Developing and validating multi-omics prediction models for late patient-reported dysphagia in head and neck radiotherapy.

Background and purpose. To investigate models developed using radiomic and dosiomic (multi-omics) features from planning and treatment imaging for late patient-reported dysphagia in head and neck radiotherapy.Materials and methods. Training (n = 64) and testing (n = 23) cohorts of head and neck cancer patients treated with curative intent chemo-radiotherapy with a follow-up time greater than 12 months were retrospectively examined. Patients completed the MD Anderson Dysphagia Inventory and a composite score ≤60 was interpreted as patient-reported dysphagia. A chart review collected baseline dysphagia and clinical factors. Multi-omic features were extracted from planning and last synthetic CT images using the pharyngeal constrictor muscle contours as a region of interest. Late patient-reported dysphagia models were developed using a random forest backbone, with feature selection and up-sampling methods to account for the imbalanced data. Models were developed and validated for multi-omic feature combinations for both timepoints.Results. A clinical and radiomic feature model developed using the planning CT achieved good performance (validation: sensitivity = 80 ± 27% / balanced accuracy = 71 ± 23%, testing: sensitivity = 80 ± 10% / balanced accuracy = 73 ± 11%). The synthetic CT models did not show improvement over the plan CT multi-omics models, with poor reliability of the radiomic features on these images. Dosiomic features extracted from the synthetic CT showed promise in predicting late patient-reported dysphagia.Conclusion. Multi-omics models can predict late patient-reported dysphagia in head and neck radiotherapy patients. Synthetic CT dosiomic features show promise in developing successful models to account for changes in delivered dose distribution. Multi-center or prospective studies are required prior to clinical implementation of these models.

Dysphagia in Head and Neck Radiotherapy: The Influence of Pharyngeal Constrictor Anatomy and Dosimetry.

The goal of this study was to identify which anatomical and dosimetric changes correlated with late patient-reported dysphagia throughout the course of head and neck chemo-radiotherapy treatment. The patient cohort (n = 64) considered oropharyngeal and nasopharyngeal patients treated with curative intent, exhibiting no baseline dysphagia with a follow-up time greater than one year. Patients completed the MD Anderson Dysphagia Inventory during a follow-up visit. A composite score was measured ranging from 20 to 100, with a low score indicating a high symptom burden; a score ≤60 indicated patient-reported dysphagia. The pharyngeal (PCM) and cricopharyngeal constrictor muscles (CPM) were contoured on a planning CT image and adapted to weekly cone-beam CT anatomy using deformable image registration and dose was accumulated using weighted dose-volume histogram curves. The PCM and CPM were examined for volume, thickness, and dosimetric changes across treatment with the results correlated to symptom group. Anatomical evaluation indicated the PCM thickness increased more during treatment for patients with dysphagia, with base of C2 vertebrae (p = 0.04) and superior-inferior middle PCM (p = 0.01) thicknesses indicating a 1.0-1.5 mm increase. The planned and delivered mean dose and DVH metrics to PCM and CPM were found to be within random error measured for the dose accumulation, indicating delivered and planned dose are equivalent. The PCM and CPM organs were found to lie approximately 5 mm closer to high dose gradients in patients exhibiting dysphagia. The volume, thickness, and high dose gradient metrics may be useful metrics to identify patients at risk of late patient-reported dysphagia.

Comparing the M.D. Anderson Symptom and Dysphagia Inventories for Head and Neck Cancer Patients.

