Wuyang Yang

Radiation dose to the floor of mouth muscles predicts swallowing complications following chemoradiation in oropharyngeal squamous cell carcinoma

OBJECTIVES While radiation dose to the larynx and pharyngeal constrictors has been the focus of swallowing complications, the suprahyoid muscles, or floor of mouth (FoM) muscles, are critical for hyoid and laryngeal elevation and effective bolus diversion, preventing penetration and aspiration. We hypothesize that radiation dose to these muscles may be important in the development of dysphagia. MATERIALS AND METHODS We studied 46 patients with OPSCC treated with CRT and who underwent baseline swallowing evaluations and post-treatment videofluoroscopic swallowing studies (VFSS) from 2007 to 2010. Patients with abnormal penetration aspiration scores (PAS>2) served as the study population and patients with normal PAS scores (≤ 2) served as the control cohort. Three suprahyoid muscles and two extrinsic tongue muscles were individually delineated and collectively referred to as the FoM muscles. Radiation dose-volume relationships for these muscles were calculated. Univariate logistic regression analysis was used to determine parameters of significance between patients with normal or abnormal PAS scores. A multivariate regression analysis was subsequently performed to isolate the most statistically critical structures associated with abnormal PAS. RESULTS Univariate analysis resulted in significance/borderline significance of multiple structures associated with abnormal PAS following irradiation. However, when a multivariate model was applied, only the mean dose to the floor of mouth and minimum dose to the geniohyoid were associated with post-radiation abnormal PAS. CONCLUSIONS The dose and volume delivered to the collective FoM muscles may be associated with an increased risk of laryngeal penetration/aspiration to a greater degree than previously recognized organs at risk.

Feasibility of rapid discharge after transoral robotic surgery of the oropharynx

To investigate the use of an algorithm for rapid discharge after transoral robotic surgery (TORS) and its effect on postoperative complications.

Automatic treatment planning implementation using a database of previously treated patients

Purpose: Using a database of prior treated patients, it is possible to predict the dose to critical structures for future patients. Automatic treatment planning speeds the planning process by generating a good initial plan from predicted dose values. Methods: A SQL relational database of previously approved treatment plans is populated via an automated export from Pinnacle3. This script outputs dose and machine information and selected Regions of Interests as well as its associated Dose-Volume Histogram (DVH) and Overlap Volume Histograms (OVHs) with respect to the target structures. Toxicity information is exported from Mosaiq and added to the database for each patient. The SQL query is designed to ask the system for the lowest achievable dose for a specified region of interest (ROI) for each patient with a given volume of that ROI being as close or closer to the target than the current patient. Results: The additional time needed to calculate OVHs is approximately 1.5 minutes for a typical patient. Database lookup of planning objectives takes approximately 4 seconds. The combined additional time is less than that of a typical single plan optimization (2.5 mins). Conclusions: An automatic treatment planning interface has been successfully used by dosimetrists to quickly produce a number of SBRT pancreas treatment plans. The database can be used to compare dose to individual structures with the toxicity experienced and predict toxicities before planning for future patients.

Browser Based Platform in Maintaining Clinical Activities - Use of The iPads in Head and Neck Clinics

Purpose: Incompatibility between documentation and clinical workflow causes physician resistance in organized data collection, which in turn complicates the use of data in patient care improvement. To resolve the gap, we developed an iPad compatible in situ browser-based platform that integrates clinical activity with data collection and analysis presentation. The ability to perform in-clinic activities and monitor decision making using the iPad was evaluated. Methods: A browser-based platform that can exchange and present analysed data from the MOSAIQ database was developed in situ, the iPads were distributed in head and neck clinics to present the browser for clinical activities, data collection and assessment monitoring. Performance of the iPads for in-clinic activities was observed. Results: All in-clinic documentation activities can be performed without workstation computers. Accessing patient record and previous assessments was significantly faster without having to open the MOSAIQ application. Patient assessments can be completed with the physician facing the patient. Graphical presentation of toxicity progression and patient radiation plans to the patient can be performed in single interface without patient leaving the seating area. Updates in patient treatment status and medical history were presented in real time without having to move paper charts around. Conclusions: The iPad can be used in clinical activities independent of computer workstations. Improvements in clinical workflow can be critical in reducing physician resistance in data maintenance. Using the iPad in providing real-time quality monitoring is intuitive to both providers and patients.

