984P - Conservative sensitivity analyses of progression-free survival (PFS) of trabectedin plus pegylated liposomal doxorubicin (PLD) vs. PLD alone in pati...

Date 29 September 2012
Event ESMO Congress 2012
Session Poster presentation I
Topics Anticancer agents
Ovarian Cancer
Biological therapy
Presenter Ignacio Romero
Authors I. Romero1, E. Pujade-Lauraine2, A. Tanovic3, J. Gómez García4, A.M. Poveda1
  • 1Area Clinica De Oncologia Ginecologica, Fundación Instituto Valenciano de Oncología, 46009 - Valencia/ES
  • 2Medical Oncology, Hopital Hotel - Dieu, FR-75004 - Paris/FR
  • 3Medical Affairs, PharmaMar, 08028 - Barcelona/ES
  • 4Bio Statistics & Data Management, PharmaMar, 28770 - Colmenar Viejo/ES



Sensitivity analyses are critical to understanding the strength of conclusions made in the primary analysis. A randomised phase III trial OVA-301 compared the efficacy of trabectedin 1.1 mg/m2 in combination with PLD 30 mg/m2 given every 3 weeks vs. PLD 50 mg/m2 every 4 weeks in patients with relapsed ovarian cancer. Primary endpoint was PFS based on independent radiology review (IRR) per RECIST. Secondary PFS analyses were based on independent oncologist (IO) and investigator's assessments (IA).


Patients were assessed symmetrically every 8 weeks in both arms. To avoid bias of the disease assessment time, two conservative sensitivity analyses of PFS were performed. Two different conservative analyses considered the scheduled date of PFS evaluation instead of the actual date of imaging moving the actual date to: #1) the last scheduled date of assessment immediately before the actual evaluation and to #2) the closest scheduled date before or after the actual evaluation.


The treatment effect based on the actual dates of assessment in favour of the combination arm is maintained across the performed sensitivity analyses. Summary of the two sensitivity PFS analyses compared with the pre-planned PFS analysis.

Analysis n Trabectedin + PLD PLD HR CI 95% LR test (p value)
PFS by IR* 645 328 317 0.789 0.646-0.963 0.0190
PFS by IR (#1) 645 328 317 0.813 0.666-0.993 0.0245
PFS by IR (#2) 645 328 317 0.822 0.674-1.004 0.0317
PFS by IO* 671 336 335 0.724 0.599-0.876 0.0008
PFS by IO (#1) 671 336 335 0.749 0.620-0.906 0.0010
PFS by IO (#2) 671 336 335 0.757 0.626-0.916 0.0013
PFS by IA* 672 337 335 0.723 0.608-0.859 0.0002
PFS by IA (#1) 672 337 335 0.742 0.624-0.882 0.0001
PFS by IA (#2) 672 337 335 0.747 0.629-0.888 0.0001

*Main PFS analyses.

#1: Conservative PFS sensitivity analysis; imputation to the previous schedule evaluation.

#2: Sensitivity analysis with imputation to the closest scheduled evaluation.

CI, confidence interval; HR, hazard ratio; IA, investigator assessment; IO, independent oncology review;

IR, independent radiology review; LR, log-rank; PFS, progression-free survival; PLD, pegylated liposomal doxorubicin.


The consistently observed significantly better PFS for the trabectedin combination based on adjusting progression times by moving assessments to scheduled times added methodological strengths and increased reliability and interpretability of the prior findings of this phase III trial.


A. Tanovic: I am an employee at PharmaMar and stock owner. J. Gómez García: I am an employee at PharmaMar and stock owner. A.M. Poveda: I received funding for research from PharmaMar. All other authors have declared no conflicts of interest.