Purpose: Selecting the better techniques to harbor optimal motion management, either a stereotactic linear accelerator delivery using TrueBeam (TBX) or magnetic resonance–guided gated delivery using MRIdian (MRG), is time-consuming and costly. To address this challenge, we aimed to develop a decision-supporting algorithm based on a combination of deep learning-generated dose distributions and clinical data. Materials and Methods: We retrospectively analyzed 65 patients with liver or pancreatic cancer who underwent both TBX and MRG simulations and planning process. We trained three-dimensional U-Net deep learning models to predict dose distributions and generated dose volume histograms (DVHs) for each system. We integrated predicted DVH metrics into a Bayesian network (BN) model incorporating clinical data. Results: The MRG prediction model outperformed the TBX model, demonstrating statistically significant superiorities in predicting normalized dose to the planning target volume (PTV) and liver. We developed a final BN prediction model integrating the predictive DVH metrics with patient factors like age, PTV size, and tumor location. This BN model an area under the receiver operating characteristic curve index of 83.56%. The decision tree derived from the BN model showed that the tumor location (abutting vs. apart of PTV to hollow viscus organs) was the most important factor to determine TBX or MRG. It provided a potential framework for selecting the optimal radiation therapy (RT) system based on individual patient characteristics. Conclusion We demonstrated a decision-supporting algorithm for selecting optimal RT plans in upper gastrointestinal cancers, incorporating both deep learning-based dose prediction and BN-based treatment selection. This approach might streamline the decision-making process, saving resources and improving treatment outcomes for patients undergoing RT.

Purpose: This study aimed to assess prognostic factors associated with combined hepatocellular-cholangiocarcinoma (cHCC-CCA) and to predict 5-year survival based on these factors. Materials and Methods: Patients who underwent definitive hepatectomy from 2006 to 2022 at a single institution was retrospectively analyzed. Inclusion criteria involved a pathologically confirmed diagnosis of cHCC-CCA. Results: A total of 80 patients with diagnosed cHCC-CCA were included in the analysis. The median progression-free survival was 15.6 months, while distant metastasis-free survival (DMFS), hepatic progression-free survival, and overall survival (OS) were 50.8, 21.5, and 85.1 months, respectively. In 52 cases of recurrence, intrahepatic recurrence was the most common initial recurrence (34/52), with distant metastasis in 17 cases. Factors associated with poor DMFS included tumor necrosis, lymphovascular invasion (LVI), perineural invasion, and histologic compact type. Postoperative carbohydrate antigen 19-9, tumor necrosis, LVI, and close/positive margin were associated with poor OS. LVI emerged as a key factor affecting both DMFS and OS, with a 5-year OS of 93.3% for patients without LVI compared to 35.8% with LVI. Based on these factors, a nomogram predicting 3-year and 5-year DMFS and OS was developed, demonstrating high concordance with actual survival in the cohort (Harrell C-index 0.809 for OS, 0.801 for DMFS, respectively). Conclusion The prognosis of cHCC-CCA is notably poor when combined with LVI. Given the significant impact of adverse features, accurate outcome prediction is crucial. Moreover, consideration of adjuvant therapy may be warranted for patients exhibiting poor survival and increased risk of local recurrence or distant metastasis.

Purpose: Selecting the better techniques to harbor optimal motion management, either a stereotactic linear accelerator delivery using TrueBeam (TBX) or magnetic resonance–guided gated delivery using MRIdian (MRG), is time-consuming and costly. To address this challenge, we aimed to develop a decision-supporting algorithm based on a combination of deep learning-generated dose distributions and clinical data. Materials and Methods: We retrospectively analyzed 65 patients with liver or pancreatic cancer who underwent both TBX and MRG simulations and planning process. We trained three-dimensional U-Net deep learning models to predict dose distributions and generated dose volume histograms (DVHs) for each system. We integrated predicted DVH metrics into a Bayesian network (BN) model incorporating clinical data. Results: The MRG prediction model outperformed the TBX model, demonstrating statistically significant superiorities in predicting normalized dose to the planning target volume (PTV) and liver. We developed a final BN prediction model integrating the predictive DVH metrics with patient factors like age, PTV size, and tumor location. This BN model an area under the receiver operating characteristic curve index of 83.56%. The decision tree derived from the BN model showed that the tumor location (abutting vs. apart of PTV to hollow viscus organs) was the most important factor to determine TBX or MRG. It provided a potential framework for selecting the optimal radiation therapy (RT) system based on individual patient characteristics. Conclusion We demonstrated a decision-supporting algorithm for selecting optimal RT plans in upper gastrointestinal cancers, incorporating both deep learning-based dose prediction and BN-based treatment selection. This approach might streamline the decision-making process, saving resources and improving treatment outcomes for patients undergoing RT.

