Purpose: Pediatric patients with medulloblastoma in LMICs are most treated with 3D conformal photon craniospinal irradiation (CSI), a time-consuming, complex treatment to plan, especially in resource-constrained settings. Therefore, we developed and tested a 3D conformal CSI autoplanning tool for varying patient lengths. Methods and Materials: Autocontours were generated with a deep learning model trained:tested (80:20 ratio) on 143 pediatric medulloblastoma CT scans (patient ages, 2-19 years, median=7 years). Using the verified autocontours, the autoplanning tool generated 2 lateral brain fields matched to a single spine field, an extended single spine field, or 2 matched spine fields. Additional spine sub-fields were added to optimize the corresponding dose distribution. Feathering was implemented (yielding 9-12 fields) to give a composite plan. Each planning approach was tested on 6 patients (ages, 3-10 years). A pediatric radiation oncologist assessed clinical acceptability of each autoplan. Results: The autocontoured structures’ average Dice similarity coefficient ranged from 0.65-0.98. The average V95 for the brain/spinal canal for single, extended, and multi-field spine configurations was 99.9±0.06%/99.9±0.10%, 99.9±0.07%/99.4±0.30%, and 99.9±0.06%/99.4±0.40%, respectively. The average maximum dose across all field configurations to the brainstem, eyes (L/R), lenses (L/R) and spinal cord were 23.7±0.08 Gy, 24.1±0.28 Gy, 13.3±5.27 Gy, 25.5±0.34 Gy, respectively (prescription=23.4 Gy/13 fractions). Of the 18 plans tested, all were scored as clinically acceptable as-is or clinically acceptable with minor, time-efficient edits preferred or required. No plans were scored as clinically unacceptable. Conclusion: The autoplanning tool successfully generated pediatric CSI plans for varying patient lengths in 3.50 ± 0.4 minutes on average, indicating potential for an efficient planning aid in resource-constrained settings.

Survival of Wilms tumor (WT) is >90% in high-resource settings but <30% in low-resource settings. Adapting a standardized surgical approach to WT is challenging in low-resource settings, but a local control strategy is crucial to improving outcomes. Objective: Provide resource-sensitive recommendations for the surgical management of WT. Methods: We performed a systematic review of PubMed and EMBASE through July 7, 2020, and used the GRADE approach to assess evidence and recommendations. Recommendations: Initiation of treatment should be expedited, and surgery should be done in a high-volume setting. Cross-sectional imaging should be done to optimize preoperative planning. For patients with typical clinical features of WT, biopsy should not be done before chemotherapy, and neoadjuvant chemotherapy should precede surgical resection. Also, resection should include a large transperitoneal laparotomy, adequate lymph node sampling, and documentation of staging findings. For WT with tumor thrombus in the inferior vena cava, neoadjuvant chemotherapy should be given before en bloc resection of the tumor and thrombus and evaluation for viable tumor thrombus. For those with bilateral WT, neoadjuvant chemotherapy should be given for 6–12 weeks. Neither routine use of complex hilar control techniques during nephron-sparing surgery, nor nephron-sparing resection for unilateral WT with a normal contralateral kidney is recommended. When indicated, postoperative radiotherapy should be administered within 14 days of surgery. Post-chemotherapy pulmonary oligometastasis should be resected when feasible, if local protocols allow omission of whole-lung irradiation in patients with non-anaplastic histology stage IV WT with pulmonary metastasis without evidence of extrapulmonary metastasis. Conclusion: We provide evidence-based recommendations for the surgical management of WT, considering the benefits/risks associated with limited-resource settings.