Development and internal validation of a clinical risk score to predict pain response after palliative radiotherapy in patients with bone metastases

Publication date: Available online 31 July 2017
Source:International Journal of Radiation Oncology*Biology*Physics
Author(s): Joanne M. van der Velden, Max Peters, Jorrit-Jan Verlaan, Anne L. Versteeg, Liying Zhang, May Tsao, Cyril Danjoux, Elizabeth Barnes, Marco van Vulpen, Edward Chow, Helena M. Verkooijen
PurposePain is a common and debilitating consequence of metastatic bone disease, affecting many patients with end-stage cancer. Radiation treatment for pain control is only effective in about 60% of patients. The aim of this study was to identify patient and tumor characteristics associated with response to radiotherapy, and to develop a risk score for response prediction.Material and MethodsA total of 965 patients with painful bone metastases undergoing palliative radiotherapy at a tertiary referral center between 1999–2007 were identified. Pain scores were measured at 1, 2, and 3 months after radiotherapy. Pain response was defined as at least 2 points decrease on a 0–10 scale in pain score, without increase in analgesics, or an analgesic decrease of at least 25% without an increase in pain score. Thirteen candidate predictors were identified from the literature and expert experience. After multiple imputation, final predictors were selected using stepwise regression and collapsed into a prediction model. Model performance was evaluated by calibration and discrimination and corrected for optimism.ResultsOverall, 462 patients (47.9%) showed a response. Primary tumor site, performance status, and baseline pain score were predictive for pain response with a corrected c-statistic of 0.63. The predicted response rates after radiotherapy increased from 37.5% for patients with the highest score to 79.8% for patients with the lowest score and were in good agreement with the observed response rates.ConclusionsA prediction score for pain response after palliative radiotherapy was developed. The model performance was moderate, showing that prediction of pain response is difficult. New biomarkers and predictors may lead to improved identification of the large group of patients who are unlikely to respond and who may benefit from other or innovative treatment options.

Teaser

Radiotherapy is effective in reducing pain in about 60% of patients with bone metastases. We developed a prediction model to help identify patients who are unlikely to respond to palliative radiotherapy. Primary tumor site, performance status, and baseline pain score were associated with pain response, and were modestly able to discriminate good and poor responders.


http://ift.tt/2veHNqk