Abstract 3268
Background
To assess the role of I for second line treatment of APNS-NSCLC.
Methods
A pooled analysis of the final data of the CA209057, the KEYNOTE-010 and the OAK trial was performed. Overall Survival (OS) was the primary end point of the trial. The outcomes of patients with PD-L1 expression of 1%-49% (PD-L1 1%-49%), PD-L1 expression <1% (PD-L1<1%) or mutated-EGFR (EGFR+) were analyzed comparing any checkpoint inhibitor with standard chemotherapy. An indirect comparison with network meta-analysis was performed between the different checkpoint inhibitors whenever a significant difference was observed in the pooled analysis. Direct and indirect comparisons were performed using a random effect model.
Results
The outcome of 1720 patients was analyzed. 313 patients had been treated with Atezolizumab (A), 292 with Nivolumab (N), 270 with Pembrolizumab (P), and 845 with Docetaxel (D). The preliminary results were detailed in the table (Legend. CI95%: 95% Confidence Interval; *: Pooled Analysis; **: Network Meta-Analysis).Table: 1476P
| OS Hazard Ratio | CI95% | ||
|---|---|---|---|
| A vs D (PD-L1 1%-49%) | 0.571 | 0.423-0.771 | P < 0.001 |
| N vs D (PD-L1 1%-49%) | 0.62 | 0.467-0.882 | P = 0.001 |
| P vs D (PD-L1 1%-49%) | 0.76 | 0.604-0.956 | P = 0.019 |
| I vs D (PD-L1 1%-49%)* | 0.66 | 0.555-0.786 | P < 0.001 |
| A vs N (PD-L1 1%-49%)** | 0.921 | 0.609-1.392 | P = 0.696 |
| A vs P (PD-L1 1%-49%)** | 0.751 | 0.541-1.043 | P = 0.087 |
| N vs P (PD-L1 1%-49%)** | 0.816 | 0.597-1.116 | P = 0.491 |
| N vs D (PD-L1<1%) | 0.9 | 0.66-1.214 | P = 0.491 |
| A vs D (PD-L1<1%) | 1.04 | 0.619-1.747 | P = 0.882 |
| I vs D (PD-L1<1%)* | 0.933 | 0.72-1.21 | P = 0.601 |
| P vs D (EGFR+) | 0.88 | 0.453-1.71 | P = 0.706 |
| N vs D (EGFR+) | 1.18 | 0.693-2.004 | P = 0.542 |
| I vs S (EGFR+)* | 1.052 | 0.695-1.594 | P = 0.81 |
Conclusions
Our data seem to confirm the role of I for APNS-NSCLC with PD-L1 1%-49%. On the contrary, not-significant benefits in terms of OS seem to emerge for patients with PD-L1<1% or EGFR+ expression. Likewise, no significant differences seem to emerge from the indirect comparisons between A, N and P for patients with a PD-L1 1%-49% expression. Although all these data need to be analyzed with caution, as expression of indirect comparisons, waiting further conformations from clinical trials they can support clinicians for daily clinical practice.
Clinical trial identification
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Editorial Acknowledgement
Disclosure
All authors have declared no conflicts of interest.