Treatment of patients with high-risk neuroblastoma (NB) is a complex challenge, and it is based on response to certain elements of therapy. The development and introduction of new treatment approaches, such as GD2-targeted immunotherapy (IT), leads to improved survival in this cohort of patients. The aim of the study was to retrospectively assess the effectiveness of therapy in patients with high-risk NB before the introduction of IT into clinical practice. We retrospectively analyzed the data of 151 NB patients stratified into a high-risk group who had received treatment in accordance with the modified NB2004 protocol of the German Society for Pediatric Oncology and Hematology (GPOH) at the Dmitry Rogachev National Medical Research Center of Pediatric Hematology, Oncology and Immunology from 01.2012 to 12.2017. This study was approved by the Independent Ethics Committee and the Academic Council of the Dmitry Rogachev National Medical Research Center of Pediatric Hematology, Oncology and Immunology. All the study subjects (or their legal representatives) signed a voluntary informed consent form indicating their agreement to treatment and use of their data for research purposes. Overall survival (OS), event-free survival (EFS), and risk factors were analyzed in the patients with high-risk NB including those who had completed multimodal therapy with autologous hematopoietic stem cell transplantation and post-consolidation therapy with isotretinoin and had achieved a satisfactory response to induction therapy (complete response (CR), very good partial response (VGPR), partial response (PR)) (population of special interest). The main unfavorable prognostic clinical and molecular genetic factors affecting survival in the high-risk NB patients were older age, MYCN gene amplification, and stage 4 of the disease. The use of the modified GPOH NB2004 protocol resulted in a satisfactory response (CR/VGPR/PR) to the induction therapy in most patients: 124/151 (82.1 %). Surgery (other than primary tumor biopsy) led to improved survival, with no statistical difference between macroscopic radical surgery and macroscopic residual tumor. At the same time, radiation therapy (RT), as the second element of local control, had a significant impact on EFS in the group of the patients with stage 4 disease: the 3-year EFS was 39.4 % (95 % confidence interval (CI) 23.1–55.4) in the patients with RT versus 25.7 % (95 % CI 17.5–34.7) in the patients without RT (p = 0.0295). The introduction of a new high-dose TreoMel chemotherapy regimen did not result in worse survival rates but led to a decrease in transplant-related toxicity. The 5-year OS and 5-year EFS were 49.4 % (95 % CI 40.9–57.3 %) and 33.3 % (95 % CI 25.9–40.9) respectively for all the study subjects, and 81.6 % (95 % CI 70.3–88.9) and 55.1 % (95 % CI 43.1–65.5) respectively for the patients from the population of special interest. The analysis of the results of therapy in the high-risk NB patients who had received treatment at the Dmitry Rogachev National Medical Research Center of Pediatric Hematology, Oncology and Immunology, yielded results comparable to those of the original GPOH NB2004 protocol. The patients with CR/VGPR/PR to the induction therapy who had completed the protocol treatment with autologous hematopoietic stem cell transplantation and isotretinoin post-consolidation therapy demonstrated higher 5-year EFS rates. However, there remains a need to develop more effective treatment regimens for high-risk NB.