OBJECTIVES: Where patient-reported outcome measures (PROMs) may be administered at multiple patient visits, it is advantageous to capture these symptoms with as few questions as possible. In this study, the M.D. Anderson Head and Neck Symptom Inventory (MDASI-HN), and the M.D. Anderson Dysphagia Inventory (MDADI) is compared to determine if using the MDASI-HN alone would overlook symptoms identified with MDADI. METHODS: The MDASI-HN and the MDADI were completed by 156 patients, postradiotherapy for head and neck cancer (HNC). Associations between the two instruments were analyzed using correlation analysis, unsupervised machine learning, and sensitivity analysis. RESULTS: Little correlation was found between the two surveys; however, there was overlap between MDASI-HN dry mouth and many MDADI items, confirming that dry mouth is an important factor in difficulty swallowing, and patient QoL. Taking longer to eat (MDADI), was the most commonly reported item overall, with 85 (54%) patients rating it as moderate-severe. Dry mouth was the most endorsed MDASI-HN item (68, 44%). There were 51 patients missed by the MDASI-HN, reporting no moderate-severe symptoms, but reported one or more moderate-severe QoL impacts on MDADI. If patients who reported a score of 2 or higher on the MDASI-HN Dry Mouth item are flagged as requiring follow-up, the number of patients missed by MDASI-HN drops to 15. CONCLUSION: In an HNC clinic where MDASI-HN is routinely administered, assessment of symptoms and QoL might be enhanced by reducing the value at which MDASI Dry Mouth is considered moderate-severe to 2. LEVEL OF EVIDENCE: 3 Laryngoscope, 132:2388-2395, 2022.

An updated approach for deriving PTV margins using image guidance and deformable dose accumulation.

Objective.To demonstrate an updated approach for deriving planning target volume (PTV) margins for a patient population treated with volumetric image-guided radiotherapy.Approach.The approach uses a semi-automated workflow within commercial radiotherapy applications that combines dose accumulation with the bidirectional local distance (BLD) metric. The patient cohort is divided into derivation and validation datasets. For each patient in the derivation dataset, a treatment plan is generated with a 0 mm PTV margin (the idealized treatment scenario without the influence of the standard margin). Deformable image registration enabled dose accumulation of these zero-margin plans. PTV margins are derived by using the BLD to calculate the geometric extent of underdosed regions of the clinical target volume (CTV). The PTV margin is validated by ensuring the specified CTV coverage criterion is met when the margin is applied to the validation dataset.Main results.The methodology was applied to two cohorts: 40 oropharyngeal cancer patients and 50 early-stage breast cancer patients. Ten patients from each cohort were used for validation. PTV margins derived for the oropharyngeal and early-stage breast cancer patient cohorts were 3 and 5 mm, respectively, and ensure that 95% of the prescription dose is delivered to 98% of the CTV for 90% of patients. Dose accumulation showed that the CTV coverage criterion was achieved for at least 90% of patients when the margins were applied.Significance.This methodology can be used to derive appropriate PTV margins for realistic treatment scenarios and any disease site, which will improve our understanding of patient outcomes.

Patient-Reported Outcomes-Guided Adaptive Radiation Therapy for Head and Neck Cancer.

PURPOSE: To identify which patient-reported outcomes (PROs) may be most improved through adaptive radiation therapy (ART) with the goal of reducing toxicity incidence among head and neck cancer patients. METHODS: One hundred fifty-five head and neck cancer patients receiving radical VMAT (chemo)radiotherapy (66-70 Gy in 30-35 fractions) completed the MD Anderson Symptom Inventory, MD Anderson Dysphagia Inventory (MDADI), and Xerostomia Questionnaire while attending routine follow-up clinics between June-October 2019. Hierarchical clustering characterized symptom endorsement. Conventional statistical approaches indicated associations between dose and commonly reported symptoms. These associations, and the potential benefit of interfractional dose corrections, were further explored via logistic regression. RESULTS: Radiotherapy-related symptoms were commonly reported (dry mouth, difficulty swallowing/chewing). Clustering identified three patient subgroups reporting: none/mild symptoms for most items (60.6% of patients); moderate/severe symptoms affecting some aspects of general well-being (32.9%); and moderate/severe symptom reporting for most items (6.5%). Clusters of PRO items broadly consisted of acute toxicities, general well-being, and head and neck-specific symptoms (xerostomia, dysphagia). Dose-PRO relationships were strongest between delivered pharyngeal constrictor Dmean and patient-reported dysphagia, with MDADI composite scores (mean ± SD) of 25.7 ± 18.9 for patients with Dmean <50 Gy vs. 32.4 ± 17.1 with Dmean ≥50 Gy. Based on logistic regression models, during-treatment dose corrections back to planned values may confer ≥5% decrease in the absolute risk of self-reported physical dysphagia symptoms ≥1 year post-treatment in 1.2% of patients, with a ≥5% decrease in relative risk in 23.3% of patients. CONCLUSIONS: Patient-reported dysphagia symptoms are strongly associated with delivered dose to the pharyngeal constrictor. Dysphagia-focused ART may provide the greatest toxicity benefit to head and neck cancer patients, and represent a potential new direction for ART, given that the existing ART literature has focused almost exclusively on xerostomia reduction.

