Immune checkpoint inhibitors (ICIs) combined with anti-angiogenic therapy have limited efficacy in treating advanced hepatocellular carcinoma (HCC). The synergistic effect from the systemic therapy and radiotherapy (RT) might resolve this problem. We aimed to investigate the effect of RT on the treatment outcomes of ICIs and anti-angiogenic combination therapy in patients with advanced-stage HCC. This retrospective observational study analyzed the medical records of 194 patients with Barcelona Clinic Liver Cancer stage C HCC who were admitted to our center from August 2018 to June 2022 and received ICIs combined with anti-angiogenic therapy as the first-line treatment. Patients who were administered RT for tumor thrombus or symptomatic metastases within 8 weeks of the commencement of combination therapy were allocated to the RT group, while those who did not receive radiotherapy were assigned to the non-radiotherapy (NRT) group. Propensity score matching (PSM) was utilized to mitigate the selection bias. The primary endpoints were progression-free survival (PFS) and overall survival (OS). The secondary endpoints included objective response rate (ORR), disease control rate (DCR), local PFS (LPFS), out-of-field PFS (OutPFS), and treatment-related adverse events (TRAEs). A total of 76 patients diagnosed with advanced-stage HCC and treated with ICIs and anti-angiogenic therapy were included in the study, with 33 patients in the RT group and 43 patients in the non-RT group. After PSM, 29 matched patient pairs were generated. The median follow-up was 15.5 months, and the RT sites were mainly located on the tumor thrombus (55.2%) and extrahepatic metastatic lesions (48.3%). The median PFS was 8.3 months (95% confidence interval (CI): 5.4-11.3) in the RT group and 4.2 months (95% CI: 3.4-5.0) in the NRT group (p<0.001). The median OS was not reached in the RT group and 9.7 months (95% CI: 4.1-15.3) in the NRT group (p=0.002). The ORR was 75.9% (95% CI: 56.5-89.7) in the RT group and 24.1% (95% CI: 10.3-43.5) in the NRT group (p<0.001). The DCR was 100% in the RT group and 75.9% (95% CI: 56.5-89.7) in the NRT group (p=0.005). The median LPFS and OutPFS was 13.2 months (95% CI: 6.3-20.1) and 10.8 months (95% CI: 7.0-14.7), respectively. RT was an independent prognostic factor for PFS (HR=0.33; 95% CI: 0.17-0.64; p <0.001) and OS (HR=0.28; 95% CI: 0.11-0.68; p=0.005), respectively. The rates of any grade TRAEs were similar between the two groups. In comparison to the combination of ICIs and anti-angiogenic therapy, the inclusion of RT has been observed to improve the DCR and survival outcomes in patients with advanced-stage HCC. The safety profile of this triple therapy was satisfactory.

For patients with locally advanced cervical cancer (LACC), precise survival prediction models could guide personalized treatment. We developed and validated CerviPro, a deep learning-based multimodal prognostic model, to predict disease-free survival (DFS) in 1018 patients with LACC receiving definitive radiotherapy. The model integrates pre- and post-treatment CT imaging, handcrafted radiomic features, and clinical variables. CerviPro demonstrated robust predictive performance in the internal validation cohort (C-index 0.81), and external validation cohorts (C-index 0.70&0.66), significantly stratifying patients into distinct high- and low-risk DFS groups. Multimodal feature fusion consistently outperformed models based on single feature categories (clinical data, imaging, or radiomics alone), highlighting the synergistic value of integrating diverse data sources. By integrating multimodal data to predict DFS and recurrence risk, CerviPro provides a clinically valuable prognostic tool for LACC, offering the potential to guide personalized treatment strategies.

Objective: Delineation of clinical target volume (CTV) and organs at risk (OARs) is important for radiotherapy but is time-consuming. We trained and evaluated a U-ResNet model to provide fast and consistent auto-segmentation. Methods: We collected 160 patients’ CT scans with breast cancer who underwent breast-conserving surgery (BCS) and were treated with radiotherapy. CTV and OARs were delineated manually and were used for model training. The dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (95HD) were used to assess the performance of our model. CTV and OARs were randomly selected as ground truth (GT) masks, and artificial intelligence (AI) masks were generated by the proposed model. Two clinicians randomly compared CTV score differences of the contour. The consistency between two clinicians was tested. Time cost for auto-delineation was evaluated. Results: The mean DSC values of the proposed method were 0.94, 0.95, 0.94, 0.96, 0.96 and 0.93 for breast CTV, contralateral breast, heart, right lung, left lung and spinal cord, respectively. The mean 95HD values were 4.31mm, 3.59mm, 4.86mm, 3.18mm, 2.79mm and 4.37mm for the above structures, respectively. The average CTV scores for AI and GT were 2.89 versus 2.92 when evaluated by oncologist A (P=0.612), and 2.75 versus 2.83 by oncologist B (P=0.213), with no statistically significant differences. The consistency between two clinicians was poor (kappa=0.282). The time for auto-segmentation of CTV and OARs was 10.03 s. Conclusion: Our proposed model (U-ResNet) can improve the efficiency and accuracy of delineation compared with U-Net, performing equally well with the segmentation generated by oncologists.