No abstract

Accurate segmentation of prostate and surrounding organs at risk is important for prostate cancer radiotherapy treatment planning. We present a fully automated workflow for male pelvic CT image segmentation using deep learning. The architecture consists of a 2D localization network followed by a 3D segmentation network for volumetric segmentation of prostate, bladder, rectum, and femoral heads. We used a multi-channel 2D U-Net followed by a 3D U-Net with encoding arm modified with aggregated residual networks, known as ResNeXt. The models were trained and tested on a pelvic CT image dataset comprising 136 patients. Test results show that 3D U-Net based segmentation achieves mean (±SD) Dice coefficient values of 90 (±2.0)% ,96 (±3.0)%, 95 (±1.3)%, 95 (±1.5)%, and 84 (±3.7)% for prostate, left femoral head, right femoral head, bladder, and rectum, respectively, using the proposed fully automated segmentation method.

By leveraging tumorgraft (patient-derived xenograft) RNA-sequencing data, we developed an empirical approach, DisHet, to dissect the tumor microenvironment (eTME). We found that 65% of previously defined immune signature genes are not abundantly expressed in renal cell carcinoma (RCC) and identified 610 novel immune/stromal transcripts. Using eTME, genomics, pathology, and medical record data involving >1,000 patients, we established an inflamed pan-RCC subtype (IS) enriched for regulatory T cells, natural killer cells, T1 cells, neutrophils, macrophages, B cells, and CD8 T cells. IS is enriched for aggressive RCCs, including -deficient clear-cell and type 2 papillary tumors. The IS subtype correlated with systemic manifestations of inflammation such as thrombocytosis and anemia, which are enigmatic predictors of poor prognosis. Furthermore, IS was a strong predictor of poor survival. Our analyses suggest that tumor cells drive the stromal immune response. These data provide a missing link between tumor cells, the TME, and systemic factors. We undertook a novel empirical approach to dissect the renal cell carcinoma TME by leveraging tumorgrafts. The dissection and downstream analyses uncovered missing links between tumor cells, the TME, systemic manifestations of inflammation, and poor prognosis. .

Radiation therapy (RT), a major modality for treating localized tumors, can induce tumor regression outside the radiation field through an abscopal effect that is thought to involve the immune system. Our studies were designed to understand the early immunological effects of RT in the tumor microenvironment using several syngeneic mouse tumor models. We observed that RT induced sterile inflammation with a rapid and transient infiltration of CD11b + Gr-1 high+ neutrophils into the tumors. RT-recruited tumor-associated neutrophils (RT-Ns) exhibited an increased production of reactive oxygen species and induced apoptosis of tumor cells. Tumor infiltration of RT-Ns resulted in sterile inflammation and, eventually, the activation of tumor-specific cytotoxic T cells, their recruitment into the tumor site, and tumor regression. Finally, the concurrent administration of granulocyte colony-stimulating factor (G-CSF) enhanced RT-mediated antitumor activity by activating RT-Ns. Our results suggest that the combination of RT and G-CSF should be further evaluated in preclinical and clinical settings.radiation therapy | tumor-associated neutrophils | G-CSF

PURPOSE Radiotherapy (RT) plus long-term androgen suppression (AS) are a standard treatment option for patients with high-risk localized prostate cancer. We hypothesized that docetaxel chemotherapy (CT) could improve overall survival (OS) and clinical outcomes among patients with high-risk prostate cancer. PATIENTS AND METHODS The multicenter randomized NRG Oncology RTOG 0521 study enrolled patients with high-risk nonmetastatic disease between 2005 and 2009. Patients were randomly assigned to receive standard long-term AS plus RT with or without adjuvant CT. RESULTS A total of 612 patients were enrolled; 563 were evaluable. Median prostate-specific antigen was 15.1 ng/mL; 53% had a Gleason score 9 to 10 cancer; 27% had cT3 to cT4 disease. Median follow-up was 5.7 years. Treatment was well tolerated in both arms. Four-year OS rate was 89% (95% CI, 84% to 92%) for AS + RT and 93% (95% CI, 90% to 96%) for AS + RT + CT (hazard ratio [HR], 0.69; 90% CI, 0.49 to 0.97; one-sided P = .034). There were 59 deaths in the AS + RT arm and 43 in the AS + RT + CT arm, with fewer deaths resulting from prostate cancer in the AS + RT + CT arm versus AS + RT (23 v 16 deaths, respectively). Six-year rate of distant metastasis was 14% for AS + RT and 9.1% for AS + RT + CT, (HR, 0.60; 95% CI, 0.37 to 0.99; two-sided P = .044). Six-year disease-free survival rate was 55% for AS + RT and 65% for AS + RT + CT (HR, 0.76; 95% CI, 0.58 to 0.99; two-sided P = .043). CONCLUSION For patients with high-risk nonmetastatic prostate cancer, CT with docetaxel improved OS from 89% to 93% at 4 years, with improved disease-free survival and reduction in the rate of distant metastasis. The trial suggests that docetaxel CT may be an option to be discussed with selected men with high-risk prostate cancer.