PD1/PD-L1 inhibitors (IO) can be prescribed as first line treatment in high PD-L1 positive NSCLC pts. There is an important need for additional predictive factors to identify pts with PDL1 weak or negative NSCLC that could benefit from these.
In this retrospective study, pts with stage III/IV NSCLC eligible for an IO were selected. All pts consented for tissue analysis. Pt characteristics and outcome were collected. A NGS panel on 52 genes was performed with an immunochemistry analysis (PDL-1 with the SP-263 clone, CD8, PTEN, beta-catenine, MSI, FOXP3, IDO-1 and CD163). We used image analysis with density results. PD-L1 was classified as negative/weak/positive if 0/1-9%/10%+ of tumor cells were stained.
Sixty-seven pts were enrolled. Median age was 64 years, 8 pts were never smokers, 90% had PS 0-1, 11.3%/58.1%/30.6% received an IO as 1st/2nd/3rd line or more, 69% had a non-squamous carcinoma. 38.7% of the tumors were PD-L1 positive, and 15% weak. Median progression-free survival (PFS) was 3.5 months (IC95%, 1.9-7.6), 12 months overall survival rate was 63.3% (IC 95% 46.6-76.1). The objective response rate (ORR) was 50.8%. In univariate analysis PS, line of IO, positivity for PD-L1 (cut off 10%), CD8 (H-SCORE ≥ 284,4), FOXP3 (H-SCORE ≥ 155,4) and IDO-1 (H-SCORE ≥ 0,4) were significantly correlated with ORR and PFS. ORR was 77% in IDO-1 positive (n = 26), 32% in IDO-1 negative (n = 25) NSCLC pts. KRAS mutation, smoking status, histological type, response to platinum-based chemotherapy were not correlated with PFS and ORR. In multivariate analysis, positive PD-L1 and IDO-1 were the only factors correlated with ORR. ORR was 87.5% if both positive (n = 16), 60% if one of them was positive, 22.7% if both negative. Only IDO-1 was correlated with PFS.
Along with PD-L1, IDO-1 appears as a promising predictive factor for IO. A prospective validation is ongoing.
Clinical trial identification
Legal entity responsible for the study
Sylvestre Le Moulec
All authors have declared no conflicts of interest.