1404PD - Regorafenib (R) versus placebo (P) in soft tissue sarcomas (STS): analysis of genetic prognostic and predictive factors

Date 10 October 2016
Event ESMO 2016 Congress
Session Sarcoma
Topics Soft Tissue Sarcomas
Presenter Thomas Brodowicz
Citation Annals of Oncology (2016) 27 (6): 483-492. 10.1093/annonc/mdw388
Authors T. Brodowicz1, B. Liegl-Atzwanger2, E. Tresch-Bruneel3, E. Bogart3, O. Mir4, J. Blay5, K. Kashofer2, A. Le Cesne6, R. Hamacher1, N. Penel7
  • 1Medical Oncology, Vienna General Hospital (AKH) - Medizinische Universität Wien, 1090 - Vienna/AT
  • 2Pathology, Medical University Graz, Graz/AT
  • 3Unité De Méthodologie Et Biostatistiques, Centre Oscar Lambret, Lille/FR
  • 4Department Of Medical Oncology, Institut Gustave Roussy, 94805 - Villejuif/FR
  • 5University Claude Bernard Lyon I, Centre Léon Bérard, 69008 - Lyon/FR
  • 6Medical Oncology, Institut de Cancérologie Gustave Roussy, 94805 - Villejuif/FR
  • 7Medical Oncology, Centre Oscar Lambret, Lille/FR



REGOSARC, a double-blind randomized phase II trial compared the activity and safety of R vs P in four STS cohorts: liposarcoma, leiomyosarcoma, synovial sarcoma and other sarcomas. In non-adipocytic sarcomas, PFS/OS were 4.0 [95%-CI: 2.6-5.5]/13.4 months [8.6-17.3] vs 1.0 [1.0-1.8]/9.0 months [6.8-12.5] (HR = 0.36 [0.26-0.53]/ = 0.67 [0.44-1.02] in R and P, respectively. R had no efficacy in the adipocytic cohort. Herein, we report the analysis of potential genetic prognostic and predictive factors of PFS and OS.


Genetic changes were investigated by Ion Torrent Next Generation Sequencing to detect hotspot mutations in 50 genes frequently mutated in cancer (Ion AmpliSeq CancerHotspot Panel v2, CHP2) as well as mutations in the full coding sequence of VEGFR1-3, FGFR1, KIT, PDGFRB, RAF1, RET1, TIE2 and TP53. Prognostic factors (independent to treatment effect) and predictive factors (with positive interaction with treatment effect) have been identified using Kaplan-Meier method and Cox models.


The study population consisted of 134 patients (pts) (71 in R-/63 in P-arm). The most frequent gene alterations were: CHP2 genes alterations (42, 31%), TP53 mutations (35, 26%), PDGFRB mutations (7, 5%), VEGFR1 (5, 4%), VEGFR2 (6, 4 %) and VEGFR3 mutations (5, 4%). Median OS was 12.0 months [7.2-16.6] vs 9.0 [7.5-12.8], in the R and P arm, respectively. Only the KIT mutation was found to be prognostic (HR = 35.9 [4.0-324.0], p = 0.001, only 1 mutated tumor). No gene alteration was found to be predictive for OS. The median PFS was 3.7 months [2.1-5.4] vs 1.3 [1.0-1.8], respectively in the R- and P-arm. One prognostic factor for PFS was detected: FGFR1 mutations (HR = 18.0 [4.0-81.0], p 


Regorafenib is an active drug. None of the tested genes (VEGFR1, VEGFR2, VEGFR3, FGFR1, KIT, PDGFRB, RAF1, RET1, TIE2, TP53 and CHP2 genes) was found to be predictive for PFS or OS. Combinatorial analysis and further subgroup testing is currently ongoing.

Clinical trial identification

EudraCT 2012-005743-24

Legal entity responsible for the study

Sarcoma Platform Austria & French Sarcoma Study Group




T. Brodowicz: Lecture fee: Roche, Amgen, Bayer, Novartis, PharmaMar, Eisai Advisory Board: Amgen, Bayer, Novartis, Eisai. O. Mir: Consultant for: Astra-Zeneca, Amgen, Bayer, BMS, Novartis, Pfizer and Roche. J-Y. Blay: Advisory Board: Roche, Novartis, Bayer, MSD, Lilly, PharmaMar, Deciphera Corporate-sponsored Research: Roche, Novartis, Bayer, MSD, Lilly, PharmaMar. A. Le Cesne: Honoraria: Novartis, PharmaMar, Lilly, Pfizer. N. Penel: Research grant from Bayer HealthCare. All other authors have declared no conflicts of interest.