New delineation of semi-quantitative scoring system for 123I-MIBG guided targeted external beam radiation treatment of bone metastases in high-risk neuroblastoma patients.

Objectives: The Curie and SIOPEN scoring systems for semi-quantitative analysis of metastases in neuroblastoma patients are well established. The prognostic value of Curie and SIOPEN scores of greater than 2 is proved to be related to poor outcome for patients. This questions the need for detailed numbering of more than 2 bone metastasis. Herewith, we propose the use of the metastatic site scoring detailed more precisely to the lesion sites on 123I-MIBG scans for planning of radiation treatment in patients with high risk neuroblastoma.

Methods: We conducted a retrospective study of patients treated between 1997 and 2011 in order to review the planar whole body and spot 123I-MIBG scintigraphic images at diagnosis, at the end of induction chemotherapy, after the bone marrow transplant and treatment failure, if such occurred. The new system further subdivided the standard Curie and SIOPEN systems into a more anatomically precise diagram, providing additional regions for the head and neck (total 7) and the proximal and distal parts of upper and lower extremities. It also separated hands and feet from arms and legs. 17 regions on the right and left sides (except vertebra) were assessed separately. Maximal score per region was 3. The regions were co-registered to the external beam targeting sites. Results were assessed for treatment response and failure. The 123I-MIBG images were also scored with the standard Curie and SIOPEN systems for comparison. Univariate cox proportional frailty regression analysis was used to identify covariates which increased the hazard for treatment failure.

Results: 76 patients with 413 metastatic sites were assessed. The MIBG-avid metastatic sites which persisted after the induction and consolidation chemotherapy and hematopoietic stem cell transplant regimen had an increased likelihood of local failure (p<0.05), relative to the lesions that cleared following the induction chemotherapy. The new detailed subdivision system provided better assessment and prediction of response to the radiation therapy compared to the standard Curie and SIOPEN systems (p<0.05). Sites at highest risk for failure included thoracic spine, pelvis, femur, and tibia. The Curie and SIOPEN scores (greater than 2) were confirmed to predict the overall poor outcome for the patients, similarly to the new system. As an additional benefit, an automated algorithm was developed to convert the new scoring system into the standard Curie and SIOPEN scores without a need for the additional manual recalculation.

Conclusion: The newly developed detailed semi-quantitative scoring system of 123I-MIBG scintigraphy provides a better prediction of response to the external beam irradiation than Curie and SIOPEN systems. The new system has a similar prognostic value for assessment of the overall patient outcome in the patients with high risk neuroblastoma. Research Support: N/A