According to the results of a study published in Radiotherapy and Oncology, a population-based model showed promise in facilitating the selection of patients with locally advanced pancreatic cancer (LAPC) at risk of exposure of healthy organs to stereotactic body radiation therapy (SBRT).
Although the tumors of patients with LAPC have not spread to distant sites, they are classified as unresectable due to local invasion of adjacent structures. Treatment options for these patients include chemotherapy, chemoradiation therapy with standard radiation therapy, and SBRT.
While SBRT more precisely targets tumors in comparison with standard radiation therapies, patient–specific anatomical variations, particularly in gastrointestinal organs, can result in high doses of radiation being delivered to other nearby healthy organs at risk (OAR) when patients are treated with SBRT.
Although the use of personalized planning to account for daily variation in individual OAR anatomy (ie, online adaptive radiotherapy) can reduce the risk of exposing nearby healthy organs to SBRT, it requires the use of additional clinical resources, including other equipment and expertise. The aim of this study was to develop and evaluate a tool to facilitate identification of those patients with locally advanced pancreatic cancer treated with SBRT who would be most likely to benefit from plan adaptation.
This study used a model for the assessment of the motion of OAR that was developed based on 133 computed tomography scans of a population of 35 patients with LAPC treated with SBRT. Specifically, an individualized model was created for each patient through use of a “leave-one-out cross-validation approach” based on imaging data from the remaining 34 patients. Hence, data from other patients with locally advanced pancreatic cancer were used to predict daily motion of OAR in an unknown patient.
A key finding from this study was that compared with clinical observations, the model accurately predicted the risk of exceeding SBRT dose tolerances for the duodenum, stomach, and bowel in 89%, 71%, and 91% of cases, respectively, resulting in an overall accuracy of 83%.
“This is the first study to derive a patient selection tool only based on the planning CT anatomical information, that estimates a prediction by quantifying the sensitivity of planned doses to individual OAR deformations,” the study authors noted.
They further commented that “correct classification of patients based on the planning CT anatomy can help to individualize treatments only when needed, eg by risk-adapted fractionation or employing online adaptation. Besides maximum patient comfort, this could be beneficial to optimize the clinical resources: avoiding the logistical challenges of altering clinical protocols for both patients and staff. An external validation of our tool and a further clinical implementation should be considered as future practical applications of our work.”
Disclosure: This study was funded in part by a research grant of Accuray Inc., Sunnyvale, USA. For a full list of disclosures, please refer to the original study.
Magallon-Baro A, Granton PV, Milder MTW, et al. A model-based patient selection tool to identify who may be at risk of exceeding dose tolerances during pancreatic SBRT [published online October 7, 2019]. Radiother Oncol. doi: 10.1016/j.radonc.2019.09.016