Determining Clinical Patient Selection Guidelines for Head and Neck Adaptive Radiation Therapy Using Random Forest Modelling and a Novel Simplification Heuristic.

PURPOSE: To determine which head and neck adaptive radiotherapy (ART) correction objectives are feasible and to derive efficient ART patient selection guidelines. METHODS: We considered various head and neck ART objectives including independent consideration of dose-sparing of the brainstem/spinal cord, parotid glands, and pharyngeal constrictor, as well as prediction of patient weight loss. Two-hundred head and neck cancer patients were used for model development and an additional 50 for model validation. Patient chart data, pre-treatment images, treatment plans, on-unit patient measurements, and combinations thereof were assessed as potential predictors of each objective. A stepwise approach identified combinations of predictors maximizing the Youden index of random forest (RF) models. A heuristic translated RF results into simple patient selection guidelines which were further refined to balance predictive capability and practical resource costs. Generalizability of the RF models and simplified guidelines to new data was tested using the validation set. RESULTS: Top performing RF models used various categories of predictors, however, final simplified patient selection guidelines only required pre-treatment information for ART predictions, indicating the potential for significant ART process streamlining. The simplified guidelines for each objective predicted which patients would experience increases in dose to: brainstem/spinal cord with sensitivity = 1.0, specificity = 0.66; parotid glands with sensitivity = 0.82, specificity = 0.70; and pharyngeal constrictor with sensitivity = 0.84, specificity = 0.68. Weight loss could be predicted with sensitivity = 0.60 and specificity = 0.55. Furthermore, depending on the ART objective, 28%-58% of patients required replan assessment, less than for previous studies, indicating a step towards more effective patient selection. CONCLUSIONS: The above ART objectives appear to be practically achievable, with patients selected for ART according to simple clinical patient selection guidelines. Explicit ART guidelines are rare in the literature, and our guidelines may aid in balancing the potential clinical gains of ART with high associated resource costs, formalizing ART trials, and ensuring the reproducibility of clinical successes.

Lasso logistic regression to derive workflow-specific algorithm performance requirements as demonstrated for head and neck cancer deformable image registration in adaptive radiation therapy.

As automation in radiation oncology becomes more common, it is important to determine which algorithms are equivalent for a given workflow. Often, algorithm comparisons are performed in isolation; however, clinical context can provide valuable insight into the importance of algorithm features and error magnification in subsequent workflow steps. We propose a strategy for deriving workflow-specific algorithm performance requirements. We considered two independent workflows indicating the need for radiotherapy treatment replanning for 15 head and neck cancer patients (15 planning CTs, 105 on-unit CBCTs). Each workflow was based on a different deformable image registration (DIR) algorithm. Differences in DIR output were assessed using three sets of QA metrics: (1) conventional, (2) workflow-specific, (3) a combination of (1) and (2). For a given set of algorithm metrics, lasso logistic regression modeled the probability of discrepant replan indications. Varying the minimum probability needed to predict a workflow discrepancy produced receiver operating characteristic (ROC) curves. ROC curves were compared using sensitivity, specificity, and the area under the curve (AUC). A heuristic then derived simple algorithm performance requirements. Including workflow-specific QA metrics improved AUC from 0.70 to 0.85, compared to the use of conventional metrics alone. Algorithm performance requirements had high sensitivity of 0.80, beneficial for replan assessments, with specificity of 0.57. This was an improvement over a naïve application of conventional QA criteria, which had sensitivity of 0.57 and specificity of 0.68. In addition, the algorithm performance requirements indicated practical refinements of conventional QA tolerances, indicated where auxiliary workflow processes should be standardized, and may be used to prioritize structures for manual review. Our algorithm performance requirements outperformed current comparison recommendations and provided practical means for ensuring workflow equivalence. This strategy may aid in trial credentialing, algorithm development, and streamlining expert adjustment of workflow output.

