825P - Relationship between progression-free (PFS) and overall survival (OS) in I-line randomized clinical trials (RCT) for advanced renal cell carcinoma (...

Date 29 September 2012
Event ESMO Congress 2012
Session Poster presentation I
Topics Renal Cell Cancer
Presenter Emilio Bria
Authors E. Bria1, F. Massari2, F. Maines2, M. Bonomi3, C. Porta4, S. Bracarda5, F. Cognetti6, D. Giannarelli7, G. Tortora2, M. Milella8
  • 1Medical Oncology, Azienda Ospedaliera Universitaria Integrata Verona-"Borgo Roma", 37134 - Verona/IT
  • 2Oncologia Medica, Azienda Ospedaliera Universitaria Integrata Verona-"Borgo Roma", 37134 - Verona/IT
  • 3Azienda Ospedaliera Universitaria Integrata Verona, IT-37134 - Verona/IT
  • 4Oncologia Medica, IRCCS San Matteo University Hospital Foundation, Pavia/IT
  • 5Department Of Oncology, Ospedale San Donato and U.O.C. of Medical Oncology, 52100 - Arezzo/IT
  • 6Division Medical Oncology A, Istituto Nazionale Tumori Regina Elena, 00144 - Roma/IT
  • 7Medical Oncology, Regina Elena National Cancer Institute, Roma/IT
  • 8Divisione Di Oncologia Medica A, Istituto Nazionale Tumori Regina Elena, Roma/IT



Targeted agents (TA) have become standard 1st-line treatment for advanced RCC based on evidence of PFS advantage over IFN or placebo. Retrospective series suggest that PFS may be considered a reliable intermediate end-point in this setting; however, correlation, surrogacy testing, and validation are required.


RCT evaluating the efficacy of TA as 1st line targeted treatment for advanced RCC were considered eligible. PFS/OS Response/Disease Control rates (ORR, DCR) and Hazard Ratios were extracted from papers/updated presentations. Correlations between 3-, 6-, 9-, and 12-mo PFS and OS rates according to parametric (Pearson's r) and non-parametric (Spearman's Rho and Kendall's Tau) coefficients (with 95% CI) were analyzed to avoid lead-time biases. Regression analysis (parametric R2) and a power-analysis-model in order to determine patients' sample necessary to determine 3%, 5% and 10% OS gain were developed.


Six RCT (4096 patients) were gathered. The best overall correlation between PFS and OS at concurrent timepoints was found at 9 months. With regard to overall rates, 3- and 6-months PFS significantly correlated with 9-mo OS, particularly in the TA arms (Pearson's 0.69/P = n.s. and 0.90/p = 0.01; Spearman's 0.94/p = 0.03 and 0.90/p = 0.04; Tau 0.85/p = 0.02 and 0.82/p = 0.03; respectively). Pearson's coefficients for the correlation between 3-mo PFS and 6-, and 12-mo OS were 0.70 (p = 0.01) and 0.67 (p = 0.01), respectively; the correlation between 6-mo PFS and 12-mo OS were significant as well (Pearson 0.74, p = 0.005; Spearman 0.83, p = 0.005; Tau 0.71, p = 0.001). The regression equation was Y = 0.391861 + 0.4914X [R2 0.44, p(slope) = 0.01]; based on this model, the demonstration of a 3-mo PFS absolute difference of 6%, 10% and 21% (corresponding to a 9-mo OS benefit of 3%, 5% and 10%) would require 2043, 696 and 155 patients, respectively. A significant correlation was found between DCR and OS.


PFS is an acceptable intermediate end-point for OS in the context of 1st line TA treatment of advanced RCC. Individual patient data analysis to verify Prentice's criteria would be required for definitive confirmation.


All authors have declared no conflicts of interest.