eTable 1. Definitions of Pain Response Based on the International Bone Metastases Consensus Working Party
eTable 2. Patient Demographics and Clinical Characteristics for Patients With Pain Response at Day 42 With QOL Data
eTable 3. Day 10 QOL Scores and Changes From Baseline in Responders and Nonresponders
eTable 4. Day 42 QOL Scores and Score Changes From Baseline in Responders and Nonresponders
eFigure. Change in Mean QOL Scores at Day 42 for Responders vs Nonresponders
Customize your JAMA Network experience by selecting one or more topics from the list below.
McDonald R, Ding K, Brundage M, et al. Effect of Radiotherapy on Painful Bone Metastases: A Secondary Analysis of the NCIC Clinical Trials Group Symptom Control Trial SC.23. JAMA Oncol. 2017;3(7):953–959. doi:10.1001/jamaoncol.2016.6770
How soon after radiotherapy for painful bone metastases can an improvement in quality of life be expected?
In this secondary analysis of 238 patients from the NCIC Clinical Trials Group Symptom Control Trial SC.23, responders to radiotherapy had significantly greater improvement in pain, pain characteristics, functional interference, and psychosocial aspects at day 10 and significant improvement in most domains of the European Organisation for Research and Treatment of Cancer Quality of Life Bone Metastases Module and the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 15 Palliative at day 42 compared with nonresponders.
Quality of life and pain may improve as early as 10 days after radiotherapy; thus, a single 8-Gy dose of radiotherapy for painful bone metastases should be offered to all patients, even those with poor survival.
Many studies that found improved quality of life (QOL) after radiotherapy of bone metastases have small sample sizes and do not use specific questionnaires. How soon after radiotherapy one can expect an improvement in QOL is unknown.
To investigate QOL at days 10 and 42 after radiotherapy with a bone metastases–specific QOL tool.
Design, Setting, and Participants
In this secondary analysis of the NCIC Clinical Trials Group Symptom Control Trial SC.23, a double-blind randomized clinical trial that investigated dexamethasone for the prophylaxis of pain flare after radiotherapy, patients were accrued from 23 Canadian centers from May 30, 2011, to December 11, 2014, and were followed up for 42 days after treatment. Participants referred for radiotherapy for bone metastases were required to have a pain score at the site(s) of treatment of at least 2 (range, 0-10).
Patients were treated with a single 8-Gy radiotherapy dose for 1 or 2 bone metastases.
Main Outcomes and Measures
Patients reported their worst pain score and analgesic intake at baseline and days 10 and 42 after treatment. Pain response was assessed with International Bone Metastases Consensus Endpoint Definitions. Self-reported QOL was completed using the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Bone Metastases Module (QLQ-BM22) and the European Organisation for Research and Treatment of Cancer Quality of Life Core 15 Palliative (QLQ-C15-PAL) at the same time points.
A total of 298 patients were accrued (median age, 68.8 [range, 32-94] years at day 10 and 68.0 [range, 34-90] years at day 42). A total of 122 patients (40.9%) responded to radiotherapy at day 10 and 116 patients (38.9%) at day 42. At day 10, compared with nonresponders, patients with a pain response had a greater reduction in pain (mean reduction, 17.0 vs 1.8; P = .002) and pain characteristics (mean reduction, 12.8 vs 1.1; P = .002), as well as greater improvements in functional interference (mean increase, 11.6 vs 3.6; P = .01) and psychosocial aspects (mean increase, 1.2 points in responders vs mean decrease of 2.2 points in nonresponders, P = .04). Comparing changes in QOL from baseline to day 42, responders had significantly greater improvements in the physical (mean increase, 6.2 vs −9.0; P < .001), emotional (mean increase, 12.3 vs −5.5; P < .001), and global domains (mean increase, 10.3 vs −4.5; P < .001) of the QLQ-C15-PAL compared with nonresponders.
Conclusions and Relevance
Forty percent of patients experienced pain reduction and better QOL at day 10 after radiotherapy with further improvements in QOL at day 42 in responders. A single 8-Gy radiotherapy dose for bone metastases should be offered to all patients, even those with poor survival.
clinicaltrials.gov Identifier: NCT01248585
Sequelae of bone metastases, including pain, hypercalcemia, pathologic fractures, and spinal cord compression, can cause significant morbidity and decreased quality of life (QOL).1,2 Radiotherapy is a key treatment modality for symptomatic uncomplicated bone metastases. Studies3-6 have demonstrated its efficacy in relieving pain, with approximately 60% to 70% of patients experiencing benefit and one-quarter to one-third of patients achieving a complete response (CR), independent of whether single or multiple fractions are prescribed.
