Evaluating pathological complete response as an surrogate endpoint for long-term survival in patients with non-small cell lung cancer: a systematic review and meta-analysis.

PURPOSE

Pathologic complete response (pCR) is deemed to associate with event-free survival (EFS) and overall survival (OS), however, whether it is suitable to serve as a surrogate endpoint for long-term survival in clinical trials of neo-adjuvant treatment for resectable NSCLC trials is still controversy. We aim to evaluate the role of pCR and its viability as a surrogate endpoint for EFS and OS in NSCLC.

METHODS

To investigate the association of pCR and EFS and OS, we performed a meta-analysis involving randomized clinical trials that have reported complete information on pCR rates with hazard ratios (HRs) for EFS and OS. A standard meta-analysis was conducted to determine the relationship between pCR rates and EFS and OS. Additionally, weighted regression analysis was performed to assess the associations between log relative risk (RR) for pCR and log HRs for EFS and OS, with the coefficient of determination (R 2 ) being used to quantify the correlations. Furthermore, the surrogate threshold effect (STE) was also used to evaluate the minimum value of the RR for pCR necessary to confidently predict a non-null effect on HRs for EFS and OS.

RESULTS

The meta-analysis included 14 randomized clinical trials. The high pCR rate group had significant improvement of EFS (HR = 0.69, 95%CI 0.55-0.86) and OS (HR = 0.82, 95%CI 0.72-0.94). A strong association was found between log RR for pCR and log HR for EFS (R 2  = 0.76; 95%CI 0.48-1.00) and a moderate correlation between log RR for pCR and log HR for OS (R 2  = 0.54; 95%CI 0.04-1.00). The STEs for pCR were 4.534 and 10.278 for EFS and OS, respectively. In the subgroup analysis, similar results were only observed in a partial set of comparisons.

CONCLUSIONS

A high pCR rate was associated with a long-term survival outcome. Strong association and moderate association were found between pCR and EFS, pCR and OS, respectively, which supports the application of pCR as a surrogate endpoint for long-term survival in RCTs for resectable NSCLC.