Clinical Acceptability of Automatically Generated Elective Lymph Node Volumes for Head and Neck Cancer Patients

PURPOSE/OBJECTIVE(S)

Manual contouring of head and neck lymph node levels is a time-intensive process prone to provider-specific variation. The purpose of this work is to generate a clinical segmentation tool while minimizing the amount of manual effort required by physicians to develop training datasets and review contours. Here we investigate an approach to curate, develop, and clinically validate an auto-contouring model for standard cervical lymph node volumes in the head and neck using a publicly available deep learning architecture. This model updates our previously validated tool to reflect modern practices in lymph node segmentation.

MATERIALS/METHODS

With the assistance of a resident physician, five radiation oncologists manually contoured individual lymph node levels on CT scans for three separate patients treated definitively with radiation or chemoradiation for oropharynx cancer, resulting in 15 unique ground truth cases. These cases were then used to train an nnUnet deep-learning model to generate automated contours for 32 additional cases. These 32 cases were reviewed, manually edited, and used to create the final model. Finally, the model was used to generate contours on the original 15 CT scans (testing cohort), and providers compared these automated contours with the ground-truth (manual) contours. Two blinded studies were performed. In a double-blinded fashion, providers were first asked to select which set of contours they would prefer to use in clinical practice as a starting point for actual cases. Second, they scored each contour on a Likert scale (1-5) to indicate clinical acceptability, ranging from completely unusable to usable without modification.

RESULTS

Across all lymph node levels (IA, IB, II, III, IV, V, RP), average Dice Similarity Coefficient ranged from 0.77 to 0.89 for AI vs manual contours in the testing cohort. These AI and manual lymph node contours were reviewed by 5 physicians each, resulting in 525 preference scores. Across all lymph nodes, the AI contour was superior to or equally preferred to the manual contours at rates ranging from 75% to 91% in the first blinded study. In the second blinded study, physician preference for the manual vs AI contour was statistically different for only the RP contours (p < 0.01). Thus, there was not a significant difference in clinical acceptability for nodal levels I-V for manual versus AI contours. Across all physician-generated contours, 82% were rated as usable with stylistic to no edits, and across all AI-generated contours, 92% were rated as usable with stylistic to no edits.

CONCLUSION

An approach to generate clinically acceptable automated contours for cervical lymph node levels in the head and neck was demonstrated. Furthermore, for nodal levels I-V, there was no significant difference in clinical acceptability in manual vs AI contours. Because we were able to generate and validate a model for each lymph node level individually, the output is applicable to a complete range of disease in which cervical lymph nodes are treated.

The Impact of Travel Time on Trimodal Therapy Compared to Cystectomy for Bladder Cancer

PURPOSE/OBJECTIVE(S)

Neoadjuvant chemotherapy followed by radical cystectomy (RC) and Transurethral Resection of Bladder Tumor (TURBT) followed by concurrent chemoradiation therapy (trimodal therapy, TMT) are both evidence-based approaches to the management of muscle-invasive bladder cancer, and no randomized evidence conclusively supports the use of one treatment over another. Literature suggests that increasing distance to radiation therapy (RT) facilities reduces the likelihood that patients will receive RT, though this has not been evaluated in bladder cancer. We hypothesize that greater travel time between a patient's home and their closest RT facility is associated with a lower likelihood of undergoing TMT.

MATERIALS/METHODS

Clinical data and the locations of patient residences and surgical facilities performing RC were obtained from the California Cancer Registry. RT facility locations were compiled from state regulatory databases and dosimetric accrediting bodies. Travel time was identified using the Google Distance Matrix API. California patients diagnosed with T2-4N0 bladder cancer in 2018-2019 who underwent TMT (> = 55Gy) or RC were included in this analysis. Chi-square tests, Kruskal-Wallis tests, and univariate and multivariate logistic regressions were performed to assess factors associated with an increased likelihood of receipt of TMT, and to compare travel times for patients receiving TMT vs RC.

