54P Efficacy of first-line immunotherapy for non-small cell lung cancer with MET exon 14 skipping according to PD-L1 expression

METΔ14ex is the driver alteration for approximately 3% of non-small cell lung cancers (NSCLC) and associated with a higher PD-L1 expression, but unclear benefit from immunotherapy (IO). Seventy-eight consecutive patients with metastatic NSCLC harboring METΔex14 who received first-line IO as monotherapy or chemoimmunotherapy (CHT+IO) in 10 German academic lung cancer centers were analyzed. The median age was 72 years (range 49-86), 34 patients (44%) were female, 47 (60%) were active or former smokers, and 23 (29%) presented with brain metastases. The Eastern Cooperative Group (ECOG) performance status was 0, 1, 2 and 3 in 27 (35%), 28 (36%), 18 (23%) and 4 (5%) cases, respectively. The most common histology was adenocarcinoma (n=61, 78%). IO was given to 43 (55%) patients as monotherapy, and to 35 (45%) combined with CHT. For patients with PD-L1 tumor proportion score (TPS) ≥50% (n=52, 67%), 1-49% (n=14, 18%) and <1% (n=12, 15%), disease control rates (DCR) were 56%, 57% and 100% (p=0.015), respectively. Other efficacy parameters including overall response rate (ORR), median progression-free survival (mPFS) and median overall survival (mOS) by PD-L1 tumor proportion score (TPS) and type of treatment are summarized in the table. Primary progressive disease/early death (before radiologic reassessment) under IO monotherapy, but not under CHT+IO, was significantly associated with never-smoker status (p=0.041). No significant correlations were found between smoking status and PD-L1 TPS (p=0.595). Table: 54PTPS≥50% / IO n=43TPS≥50% / CHT+IO n=9TPS 1-49% / CHT+IO n=14TPS 0% / CHT+IO n=12p-valueORR (%)355643500.599DCR (%)5467571000.030mPFS (mo)346150.520mOS (mo)14515160.690 Open table in a new tab Our exploratory analysis suggests an association between higher PD-L1 TPS and worse clinical outcomes under IO in patients with NSCLC harboring METΔ14ex. Although these results should be interpreted with caution, they contrast the favorable effect of PD-L1 expression for IO efficacy in other NSCLC and underline the need for alternative biomarkers for IO in this patient population.