Data clustering to select clinically-relevant test cases for algorithm benchmarking and characterization.

Algorithm benchmarking and characterization are an important part of algorithm development and validation prior to clinical implementation. However, benchmarking may be limited to a small collection of test cases due to the resource-intensive nature of establishing 'ground-truth' references. This study proposes a framework for selecting test cases to assess algorithm and workflow equivalence. Effective test case selection may minimize the number of ground-truth comparisons required to establish robust and clinically relevant benchmarking and characterization results. To demonstrate the proposed framework, we clustered differences between two independent workflows estimating during-treatment dose objective violations for 15 head and neck cancer patients (15 planning CTs, 105 on-unit CBCTs). Each workflow used a different deformable image registration algorithm to estimate inter-fractional anatomy and contour changes. The Hopkins statistic tested whether workflow output was inherently clustered and k-medoid clustering formalized cluster assignment. Further statistical analyses verified the relevance of clusters to algorithm output. Data at cluster centers ('medoids') were considered as candidate test cases representative of workflow-relevant algorithm differences. The framework indicated that differences in estimated dose objective violations were naturally grouped (Hopkins = 0.75, providing 90% confidence). K-medoid clustering identified five clusters which stratified workflow differences (MANOVA: p  < 0.001) in estimated parotid gland D50%, spinal cord/brainstem Dmax, and high dose CTV coverage dose violations (Kendall's tau: p  < 0.05). Systematic algorithm differences resulting in workflow discrepancies were: parotid gland volumes (ANOVA: p  < 0.001), external contour deformations (t-test: p  = 0.022), and CTV-to-PTV margins (t-test: 0.009), respectively. Five candidate test cases were verified as representative of the five clusters. The framework successfully clustered workflow outputs and identified five test cases representative of clinically relevant algorithm discrepancies. This approach may improve the allocation of resources during the benchmarking and characterization process and the applicability of results to clinical data.

Radiation Oncology Device Approval in the United States and Canada.

Background Medical devices are a crucial component in the field of radiation oncology. The review and licensing of radiation oncology devices (RODs) is managed on a national basis in Canada by Health Canada and in the United States by the Food and Drug Administration (FDA). The purpose of this study was to examine differences in ROD licensing timelines between Health Canada and the FDA that may impact the ability of Canadians to access the most up-to-date radiation oncology care. Methods A list of ROD was compiled by searching keywords, manufacturers, and proprietary device names in the publicly accessible Canadian Medical Devices Active Licence Listing (MDALL) and the American Establishment Registration & Device Listing and the 510(k) Premarket Notification database. ROD licensing dates were then obtained through both databases. ROD were included if they were licensed in both countries. Results A total of 51 RODs were included in this study and it was found that 71% (36/51) were issued licenses for sale in the United States before Canada, at a mean of 506 days sooner (median [IQR] = 282 [326.5]). No trends in licensing dates were found by stratifying devices by type. Analyses were limited to the date of licensing only, as Health Canada provided no publicly-available information regarding submission milestones such as first submission date for the RODs studied. Conclusions The majority of radiation oncology devices examined were licensed for sale in the USA before Canada. Due to the absence of publicly available information regarding initial ROD application date, we cannot evaluate the impact of the approval process on the overall difference in licensing date. Importantly, this research highlights a lack of publicly-available information from Health Canada regarding the medical device approval process for the radiation oncology devices studied herein.