WE‐G‐108‐02: Oncospace: A Database Designed for Personalized Medicine in Radiation Oncology

PURPOSE The Oncospace database aggregates treatment planning and clinical information about prior patients facilitating extraction of knowledge from prior courses of care. The goal is to use this knowledge to influence clinical decisions, quality and safety of care for new patients. METHODS The Oncospace website, built in C# and ASP.NET, accesses the MS SQLServer database for analysis. The data tables are designed to support patient geometry, targets and organs at risk (OAR) and their spatial relationships, dose distributions, toxicities, diagnosis and disease progression, chemotherapy and medications, laboratory values, patient histories and demographics. Data is collected directly from the treatment planning system and the oncology information system (OIS). Point of service data collection is facilitated with tablet (iPad) forms linked to the OIS. RESULTS Web pages have been built to answer the following: A) For a selected toxicity and OAR, display the dose volume histogram (DVH) and colorize them by the maximum toxicity grade of the patient while identifying a selected patients DVH B) For a selected OAR and percent volume (%V), find the lowest dose achieved from all patients whose %V is closer to the selected target volume? C) For a given diagnosis, toxicity and treatment, display the aggregate trend in toxicity from start of treatment (acute) through several year follow-up (late)? Oncospace currently supports: and a full H&N database, a multi-institutional pancreatic stereotactic trial, and shape database for automated planning pancreas cancer. CONCLUSION Oncospace can provide fast access to large amounts of data through complex queries designed to answer specific clinical questions to influence the safety and quality of care for new patients. This system can be used in the clinical setting to assess both plan quality and outcome expectations for new patients based on the data of prior patients. Funding for this research was provided by the Commonwealth Foundation, Elekta/IMPAC Medical Systems, Inc., Philips Radiation Oncology Sytems, and the Johns Hopkins University.

SU‐E‐T‐212: Clinical Deployment of an Automatic Planning Interface for Overlap Volume Histogram Based Treatment Planning

PURPOSE To develop and deploy an interface to support automatic treatment planning which predicts achievable dose levels for organs at risk (OARs) from patients with similar or more complicated anatomies queried from a database. This interface will provide an easy to use method of selecting the best known achievable dose values for a given patient, and use them to automate the planning process. METHODS An overlap volume histogram (OVH) describes the distance a target structure can be expanded with the volume of the compared overlap structure. An OVH is generated for each target/critical structure pair and stored in a database with dose-volume histograms (DVHs) for each patient. For all patients, structures are consistently named by mapping ROI names to a set of common names. For a new patient, the patient database is queried for the lowest achievable dose for each OAR from patients in the database with the same or lower overlap distance. The plan parameters and generated objectives are then automatically loaded into treatment planning system for optimization. The final clinical plan from each patient is added to the database to improve the results of future queries. RESULTS The system has been accepted by the dosimetrists for clinical use. Automatically generated plans required less dosimetrist interaction to achieve similar coverage to manually generated plans while OAR doses were reduced or no worse than the manually generated plans. CONCLUSION Automatic planning tools can aid dosimetrists in quickly generating plans which maintain target coverage and produce comparable or reduced dose to OARs. Our interface has simplified the process enabling the broader use of the system across our dosimetry staff. Philips stock ownership Philips Sponsored Research Elekta Sponsored Research Elekta Patent License Accuray (Tomotherapy) Patent License.