Purpose: This study aimed to assess prognostic factors associated with combined hepatocellular-cholangiocarcinoma (cHCC-CCA) and to predict 5-year survival based on these factors. Materials and Methods: Patients who underwent definitive hepatectomy from 2006 to 2022 at a single institution was retrospectively analyzed. Inclusion criteria involved a pathologically confirmed diagnosis of cHCC-CCA. Results: A total of 80 patients with diagnosed cHCC-CCA were included in the analysis. The median progression-free survival was 15.6 months, while distant metastasis-free survival (DMFS), hepatic progression-free survival, and overall survival (OS) were 50.8, 21.5, and 85.1 months, respectively. In 52 cases of recurrence, intrahepatic recurrence was the most common initial recurrence (34/52), with distant metastasis in 17 cases. Factors associated with poor DMFS included tumor necrosis, lymphovascular invasion (LVI), perineural invasion, and histologic compact type. Postoperative carbohydrate antigen 19-9, tumor necrosis, LVI, and close/positive margin were associated with poor OS. LVI emerged as a key factor affecting both DMFS and OS, with a 5-year OS of 93.3% for patients without LVI compared to 35.8% with LVI. Based on these factors, a nomogram predicting 3-year and 5-year DMFS and OS was developed, demonstrating high concordance with actual survival in the cohort (Harrell C-index 0.809 for OS, 0.801 for DMFS, respectively). Conclusion The prognosis of cHCC-CCA is notably poor when combined with LVI. Given the significant impact of adverse features, accurate outcome prediction is crucial. Moreover, consideration of adjuvant therapy may be warranted for patients exhibiting poor survival and increased risk of local recurrence or distant metastasis.

Purpose: Selecting the better techniques to harbor optimal motion management, either a stereotactic linear accelerator delivery using TrueBeam (TBX) or magnetic resonance–guided gated delivery using MRIdian (MRG), is time-consuming and costly. To address this challenge, we aimed to develop a decision-supporting algorithm based on a combination of deep learning-generated dose distributions and clinical data. Materials and Methods: We retrospectively analyzed 65 patients with liver or pancreatic cancer who underwent both TBX and MRG simulations and planning process. We trained three-dimensional U-Net deep learning models to predict dose distributions and generated dose volume histograms (DVHs) for each system. We integrated predicted DVH metrics into a Bayesian network (BN) model incorporating clinical data. Results: The MRG prediction model outperformed the TBX model, demonstrating statistically significant superiorities in predicting normalized dose to the planning target volume (PTV) and liver. We developed a final BN prediction model integrating the predictive DVH metrics with patient factors like age, PTV size, and tumor location. This BN model an area under the receiver operating characteristic curve index of 83.56%. The decision tree derived from the BN model showed that the tumor location (abutting vs. apart of PTV to hollow viscus organs) was the most important factor to determine TBX or MRG. It provided a potential framework for selecting the optimal radiation therapy (RT) system based on individual patient characteristics. Conclusion We demonstrated a decision-supporting algorithm for selecting optimal RT plans in upper gastrointestinal cancers, incorporating both deep learning-based dose prediction and BN-based treatment selection. This approach might streamline the decision-making process, saving resources and improving treatment outcomes for patients undergoing RT.

