Aspectos clínicos y moleculares de la reintervención en pacientes afectos de glioblastoma

Abstract

Background High-grade gliomas continue to be a poor-prognosis, although with marked differences due to their molecular attributes, such as IDH mutation and the degree of methylation of the MGMT promoter gene. Despite standard treatment, consisting of surgery followed by radiotherapy and temozolomide regimens, recurrence or progression is the norm in almost all patients. Multiple treatment options at glioblastoma progression exist, including reintervention, reirradiation, additional systemic therapy, and novel strategies. To date, no alternative has shown clear superiority of providing greater survival after progression. A second surgery has shown conflicting evidence in the literature regarding its prognostic impact, possibly affected by selection bias, and might benefit a sparse subset of patients with recurrent glioblastoma. On the other hand, variations in the MGMT methylation profile have been described in reinterventions in response to different factors, including exposure to temozolamide. Some authors have tried to quantify these alterations with inconsistent results. The present work aims to determine the prognostic influence of reintervention in a series of patients affected by high-grade glioma and subsequently to analyse changes in the methylation profile of the MGMT promoter that may occur in reoperated patients after exposure to radio and chemotherapy. Research content Methods: Four hundred twenty-seven patients were diagnosed with high-grade glioma between 2005 and 2019 in our institution. In order to avoid inherent selection bias, subjects diagnosed by biopsy, those who did not receive oncological treatment after the first surgery, or those who were not actively treated at progression were excluded from the analysis. In glioblastoma reoperated patients, a retrospective analysis of the degree of methylation in paired samples was performed, using pyrosequencing as the molecular diagnostic technique. Univariate and multivariate Cox regression models were applied, modelling overall survival and post-progression survival, also considering reintervention as a time-dependent covariate. Results: At progression, 33 glioblastoma patients underwent reoperation, and 84 received second-line chemotherapy. Overall survival, but not post-progression survival, was higher in the reoperated group. However, treatment modality did not show significative impact our multivariate Cox regression models considering overall or post-progression survival as the endpoints. Regarding grade III gliomas, a new procedure could provide better survival, although the small sample size limits the results. Analysis of the degree of MGMT promoter methylation was performed in paired samples of 25 glioblastoma reoperated patients. Albeit all patients had received temozolomide, an increase or a decrease in the degree of methylation was observed in 40% of the cases when comparing samples from the second surgery with those from the first surgery. Likewise, considering methylation as a continuous or categorical variable, a correlation between the degree of methylation of the promoter in the first surgery and overall survival was observed. Conclusions Despite selective indications and good technical results in reinterventions in our series, we did not observe a clear prognostic benefit. The association of this procedure with increased survival in some series of the literature could be influenced by patient selection biases and with the fact that reintervention has not been considered as a time-dependent covariate. Variations in the MGMT promoter methylation pattern occurred in some patients and both directions, despite all patients being treated with temozolamide.