Integración en la rutina diagnóstica de la clasificación molecular de los tumores gliales

Résumé

ABSTRACT Diffuse gliomas are the most common primary brain tumors, they represent 32% of all brain and central nervous system tumors and 81% of all malignant tumors. Of all gliomas, glioblastoma, represents the majority encompassing more than 50 % of all cases, in which 5 year survival rate is less than 5%. This study aimed to analyze the clinical, histological and molecular factors associated with outcome of patients with diffuse glioma, and emphasize the role of biomarkers in current clinical practice and therapeutic decision making, consistent with the 2016 WHO classification of central nervous system tumors. Paraffin-embedded material of 138 diffuse gliomas from 108 patients was analyzed in this retrospective study. In 30 cases, material from both the first operation and recurrence was investigated. The patients underwent tumor resection or biopsy between 1991 and 2014 at our center. Molecular data collected included IDH1 mutation status, 1p/19q chromosome loss, p53 protein expression, loss of ATRX protein expression, EGFR amplification, detection of EGFRvIII, MGMT methylation and proliferative activity (Ki-67). IDH1 mutations occurred in high percentage of grade II and III glioma (87,5 and 56,3%) and were rare in glioblastoma (4,4%), 1p/19q codeletion was present in oligodendrogliomas and 89,5% of codeleted tumors were IDH1 mutant, loss of ATRX and overexpression of p53 were found in astrocytomas (84,2% and 57,1%). Overall survival was longer in patients with IDH1 mutant (p<0,001), 1p19q codeleted (p<0,001) and ATRX loss tumors (p=0,009). EGFR amplification was associated with lower overall survival (p<0,001) and patients with MGMT methylated tumors had better response to adjuvant treatment following tumor resection (p<0,001). On the basis of IDH1 mutation, 1p/19q codeletion, and ATRX expression, we defined 3 different molecular and prognostic groups of gliomas. The group with IDH1 mutant and 1p/19q codeleted exhibited the best prognosis, the second group with IDH1 mutated, loss of ATRX and 1p/19q intact tumors exhibited intermediate prognosis. Finally, the group of tumors characterized by IDH1 wild type, demonstrated the poorest prognosis. Cox regression analysis to estimate relative risk for the molecular and the histological classification showed that the molecular subgrouping of diffuse gliomas allows to stratify patients into prognostically distinct groups better than histological classification. Finally in multivariate analysis, only age, extent of surgical resection, and molecular subgrouping were independent prognostic factors for overall survival. In conclusion, the integrated diagnostic approach allows for much more robust recognition of biologically different subgroups within the spectrum of diffuse gliomas, thereby facilitating tailored treatment for individual patients.