Purpose: This study aimed to assess prognostic factors associated with combined hepatocellular-cholangiocarcinoma (cHCC-CCA) and to predict 5-year survival based on these factors. Materials and Methods: Patients who underwent definitive hepatectomy from 2006 to 2022 at a single institution was retrospectively analyzed. Inclusion criteria involved a pathologically confirmed diagnosis of cHCC-CCA. Results: A total of 80 patients with diagnosed cHCC-CCA were included in the analysis. The median progression-free survival was 15.6 months, while distant metastasis-free survival (DMFS), hepatic progression-free survival, and overall survival (OS) were 50.8, 21.5, and 85.1 months, respectively. In 52 cases of recurrence, intrahepatic recurrence was the most common initial recurrence (34/52), with distant metastasis in 17 cases. Factors associated with poor DMFS included tumor necrosis, lymphovascular invasion (LVI), perineural invasion, and histologic compact type. Postoperative carbohydrate antigen 19-9, tumor necrosis, LVI, and close/positive margin were associated with poor OS. LVI emerged as a key factor affecting both DMFS and OS, with a 5-year OS of 93.3% for patients without LVI compared to 35.8% with LVI. Based on these factors, a nomogram predicting 3-year and 5-year DMFS and OS was developed, demonstrating high concordance with actual survival in the cohort (Harrell C-index 0.809 for OS, 0.801 for DMFS, respectively). Conclusion The prognosis of cHCC-CCA is notably poor when combined with LVI. Given the significant impact of adverse features, accurate outcome prediction is crucial. Moreover, consideration of adjuvant therapy may be warranted for patients exhibiting poor survival and increased risk of local recurrence or distant metastasis.

Purpose: Perspectives of radiation oncologists on oligometastatic disease was investigated using multi-layered survey. Materials and Methods: Online survey on the oligometastatic disease was distributed to the board-certified regular members of the Korean Society for Radiation Oncology. The questionnaire consisted of four domains: five questions on demographics; five on the definition of oligometastatic disease; four on the role of local therapy; and three on the oligometastatic disease classification, respectively. Results: A total of 135 radiation oncologists participated in the survey. The median length of practice after board certification was 22.5 years (range, 1 to 44 years), and the vast majority (94.1%) answered affirmatively to the clinical experience in oligometastatic disease management. Nearly two-thirds of the respondents considered the number of involved organs as an independent factor in defining oligometastasis. Most frequently perceived upper limit on the numerical definition of oligometastasis was 5 (64.2%), followed by 3 (26.0%), respectively. Peritoneal and brain metastasis were nominated as the sites to be excluded from oligometastastic disease by 56.3% and 12.6% of the participants, respectively. Vast majority (82.1%) agreed on the role of local treatment in the management of oligometastatic disease. Majority (72%) of the participants acknowledged the European Society for Radiotherapy and Oncology (ESTRO)–European Organisation for Research and Treatment of Cancer (EORTC) classification of oligometastatic disease, however, only 43.3% answered that they applied this classification in their clinical practice. Underlying reasons against the clinical use were ‘too complicated’ (66.0%), followed by ‘insufficient supporting evidence’ (30.0%), respectively. Conclusion While most radiation oncologists supported the role of local therapy in oligometastatic disease, there were several inconsistencies in defining and categorizing oligometastatic disease. Continued education and training on oligometastatic disease would be also required to build consensus among participating caregivers.

