Deep Learning-Based Objective and Reproducible Osteosarcoma Chemotherapy Response Assessment and Outcome Prediction.

Osteosarcoma is the most common primary bone cancer, whose standard treatment includes pre-operative chemotherapy followed by resection. Chemotherapy response is used for prognosis and management of patients. Necrosis is routinely assessed after chemotherapy from histology slides on resection specimens, where necrosis ratio is defined as the ratio of necrotic tumor/overall tumor. Patients with necrosis ratio ≥90% are known to have a better outcome. Manual microscopic review of necrosis ratio from multiple glass slides is semiquantitative and can have intraobserver and interobserver variability. In this study, an objective and reproducible deep learning-based approach was proposed to estimate necrosis ratio with outcome prediction from scanned hematoxylin and eosin whole slide images (WSIs). To conduct the study, 103 osteosarcoma cases with 3134 WSIs were collected. Deep Multi-Magnification Network was trained to segment multiple tissue subtypes, including viable tumor and necrotic tumor at a pixel level and to calculate case-level necrosis ratio from multiple WSIs. Necrosis ratio estimated by the segmentation model highly correlates with necrosis ratio from pathology reports manually assessed by experts. Furthermore, patients were successfully stratified to predict overall survival with P = 2.4 × 10-6 and progression-free survival with P = 0.016. This study indicates that deep learning can support pathologists as an objective tool to analyze osteosarcoma from histology for assessing treatment response and predicting patient outcome.

Off-the-shelf EBV-specific T cell immunotherapy for rituximab-refractory EBV-associated lymphoma following transplantation.

BACKGROUNDAdoptive transfer of donor-derived EBV-specific cytotoxic T-lymphocytes (EBV-CTLs) can eradicate EBV-associated lymphomas (EBV-PTLD) after transplantation of hematopoietic cell (HCT) or solid organ (SOT) but is unavailable for most patients.METHODSWe developed a third-party, allogeneic, off-the-shelf bank of 330 GMP-grade EBV-CTL lines from specifically consented healthy HCT donors. We treated 46 recipients of HCT (n = 33) or SOT (n = 13) with established EBV-PTLD, who had failed rituximab therapy, with third-party EBV-CTLs. Treatment cycles consisted of 3 weekly infusions of EBV-CTLs and 3 weeks of observation.RESULTSEBV-CTLs did not induce significant toxicities. One patient developed grade I skin graft-versus-host disease. Complete remission (CR) or sustained partial remission (PR) was achieved in 68% of HCT recipients and 54% of SOT recipients. For patients who achieved CR/PR or stable disease after cycle 1, one year overall survival was 88.9% and 81.8%, respectively. In addition, 3 of 5 recipients with POD after a first cycle who received EBV-CTLs from a different donor achieved CR or durable PR (60%) and survived longer than 1 year. Maximal responses were achieved after a median of 2 cycles.CONCLUSIONThird-party EBV-CTLs of defined HLA restriction provide safe, immediately accessible treatment for EBV-PTLD. Secondary treatment with EBV-CTLs restricted by a different HLA allele (switch therapy) can also induce remissions if initial EBV-CTLs are ineffective. These results suggest a promising potential therapy for patients with rituximab-refractory EBV-associated lymphoma after transplantation.TRIAL REGISTRATIONPhase II protocols (NCT01498484 and NCT00002663) were approved by the Institutional Review Board at Memorial Sloan Kettering Cancer Center, the FDA, and the National Marrow Donor Program.FUNDINGThis work was supported by NIH grants CA23766 and R21CA162002, the Aubrey Fund, the Claire Tow Foundation, the Major Family Foundation, the Max Cure Foundation, the Richard "Rick" J. Eisemann Pediatric Research Fund, the Banbury Foundation, the Edith Robertson Foundation, and the Larry Smead Foundation. Atara Biotherapeutics licensed the bank of third-party EBV-CTLs from Memorial Sloan Kettering Cancer Center in June 2015.