Immune checkpoint inhibitors (ICIs) have significantly improved survival for advanced wild-type non-small cell lung cancer, but there is no direct comparison to confirm which first-line treatment may lead to the longest overall survival. What qualifies as long-term survival (LS) is even unclear.
By searching PubMed, Embase, and the Cochrane Central Register of Controlled Trials from January 2005 to December 2020, we included randomized controlled trials (RCTs) of first-line ICI-containing treatments to perform an integrated analysis (IA) to determine the criterion of LS and then screened regimens with LS for network meta-analysis (NMA). The main outcomes for NMA were median overall survival (mOS), 1-year survival rate (1ySR), and 2-year survival rate (2ySR); those for IA were the pooled mOS (POS), 1ySR (P1SR), and 2ySR (P2SR).
By IA of 16 first-line ICIs from 20 RCTs, the POS was 16.20 (95% CI 14.79–17.60) months, with P1SR of 63% (95% CI 59–66%) and P2SR of 37% (33–41%). Thus, we defined LS as mOS ≥ POS (16.20 m) for regimens and screened for RCTs with outcomes meeting this criterion. Eleven ICI-based regimens can bring LS for the overall population, among which ICI with bevacizumab and chemotherapy achieved the longest POS of 19.50 m (16.90–22.10 m) and the highest P1SR (74%, 61%–87%) and P2SR (49%, 38%–61%). Pembrolizumab with chemotherapy ranked first in mOS and 1ySR, while atezolizumab plus bevacizumab and chemotherapy ranked first in 2ySR.
Through the IA of first-line treatment regimens, a POS of 16.20 m can be determined as the LS standard. Further considering 1ySR and 2ySR, atezolizumab combined with bevacizumab and chemotherapy or pembrolizumab plus chemotherapy are likely to bring the longest LS in the overall population, while single ICI may be adequate for patients with a high PD-L1 expression. ICIs with bevacizumab and chemotherapy may be the best combination for LS for its further advantage over time.