Objective — to present final overall survival rates. Subject and methods. Since August 2013, the P.A. Herzen Moscow Oncology Research Institute has treated 50 patients with verified high-grade gliomas; grades 3 and 4 glioma was diagnosed in 11 (22%) and 39 (78%) patients, respectively. For each patient, the total isoeffective doses were calculated according to the individual average single dose using the LQED2 formula (H. Withers, 1992) with α/β = 8.5 Gy. Temozolomide is included in a first-line drug treatment program for 45 (90%) patients; the remaining cases used other drugs. Results. A total of 44 (88%) of the 50 patients died; the median survival (MS) was 17.6 (10.9—24.3) months, that in glioblastoma (GBL) and grade 3 gliomas was 17.1 and 46.9 months, respectively (p=0.024). The Cox regression survival model was used to identify two of the most significant factors influencing overall survival rates. These predictors were the onset of a first recurrence in months (Exp(B) 0.892; p=0.000) and bevacizumab (avastin) used as a second-line drug (Exp(B) 0.116; p=0.000). The prescription of avastin as a second line drug is the main predictor of long-term survival. Unifactorial analysis revealed that MS in the avastin group (n=30) was significantly higher: 27.2 versus 10.5 months in the non-bevacizumab group (p=0.000). MS was 23.5 months in patients with GBL receiving an isoeffective dose of 60 Gy and more and only 11.8 months in those having a lower dose of less than 60 Gy (Log Rank 0.049, Breslow 0.004). None of the 21 patients treated with an isoeffective dose of more than 60 Gy survived; therefore the MS of 23.5 months should be recognized to be definitive. In the LQED2 group (receiving less than 60 Gy) two (11.1%) patients were alive as of January 2019. Further significant factors influencing overall survival rates have been identified. In addition to the level of isoeffective dose (p=0.044) and bevacizumab therapy (p=0.000), these parameters included age (<50 years≥; p=0.003), Karnofsky performance status before special treatment (<80%≥; p=0.001) and prior to the start of radiation therapy (<70%≥; p=0.016). A pilot study recorded a total of 5 radionecroses; 4 of them were symptomatic. The median overall survival in patients with radionecrosis was 25.02 (16.3—34.1) months. Conclusion. There is evidence that it is feasible to use more intensive radiation therapy regimens for gliomas with a set of unfavorable prognostic factors. Age older than 50 years and a morphologically more malignant form of a tumor (glioblastoma) are predictors that allow the isoeffective dose level to be increased. At the same time, the patient’s functional status before the start of radiation therapy should not be below 70%.