Framework for the quantitative assessment of adaptive radiation therapy protocols.

BACKGROUND: Adaptive radiation therapy (ART) "flags," such as change in external body contour or relative weight loss, are widely used to identify which head and neck cancer (HNC) patients may benefit from replanned treatment. Despite the popularity of ART, few published quantitative approaches verify the accuracy of replan candidate identification, especially with regards to the simple flagging approaches that are considered current standard of practice. We propose a quantitative evaluation framework, demonstrated through the assessment of a single institution's clinical ART flag: change in body contour exceeding 1.5 cm. METHODS: Ground truth replan criteria were established by surveying HNC radiation oncologists. Patient-specific dose deviations were approximated by using weekly acquired CBCT images to deform copies of the CT simulation, yielding during treatment "synthetic CTs." The original plan reapplied to the synthetic CTs estimated interfractional dose deposition and truth table analysis compared ground truth flagging with the clinical ART metric. This process was demonstrated by assessing flagged fractions for 15 HNC patients whose body contour changed by >1.5 cm at some point in their treatment. RESULTS: Survey results indicated that geometric shifts of high-dose volumes relative to image-guided radiation therapy alignment of bony anatomy were of most interest to HNC physicians. This evaluation framework successfully identified a fundamental discrepancy between the "truth" criteria and the body contour flagging protocol selected to identify changes in central axis dose. The body contour flag had poor sensitivity to survey-derived major violation criteria (0%-28%). The sensitivity of a random sample for comparable violation/flagging frequencies was 27%. CONCLUSIONS: These results indicate that centers should establish ground truth replan criteria to assess current standard of practice ART protocols. In addition, more effective replan flags may be tested and identified according to the proposed framework. Such improvements in ART flagging may contribute to better clinical resource allocation and patient outcome.

Clinical implementation of AXB from AAA for breast: Plan quality and subvolume analysis.

PURPOSE: Two dose calculation algorithms are available in Varian Eclipse software: Anisotropic Analytical Algorithm (AAA) and Acuros External Beam (AXB). Many Varian Eclipse-based centers have access to AXB; however, a thorough understanding of how it will affect plan characteristics and, subsequently, clinical practice is necessary prior to implementation. We characterized the difference in breast plan quality between AXB and AAA for dissemination to clinicians during implementation. METHODS: Locoregional irradiation plans were created with AAA for 30 breast cancer patients with a prescription dose of 50 Gy to the breast and 45 Gy to the regional node, in 25 fractions. The internal mammary chain (IMCCTV ) nodes were covered by 80% of the breast dose. AXB, both dose-to-water and dose-to-medium reporting, was used to recalculate plans while maintaining constant monitor units. Target coverage and organ-at-risk doses were compared between the two algorithms using dose-volume parameters. An analysis to assess location-specific changes was performed by dividing the breast into nine subvolumes in the superior-inferior and left-right directions. RESULTS: There were minimal differences found between the AXB and AAA calculated plans. The median difference between AXB and AAA for breastCTV V95% , was <2.5%. For IMCCTV , the median differences V95% , and V80% were <5% and 0%, respectively; indicating IMCCTV coverage only decreased when marginally covered. Mean superficial dose increased by a median of 3.2 Gy. In the subvolume analysis, the medial subvolumes were "hotter" when recalculated with AXB and the lateral subvolumes "cooler" with AXB; however, all differences were within 2 Gy. CONCLUSION: We observed minimal difference in magnitude and spatial distribution of dose when comparing the two algorithms. The largest observable differences occurred in superficial dose regions. Therefore, clinical implementation of AXB from AAA for breast radiotherapy is not expected to result in changes in clinical practice for prescribing or planning breast radiotherapy.