181P - Validation of modeled early longitudinal CA-125 kinetics for predicting survival in ovarian cancer (OC) phase III trials: Could the failure of addi...

Date 28 September 2014
Event ESMO 2014
Session Poster Display session
Topics Anticancer Agents
Ovarian Cancer
Translational Research
Basic Principles in the Management and Treatment (of cancer)
Biological Therapy
Presenter Melanie Wilbaux
Citation Annals of Oncology (2014) 25 (suppl_4): iv58-iv84. 10.1093/annonc/mdu326
Authors M. Wilbaux1, B. You2, E. Hénin3, A. Hardy-Bessard4, A. Lortholary5, E. Pujade-Lauraine6, G. Freyer2, M. Tod1, A. Reuss7, A. du Bois8, J. Sehouli9, U. Wagner10, J. Pfisterer11
  • 1Emr 3738 - Ciblage Thérapeutique En Oncologie, Faculté de médecine Lyon Sud, Lyon 1, 69921 - Oullins/FR
  • 2Oncologie Médicale, CHU Lyon Sud, 69000 - Lyon/FR
  • 3Emr 3738 - Ciblage Thérapeutique En Oncologie, Faculté de médecine Lyon Sud, Lyon 1, Oullins/FR
  • 4Clinique Armoricaine De Radiologie, Clinique armoricaine de radiologie, Saint-Brieuc/FR
  • 5Medical Oncology, Catherine de Sienne Institute, Nantes/FR
  • 6Cancer De La Femme Et Recherche Clinique, AP-HP, Hôpitaux Universitaires Paris Centre site Hôtel-Dieu, 75004 - Paris/FR
  • 7Coordinating Center For Clinical Trials, Philipps-University Marburg, Marburg/DE
  • 8Department Of Gynecology & Gynecologic Oncology, Kliniken Essen Mitte, Essen/DE
  • 9Gynecology Department, Charite Medical University, Berlin/DE
  • 10Dpt Of Gynecology, Gynecologic Endocrinology And Oncology, University Hospital of Giessen and Marburg, Marburg/DE
  • 11Oncology, Zentrum fur Gynakologische Onkologie, Kiel/DE



Prediction of the survival benefit to expect based on early changes in CA125 titer in ovarian cancer (OC) patients may be helpful for drug development decision making (go/no-go). The mathematical modeling of relationships between CA125 longitudinal kinetics and PFS in recurrent OC patients (reported in Wilbaux et al. Gynecol Oncol 2014) offers this opportunity. The present study was run to assess the external validity of this model of survival benefit prediction in phase III trials of first line advanced OC.


The data from AGO-OVAR 5 and 7 trials investigating the addition of a third drug (epirubicin and topotecan, respectively) to the standard first-line carboplatin-paclitaxel (CP) in advanced OC patients were used to adjust the previously reported parametric survival model. This model links longitudinal modeled CA125 kinetics during the first treatment 6 weeks and PFS outcomes in recurrent OC patients. The predictive performance of the resulting model, adjusted to first line setting, was externally assessed with the data from AGO-OVAR 9, where the addition of gemcitabine to first line CP resulted in a slightly inferior PFS compared to CP alone in the assessable population (400 vs 414 days, relative difference -3.4 %).


The data of 369 advanced OC patients enrolled in AGO-OVAR 5 and 7 trials were successfully used to adjust the model linking longitudinal CA125 kinetics and survival with log-logistic distribution. Based on the longitudinal CA125 changes observed during the first 6 weeks in 435 advanced OC patients enrolled in AGO-OVAR 9 trial, the model could forecast that the combination of gemcitabine to CP would result in a slightly lower PFS in the experimental arm (360 vs 371 days, predicted relative difference -2.9 %).


This is the first externally validated parametric survival model linking longitudinal CA125 kinetics and PFS in advanced OC during first-line chemotherapy. It may be an early predictive tool for drug development go/no-go decisions (A GINEGEPS study).


All authors have declared no conflicts of interest.