F Petrelli1, S Barni. 1. Oncology Unit, Azienda Ospedaliera di Treviglio, Treviglio, Italy. faupe@libero.it
Abstract
BACKGROUND: Polychemotherapy and biological drugs have increased therapeutic options and outcomes of advanced colorectal cancer (CRC). We examined the relation between progression-free survival (PFS), post-progression survival (PPS) and overall survival (OS) in trials of modern (oxaliplatin- and irinotecan-based) chemotherapy alone or with targeted therapies for advanced CRC. We also evaluated surrogacy of PFS and OS. PATIENTS AND METHODS: A PubMed search identified 34 randomized trials. We split the OS, PFS and PPS and evaluated the correlation between OS and either PFS or PPS. RESULTS: The median PPS and PFS were 10.75 and 8.4 months, respectively. For all trials, PPS was strongly associated with OS [regression coefficient (R2)=0.8; Spearman's rank correlation coefficient (r)=0.88], whereas PFS was moderately associated with OS (R2)=0.43; r=0.64). In trials with targeted therapies, the correlation of PPS with OS was 0.88. However, across all trials, correlation between differences in median PFS (ΔPFS) and median OS (ΔOS) is 0.59 (P=0.0007), confirming PFS/OS surrogacy. CONCLUSION: Our findings indicate that in recent first-line, phase III, trials, OS becomes more associated with PPS than PFS. However, improvements in PFS are strongly associated with improvements in OS. In this setting so, PFS may be an appropriate surrogate for OS.
BACKGROUND: Polychemotherapy and biological drugs have increased therapeutic options and outcomes of advanced colorectal cancer (CRC). We examined the relation between progression-free survival (PFS), post-progression survival (PPS) and overall survival (OS) in trials of modern (oxaliplatin- and irinotecan-based) chemotherapy alone or with targeted therapies for advanced CRC. We also evaluated surrogacy of PFS and OS. PATIENTS AND METHODS: A PubMed search identified 34 randomized trials. We split the OS, PFS and PPS and evaluated the correlation between OS and either PFS or PPS. RESULTS: The median PPS and PFS were 10.75 and 8.4 months, respectively. For all trials, PPS was strongly associated with OS [regression coefficient (R2)=0.8; Spearman's rank correlation coefficient (r)=0.88], whereas PFS was moderately associated with OS (R2)=0.43; r=0.64). In trials with targeted therapies, the correlation of PPS with OS was 0.88. However, across all trials, correlation between differences in median PFS (ΔPFS) and median OS (ΔOS) is 0.59 (P=0.0007), confirming PFS/OS surrogacy. CONCLUSION: Our findings indicate that in recent first-line, phase III, trials, OS becomes more associated with PPS than PFS. However, improvements in PFS are strongly associated with improvements in OS. In this setting so, PFS may be an appropriate surrogate for OS.