit: Research funding from AstraZeneca. Honoraria or consulting Amgen. H.T. Sørensen: Research funding from Epidstat. All other authors have declared no conflicts of interest.