RESULTS

A total of 536 patients met inclusion criteria, and 27.1% underwent TMT. Patients receiving TMT were more likely to be older, White, male, and live further from their nearest RT facility. After adjusting for age, sex, and race/ethnicity, patients undergoing TMT were more likely than RC patients to live more than 30 minutes away from the nearest RT facility (OR 1.53, p = 0.21, ref<15 minutes). On multivariate analysis, patients receiving TMT were less likely than surgical patients to live 30-60 minutes or >60 minutes from their treatment facility ((OR 0.5, p-value<0.01, and OR 0.23, p-value <.001), ref: <30 minutes). Only 26.2% of patients who received TMT were treated at their nearest RT facility, while 13.3% of patients undergoing RC were treated at their nearest surgical facility.

CONCLUSION

In a modern cohort of bladder cancer patients in California, about a quarter of T2-4N0 patients are undergoing bladder preservation. Patients undergoing TMT lived further from their nearest RT facility than patients undergoing surgery. Surgical patients were more likely to live >30 minutes from their treating facility than TMT patients. A minority of patients underwent treatment at the facility nearest to them, regardless of treatment modality chosen. Taken together, this suggests that proximity to the nearest RT facility may not be associated with a higher likelihood of undergoing TMT. Examination of actual treatment facilities (as opposed to the closest facility) shows that patients are traveling further for surgical care than TMT.

Patient Impressions of Telemedicine in Radiation Oncology

PURPOSE/OBJECTIVE(S)

The increasing use of telemedicine in radiation oncology (RO) since the start of the COVID-19 pandemic highlights the need to understand its role in routine practice. The objective of this study was to assess patient satisfaction with telemedicine in radiation oncology. Secondary objectives included understanding patient motivations for using telemedicine, quantifying the value of telemedicine as an option for care delivery, and characterizing preferences regarding how it should be utilized in their care.

MATERIALS/METHODS

We surveyed patients at our institution about their experience with telemedicine video visits (VVs) in the RO department. Patients were eligible for inclusion if they were English-speaking and participated in a VV with our RO department after January 2020. A convenience sample was obtained, with preference for enrolling patients who completed a VV within 6 months of survey administration (rolling) and for patients treated for a diversity of cancer types. Surveys were administered from August 2021 to October 2022.

RESULTS

A total of 155 patients were approached to complete the survey. Interested patients were emailed a link to a consent form and survey, and 25 patients (16%) completed both. The study closed early due to lack of funding for staff to approach and enroll patients. 96% of patients reported being satisfied with their telemedicine experience in RO and 92% would prefer to have the option of VVs after the resolution of the COVID-19 pandemic. Regarding visit type, 50% of patients preferred in-person visits over VVs for initial consults, while 68% and 77% preferred VVs for weekly on-treatment-visits and follow-up visits, respectively. All patients agreed that their health issues could be adequately addressed with VVs, but only 80% felt that their issues could be addressed with a phone call or email message. The most reported reason for preferring VVs was reduced travel burden. 60% of patients stated they would be more likely to seek cancer treatment at a facility offering VVs and 52% would pay more for a VV than an in-person visit. 92% of patients felt their connection with their physician was as strong as or stronger with VVs than with in-person visits, and 96% trusted their physician to determine when a VV would be medically appropriate.

CONCLUSION

Surveyed patients were satisfied with VVs in RO at our institution, consistent with reports published by our group and others suggesting high rates of satisfaction with VVs by both providers and patients. Our data also suggest that patients find VVs convenient and even prefer them to in-person visits in some cases, without an impact on their perceived ability to have their medical issues addressed or a degradation of their relationship with their physician. The availability of telemedicine options may influence where some patients seek care and their willingness to pay. This study supports the need for ongoing reimbursement for VVs to ensure that patients continue to have access to their cancer care teams.