BACKGROUND: Demonstrating improved overall survival (OS) with new multiple myeloma (MM) treatments is becoming difficult because of extended survival, so progression-free survival (PFS) is commonly used as a surrogate endpoint for OS. We evaluated PFS as a potential surrogate for OS by examining whether observed treatment effects on PFS are positively associated with treatment effects on OS in MM. METHODS: A systematic literature review identified 21 randomized control trials reporting hazard ratios (HRs) for treatment effects on PFS and OS. Pearson's r estimated the relationship between HRs (HRPFS and HROS), and between log-transformed HRs (log(HRPFS) and log(HROS)). R(2) values were estimated from linear regression models of the HR and the log(HR) relationships. Sensitivity and subgroup analyses examined the robustness of the HR findings. RESULTS: Positive correlations were found between HRPFS and HROS (r = 0.82; p < 0.0001) and between log(HRPFS) and log(HROS) (r = 0.80; p < 0.0001). Linear regression models produced R(2) values of 0.67 and 0.63 when regressing HROS on HRPFS, and log(HROS) on log(HRPFS), respectively. Sensitivity analyses supported the HR findings. CONCLUSION: This analysis provides evidence for a positive association between treatment effects on PFS and OS. Studies involving patient level data are necessary to confirm whether PFS is a valid surrogate for OS in MM.
BACKGROUND: Demonstrating improved overall survival (OS) with new multiple myeloma (MM) treatments is becoming difficult because of extended survival, so progression-free survival (PFS) is commonly used as a surrogate endpoint for OS. We evaluated PFS as a potential surrogate for OS by examining whether observed treatment effects on PFS are positively associated with treatment effects on OS in MM. METHODS: A systematic literature review identified 21 randomized control trials reporting hazard ratios (HRs) for treatment effects on PFS and OS. Pearson's r estimated the relationship between HRs (HRPFS and HROS), and between log-transformed HRs (log(HRPFS) and log(HROS)). R(2) values were estimated from linear regression models of the HR and the log(HR) relationships. Sensitivity and subgroup analyses examined the robustness of the HR findings. RESULTS: Positive correlations were found between HRPFS and HROS (r = 0.82; p < 0.0001) and between log(HRPFS) and log(HROS) (r = 0.80; p < 0.0001). Linear regression models produced R(2) values of 0.67 and 0.63 when regressing HROS on HRPFS, and log(HROS) on log(HRPFS), respectively. Sensitivity analyses supported the HR findings. CONCLUSION: This analysis provides evidence for a positive association between treatment effects on PFS and OS. Studies involving patient level data are necessary to confirm whether PFS is a valid surrogate for OS in MM.
Authors: Patrick Daniele; Carla Mamolo; Joseph C Cappelleri; Timothy Bell; Alexander Neuhof; Gabriel Tremblay; Mihaela Musat; Anna Forsythe Journal: PLoS One Date: 2022-05-12 Impact factor: 3.752