Progression-free survival is a surrogate for overall survival in metastatic breast cancer
Progression-free survival is a well established endpoint in assessing clinical efficacy of anti-cancer drugs. However, the surrogacy of progression-free survival for overall survival remains a matter of discussion in the setting of metastatic breast cancer. This question on surrogacy is becoming increasingly pressing given the numerous clinical trials showing improved progression-free survival, but no gain in overall survival. Data from a study assessing the relationship between progression-free and overall survival in the setting of metastatic breast cancer presented at this year's ESMO meeting now supports surrogacy.
In order to investigate the relationship between progression-free and overall survival, a systematic literature search was performed. This literature search was performed using the PICO method, with the population being women with metastatic breast cancer, the interventions and comparators were standard treatments for metastatic breast cancer or best supportive care and the outcomes of interest were median progression-free survival, time to progression and median overall survival. In a second step, a correlation analysis was performed to assess the relationship between median progression-free and overall survival and subgroup analyses were conducted to explore possible reasons for heterogeneity. In a last step, a linear regression analysis was performed to predict the effects of anti-cancer drugs on overall survival, based on its effects on progression-free survival.
In total, 5,041 studies were identified of which 144 fulfilled the eligibility criteria. The selected studies included a total of 315 treatment arms, representing 43,459 patients with metastatic breast cancer. The correlation analysis revealed that a significant relationship exists between median progression-free and overall survival across the included studies (r=0.428; p<0.01). Interestingly this correlation was stronger in studies evaluating chemotherapy alone (r=0.575; p<0.01) or combination (r=0.632; p=0.01), compared to assessing hormone therapy (non= significant correlation).
The correlation coefficient for the treatment effects on progression-free and overall survival was estimated at 0.427 (p<0.01). Results of the regression analysis performed on this large dataset predicts that a difference in median progression-free survival or time to progression of 5, 10, 15 and 20 months would translate into a difference in median overall survival of 8.7, 17.4, 26.2 and 35.0 months respectively.
In summary, this statistical analysis on a large set of metastatic breast cancer patients demonstrates surrogacy of progression-free survival for overall survival. As such, demonstration of improved progression-free survival coupled with a favourable toxicity profile appears to constitute enough evidence for superiority of an investigational treatment.
Reference
Beauchemin C, Cooper D, Lapierre M-E et al. Progression-free survival as a surrogate for overall survival in metastatic breast cancer. Presented at ESMO 2012, Abstract 328PD.
