1416PD - Integrating genotype in risk classification for GIST recurrence. A Spanish Group for Sarcoma Research (GEIS) study

Date 27 September 2014
Event ESMO 2014
Session Sarcoma
Topics GIST
Pathology/Molecular Biology
Translational Research
Presenter Javier Martin Broto
Citation Annals of Oncology (2014) 25 (suppl_4): iv494-iv510. 10.1093/annonc/mdu354
Authors J. Martin Broto1, S. Calabuig2, J. Rubio3, A. Gutierrez4, J. Duran5, F. Garcia5, J. Martinez-Trufero6, J. Maurel7, X. Garcia Del Muro8, J. Cruz9, R. Cubedo10, A.M. Poveda11, C. Valverde Morales12, L.M. De Sande González13, A. De Juan14, J.A. Lopez-Guerrero15
  • 1Medical Oncology, Hospital Universitario Son Espases, 07010 - Palma de Mallorca/ES
  • 2Medical Oncology, Hospital General de Valencia, Valencia/ES
  • 3Medical Oncology, Hospital ICO, Girona/ES
  • 4Hematology, Hospital Universitario Son Espases, Palma de Mallorca/ES
  • 5Medical Oncology, Hospital Universitario Son Espases, Palma de Mallorca/ES
  • 6Medical Oncology, Hospital Miguel Servet, Zaragoza/ES
  • 7Medical Oncology, Hospital Clinic, Barcelona/ES
  • 8Medical Oncology, Hospital ICO, Barcelona/ES
  • 9Medical Oncology, Hospital Universitario de Canarias, Santa Cruz de Tenerife/ES
  • 10Medical Oncology, Hospital Puerta de Hierro, Madrid/ES
  • 11Medical Oncology, Hospital IVO, Valencia/ES
  • 12Oncologia Médica, Vall d'Hebron University HospitalInstitut d'Oncologia, ES-08035 - Barcelona/ES
  • 13Medical Oncology, Complejo Hospitalario de León, León/ES
  • 14Medical Oncology, Hospital Marqués de Valdecilla, Santander/ES
  • 15Molecular Biology, Hospital IVO, Valencia/ES

 

Abstract

Aim

Relevant prognostic factors for relapse free survival (RFS) in GIST relies on clinical and pathologic variables as size, mitoses, location or tumor rupture. However, being GIST a solid tumor model for molecular research and targeted therapies, it seems legitimate to explore the integration of relevant molecular prognostic factors into the risk classification. Several authors have pointed out the detrimental prognostic role of deletion type mutations involving 557/558 (DEL-5758) codons of exon 11 in KIT gene. On the other hand, mutations of PDGFR (MUT-PDGFR) gene have been associated with lower risk of recurrence. We will analyze the influence of these genotype factors for each Miettinen risk category (MRC).

Methods

Clinical data, therapeutic and follow-up procedures stemmed from the GIST Registry of GEIS. Main inclusion criteria were: localized GIST, adequate surgery, size>2 cm, complete genotype for KIT and PDGFR genes, no adjuvant imatinib and minimum follow-up of 3 years. RFS was measured by Kaplan-Meier method. For each MRC, RFS was estimated for those harboring DEL-5758 or MUT-PDGFR. Univariate and multivariate analysis were performed using log-rank test and Cox regression.

Results

429 patients were identified according to the inclusion criteria, 35 of them had insufficient data, thus 394 patients entered into the analysis. Median size was 7 cm and with a median follow-up of 84 months there were 137 recurrences (37%). DEL-5758 and MUT-PDGFR were present in 65 and 25 cases respectively. The 7-year RFS for low MRC were 93% in MUT-PDGFR, 76% in DEL-5758 and 90% in the remainder (p=0.35), for Intermediate MRC were 80%, 26% and 64% (p=0-009) respectively and for high MRC were 40%, 21% and 28% (p=0.79) respectively. On univariate analysis: mitoses, size, location (gastric vs non gastric) and genotype were found to be significantly correlated with RFS. Multivariate analyses revealed that size (HR 1.75; CI 1.04-2.90), location (HR 1.79; CI 1.24-2.60), mitoses (RR 3.4; CI 2.31-5.05) and DEL-5758 (RR 1.5; CI 1.02-2.32) were independent prognostic factors for RFS.

Conclusions

Genotype significantly affects prognosis in localized GIST and therefore should be integrated into the risk classification especially in intermediate MCR.

Disclosure

All authors have declared no conflicts of interest.