Presenter(s)
J. P. Harris1, R. Stitzlein2, M. Golshan Momeni3, M. P. Reilly1, J. Salas1, T. Scharschmidt4, N. Jara Ormazabal5, C. S. Ha6, A. Kalbasi7, and D. J. Konieczkowski8; 1Department of Radiation Oncology, University of California - Irvine, Orange, CA, 2Department of Orthopedic Surgery, University of California - Irvine, Orange, CA, 3Department of Radiology, University of California - Irvine, Orange, CA, 4Department of Orthopedics, Nationwide Children’s Hospital, Colombus, OH, 5Departamento de Medicina Interna Norte, Universidad de Chile, Santiago, NA, Chile, 6University of Texas Health Science Center at San Antonio, San Antonio, TX, 7Department of Radiation Oncology, Stanford University School of Medicine, Stanford, CA, 8Department of Radiation Oncology, James Cancer Hospital/Wexner Medical Center, The Ohio State University, Columbus, OH
Purpose/Objective(s): Radiation combined with surgical resection can result in loss of joint range of motion (ROM). Although contemporary treatment includes joint dose constraints as employed in major clinical trials, there remains a lack of standardized contouring approaches. We sought to develop evidence-based contouring guidelines for major extremity joints and evaluate dose-toxicity relationships.
Materials/Methods: Anatomically-based guidelines were developed for the shoulder (glenohumeral), elbow, hip, and knee joint capsules using anatomic references, with input from board-certified orthopedic surgeons and a musculoskeletal radiologist. Contour guidelines were retrospectively applied to patients with extremity sarcoma treated with combined surgery and radiation at a single academic institution. Dosimetric parameters included maximum dose (D0.03cc), mean dose (DMean), and volumetric doses (D30%, D50%, D60%, D100%). Joint stiffness and arthralgia were graded using CTCAE v6. Hypofractionated doses were converted to 25-fraction equivalents using the linear-quadratic equation with alpha/beta of 3 Gy. Dose-toxicity relationships were evaluated using Wilcoxon rank-sum testing and logistic regression with receiver operator curve analysis. Normal tissue complication probability (NTCP) modeling was performed using logistic regression models with bootstrapping (1000 iterations) to calculate 95% confidence intervals.
Results: Out of 97 consecutively treated patients from 9/2019-6/2025, 61 (62.9%) had planning target volume (PTV) co-planar in axial imaging with 66 joints. Average joint volumes were: shoulder 76 mL (range 57-105 mL), elbow 51 mL (39-71 mL), hip 92 mL (66-130 mL), and knee 305 mL (213-455 mL), with males demonstrating larger volumes than females across all joints. Grade =2 joint stiffness developed in 8 joints (14.8%) including one Grade 3 case, and Grade 2 arthralgia developed in 2 joints (3.7%). Joints with Grade =2 stiffness received significantly higher mean doses (3092 cGy vs 1771 cGy, p=0.01) and D50% doses (3048 cGy vs 1462 cGy, p=0.007) compared to joints without stiffness. The median joint dose (D50%) demonstrated the highest area under the curve for predicting joint stiffness among all evaluated dosimetric parameters (AUC=0.7935). NTCP estimates were computed for 5%, 10%, and 20% joint stiffness rates, with upper bounds on 95% confidence intervals of 2599 cGy, 3416 cGy, and 5470 cGy, respectively.
Conclusion: Standardized, anatomically-based contouring guidelines for major extremity joints were developed and demonstrate clinically relevant dose-toxicity relationships, with joint D50% emerging as the best predictor of toxicity. These guidelines will be prospectively employed in the upcoming phase III SWOG trial S2505 comparing ultra-hypofractionated to conventional pre-operative radiation and can inform treatment planning strategies to minimize joint morbidity in extremity sarcoma patients.