MO-A-BRD-09: A Data-Mining Algorithm for Large Scale Analysis of Dose-Outcome Relationships in a Database of Irradiated Head-And-Neck (HN) Cancer Patients

PURPOSE To develop a framework for automatic extraction of clinically meaningful dosimetric-outcome relationships from an in-house, analytic oncology database. METHODS Dose-volume histograms (DVH) and clinical outcome-related structured data elements have been routinely stored to our database for 513 HN cancer patients treated from 2007 to 2014. SQL queries were developed to extract outcomes that had been assessed for at least 100 patients, as well as DVH curves for organs-at-risk (OAR) that were contoured for at least 100 patients. DVH curves for paired OAR (e.g., left and right parotids) were automatically combined and included as additional structures for analysis. For each OAR-outcome combination, DVH dose points, D(Vt ), at a series of normalized volume thresholds, Vt =[0.01,0.99], were stratified into two groups based on outcomes after treatment completion. The probability, P[D(Vt )], of an outcome was modeled at each Vt by logistic regression. Notable combinations, defined as having P[D(Vt )] increase by at least 5% per Gy (p<0.05), were further evaluated for clinical relevance using a custom graphical interface. RESULTS A total of 57 individual and combined structures and 115 outcomes were queried, resulting in over 6,500 combinations for analysis. Of these, 528 combinations met the 5%/Gy requirement, with further manual inspection revealing a number of reasonable models based on either reported literature or proximity between neighboring OAR. The data mining algorithm confirmed the following well-known toxicity/outcome relationships: dysphagia/larynx, voice changes/larynx, esophagitis/esophagus, xerostomia/combined parotids, and mucositis/oral mucosa. Other notable relationships included dysphagia/pharyngeal constrictors, nausea/brainstem, nausea/spinal cord, weight-loss/mandible, and weight-loss/combined parotids. CONCLUSION Our database platform has enabled large-scale analysis of dose-outcome relationships. The current data-mining framework revealed both known and novel dosimetric and clinical relationships, underscoring the potential utility of this analytic approach. Multivariate models may be necessary to further evaluate the complex relationship between neighboring OARs and observed outcomes. This research was supported through collaborations with Elekta, Philips, and Toshiba.

SU‐E‐T‐687: Reverse Auto‐Planning: Assessing Dose Escalation in Pancreatic Stereotactic Radiotherapy From a Shape Database of Prior Patients

PURPOSE Automatic treatment planning can be used to determine achievable dose values before optimization. Reverse auto planning instead looks to find the highest target dose while still meeting critical structure objectives. A tool incorporating reverse auto planning is proposed for prediction of planning difficulty. METHODS A SQL database of 53 pancreas stereotactic body radiotherapy patients is populated with dose and structure information. Overlap volume histograms (OVH) are generated for each organ at risk (OAR) and target pair. For each structure, the lowest achievable target dose, which was calculated by selecting the lowest dose from all patients with a smaller distance to overlap, was queried together with maximum target dose. The lowest target dose from the list of maximum target dose per structure represents the highest achieved dose from the database population. Additional scaling can be used which scales the maximum achievable target dose per structure to the limiting dose of the OAR. RESULTS For patients analyzed using this tool, the predicted maximum target dose serves as a predictor of planning difficulty. If the predicted maximum target dose is high, plans which meet OAR objectives are easier to generate. If the predicted maximum target dose is low, plans are more difficult and planning objectives may not be achievable. CONCLUSION A tool for predicting the maximum achievable target dose is developed. This tool estimates plan difficulty and achievability prior to planning and can be used to determine if dose escalation may be appropriate or if alternate treatment strategies should be used due to difficulty in achieving desired goals. Support by Philips Healthcare: Philips Radiation Oncology Services.

Feasibility and efficiency of automatic treatment planning for stereotactic body radiation therapy (SBRT) in pancreatic cancer patients.

323 Background: Overlap volume histogrmas (OVHs) allow for new plans to be generated based upon prior similar treatment plans. The purpose of this project was to clinically deploy a generic and user-friendly interface for using a database populated with OVHs for planning SBRT pancreas patients. An interface for evaluating adherence to protocol tolerances is also developed to aid in the planning process. Methods: A database of previously treated SBRT pancreas patients is used to query the organ at risk (OAR) dose from patients with similar or harder to plan OVHs. For each OAR in a new plan, the database is queried to find the lowest achievable structure dose from all previous patients with a greater than or equal overlap between the selected structure and the target structure. Queried values are then automatically loaded into the inverse planning optimizer objectives and used to generate an optimized plan. Plans are then evaluated using a protocol interface which queries relevant protocol values and displa...