Background and Importance Pembrolizumab showed longer overall survival compared with chemotherapy in the first-line treatment of advanced non-small-cell lung cancer (aNSCLC)

Aim and Objectives To evaluate the effectiveness of pembrolizumab in highly expressing patients with aNSCLC, comparing patients with PD-L1 expression ≥ 90% (very high) vs those with PD-L1 50-89% (high) in a tertiary hospital.

Material and Methods Observational, retrospective study. Inclusion criteria: patients with NSCLCa with pembrolizumab from August 2018 to August 2022. The clinical database of the Andalusian Health System (Diraya), its analytical module (Modulab) and the pharmaceutical validation program (Farmis-Oncofarm) were consulted. Variables collected: sex, age, ECOG (initial), smoking (current/past/non-smoker), percentage of PD-L1 expression and date of administration (first/last). Statistical analysis using the nonparametric Kaplan-Meier model with random censoring studying whether there is an increase in overall survival progression in very high versus high PD-L1 groups. A Cox regression model was included to analyse whether the rest of the variables studied affect overall survival.

Results 65 patients enrolled, 40 were included, 16 with very high PD-L1 and 24 with high PD-L1 expression (excluded 14 patients PD-L1<50% and 11 with 1 single administration of pembrolizumab). 73.17% patients were male with median age 43 years [80-37] and ECOG=1 [0-2]. 39.02% were current smokers, 53.65% were former smokers and 4.8% were non-smokers. Median overall survival in PD-L1(high) patients was 16.46 months vs 21.57 months median overall survival in PD-L1(very high) patients. P=0.92 is obtained from the PD-L1(High) versus PD-L1(very high) survival curves. Current smoking is the only variable with p=0.49 that positively affects the probability of death with respect to those studied (age, sex, ECOG, past smoking/non-smoking, PDL).

Conclusion and Relevance Survival results in PD-L1 patients (≥ 90%) compared to less expressors were positive but without statistically significant differences. It could be due to the small study sample. However, the median survival obtained is consistent with data from previous studies, it would be advisable to study this hypothesis in larger cohorts.

Conflict of Interest No conflict of interest.