Purpose: Risk factors predicting distant metastasis (DM) in extrahepatic bile duct cancer (EHBDC) patients treated with curative resection were investigated. Materials and Methods: Medical records of 1,418 EHBDC patients undergoing curative resection between Jan 2000 and Dec 2015 from 14 institutions were reviewed. After resection, 924 patients (67.6%) were surveilled without adjuvant therapy, 297 (21.7%) were treated with concurrent chemoradiotherapy (CCRT) and 148 (10.8%) with CCRT followed by chemotherapy. To exclude the treatment effect from innate confounders, patients not treated with adjuvant therapy were evaluated. Results: After a median follow-up of 36.7 months (range, 2.7 to 213.2 months), the 5-year distant metastasis-free survival (DMFS) rate was 57.7%. On multivariate analysis, perihilar or diffuse tumor (hazard ratio [HR], 1.391; p=0.004), poorly differentiated histology (HR, 2.014; p < 0.001), presence of perineural invasion (HR, 1.768; p < 0.001), positive nodal metastasis (HR, 2.670; p < 0.001) and preoperative carbohydrate antigen (CA) 19-9 ≥ 37 U/mL (HR, 1.353; p < 0.001) were significantly associated with inferior DMFS. The DMFS rates significantly differed according to the number of these risk factors. For validation, patients who underwent adjuvant therapy were evaluated. In patients with ≥ 3 factors, additional chemotherapy after CCRT resulted in a superior DMFS compared with CCRT alone (5-year rate, 47.6% vs. 27.7%; p=0.001), but the benefit of additional chemotherapy was not observed in patients with 0-2 risk factors. Conclusion Tumor location, histologic differentiation, perineural invasion, lymph node metastasis, and preoperative CA 19-9 level predicted DM risk in resected EHBDC. These risk factors might help identifying a subset of patients who could benefit from additional chemotherapy after resection.

Purpose: There is increasing interest in the efficacy of stereotactic ablative radiotherapy (SABR) for treating colorectal cancer (CRC) patients with oligometastases (OM), recently. The purpose of this meta-analysis was to evaluate local control (LC), progression-free survival (PFS), and overall survival (OS) of CRC patients with pulmonary OM treated with SABR and toxicities. Materials and Methods: Studies that reported SABR for CRC patients with pulmonary OM were searched from MEDLINE and Embase. Treatment outcomes including LC, PFS, OS, and toxicities of grade 3 or higher were assessed. Results: A total of 19 studies with 1,668 patients were chosen for this meta-analysis. Pooled 1-, 2-, and 3-year LC rates were 83.1%, 69.3%, and 63.9%, respectively. PFS rates were 44.8%, 26.5%, and 21.5% at 1, 2, and 3 years, respectively. OS rates at 1-, 2-, and 3-year were 87.5%, 69.9%, and 60.5%, respectively. The toxicity rate of grade 3 or higher was 3.6%. The effect of dose escalation was meta-analyzed using available studies. Conclusion Application of SABR to CRC patients with pulmonary OM achieved modest local control with acceptable toxicity according to the present meta-analysis. Further studies establishing the clinical efficacy of SABR are guaranteed.

Purpose: This study aimed to assess the prognostic significance of bulky nodal involvement in patients with anal squamous cell carcinoma treated with definitive chemoradiotherapy. Materials and Methods: We retrospectively analyzed medical records of patients diagnosed with anal squamous cell carcinoma who underwent definitive chemoradiotherapy at three medical centers between 2004 and 2021. Exclusion criteria included distant metastasis at diagnosis, 2D radiotherapy, and salvage treatment for local relapse. Bulky N+ was defined as nodes with a long diameter of 2 cm or greater. Results: A total of 104 patients were included, comprising 51 with N0, 46 with non-bulky N+, and seven with bulky N+. The median follow-up duration was 54.0 months (range, 6.4 to 162.2 months). Estimated 5-year progression-free survival (PFS), loco-regional recurrence-free survival (LRRFS), and overall survival (OS) rates for patients with bulky N+ were 42.9%, 42.9%, and 47.6%, respectively. Bulky N+ was significantly associated with inferior PFS, LRRFS, and OS compared to patients without or with non-bulky N+, even after multivariate analysis. We proposed a new staging system incorporating bulky N+ as N2 category, with estimated 5-year LRRFS, PFS, and OS rates of 81.1%, 80.6%, and 86.2% for stage I, 67.7%, 60.9%, and 93.3% for stage II, and 42.9%, 42.9%, and 47.6% for stage III disease, enhancing the predictability of prognosis. Conclusion Patients with bulky nodal disease treated with standard chemoradiotherapy experienced poor survival outcomes, indicating the potential necessity for further treatment intensification.