Although reduction of pain is an important end point, palliative-intent treatment should also focus on maintenance and improvement of QOL.7 Understanding QOL has historically been difficult because of its many dimensions and potential confounding factors. More recently, tools have been designed to measure the QOL of patients with palliative cancer and, even more specifically, those with bone metastases. These tools include the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 15 Palliative (QLQ-C15-PAL)8 and the EORTC Quality of Life Questionnaire Bone Metastases Module (QLQ-BM22).9
Because increased pain is associated with a lower QOL in patients with symptomatic bone metastases,10 a reduction in pain after radiotherapy is hypothesized to translate to an improvement in QOL. A few studies11-13 conducted to date have demonstrated this; however, many involve small patient populations and do not use the bone metastases–specific QOL measurement tools currently available. Furthermore, the time points at which QOL have historically been evaluated are not the same among all studies. Previous studies have found that pain response may occur 1 week after radiotherapy14 and that QOL improves between 1 week and 1 month.13,15 However, these studies13-15 accounted for analgesics differently than what the current international consensus end point definitions suggest.16 Similarly, they did not use QOL instruments that were validated specifically for patients with advanced cancer and, in particular, for those with bone metastases. The objective of our study was to investigate pain and QOL at days 10 and 42 after radiotherapy using the QLQ-C15-PAL and QLQ-BM22 as measurement tools. Specifically, we aimed to determine whether there are differences in QOL between responders and nonresponders as determined by the International Bone Metastases Consensus Endpoint Definitions.16
This is a secondary analysis of the NCIC Clinical Trials Group Symptom Control Trial SC.23 (NCIC-CTG-SC.23), which was an intent-to-treat, randomized, double-blind, phase 3 trial that investigated the use of dexamethasone for prophylaxis of pain flare after radiotherapy for painful bone metastases.17 Patients were accrued from May 30, 2011, to December 11, 2014, and were followed up for 42 days after treatment. All participating centers received approval from their local research ethics boards (Research Ethics Board of the Vitalité Health Network, Comité d'éthique de la recherche en santé chez l'humain du CHUS, Comité d'éthique de la recherche du CHU de Québec, Hôpital Maisonneuve-Rosemont Centre affilié a l'Université de Montréal Comité d'éthique de la Recherche, Comité d'éthique de la Recherche du CHUM, Ontario Cancer Research Ethics Board, The University of Manitoba Biomedical Research Ethics Board, University of Saskatchewan Biomedical Research Ethics Board, Health Research Ethics Board of Alberta–Cancer Committee [formerly Alberta Cancer Research Ethics Committee], and University of British Columbia–British Columbia Cancer Agency Research Ethics Board). Written informed consent was obtained from all participants, and all data were deidentified.
All patients were prescribed single 8-Gy (to convert to rads, multiply by 100) radiotherapy to 1 or 2 painful bone metastases. Patients were randomized to receive two 4-mg tablets of dexamethasone or 2 placebo tablets at least 1 hour before radiotherapy, then every day for 4 days after radiotherapy. To be eligible for participation, patients were required to have a worst pain score of at least 2 on a scale of 0 (no pain) to 10 (worst possible pain). Patients reported their worst pain score and daily opioid analgesic intake at baseline, days 1 to 10 after treatment, and day 42 after treatment. Patient-reported QOL was evaluated using the QLQ-C15-PAL and QLQ-BM22 at baseline and days 10 and 42 after treatment.
Shortened from the parent EORTC QOL Questionnaire Core 30, the QLQ-C15-PAL contains 15 questions that assess QOL in patients with palliative cancer.8 Five symptoms (nausea and vomiting, dyspnea, insomnia, appetite loss, and constipation) are assessed as single-item symptom scales, whereas 2 symptoms (fatigue and pain) are assessed on multi-item symptom scales. The questionnaire also consists of 2 multi-item functional scales (physical and emotional functioning) and 1 question that pertains to overall QOL. Each question is rated on a Likert scale ranging from 1 (not at all) to 4 (very much), except for overall QOL, which is assessed on a scale of 1 (very poor) to 7 (excellent). All scores were linearly transformed to a scale ranging from 0 to 100. Higher scores for the multi- and single-item symptom scales indicate higher symptom burden, whereas higher scores for the multi-item functional scales and overall QOL indicate better functioning and QOL.
The QLQ-BM22 is a module designed specifically for patients with bone metastases.9 It consists of 22 questions divided into 4 subscales: painful sites, pain characteristics, functional interference, and psychosocial aspects. Similar to the QLQ-C15-PAL, all items in the QLQ-BM22 are rated on a 4-point Likert scale ranging from 1 (not at all) to 4 (very much). All items were linearly transformed to a scale ranging from 0 to 100. For painful sites and pain characteristics, higher scores signify greater symptoms. Conversely, greater scores indicate better functioning for functional interference and psychosocial aspects.
Changes in the QOL scores reported using the QLQ-C15-PAL and QLQ-BM22 were interpreted according to the guidelines suggested by Osoba et al.18 Changes in QOL scores were calculated after all items were transformed to a linear scale from 0 to 100. Clinically meaningful changes were those that differed by 10 or more points from the baseline. For the symptom domains of the QLQ-C15-PAL and QLQ-BM22 (painful sites and pain characteristics), a decrease of 10 or more points indicates less symptom burden, whereas an increase of 10 or more points indicates greater symptom burden. On the other hand, for the functional interference domains of the QLQ-C15-PAL (functional scales and QOL) and QLQ-BM22 (functional interference and psychosocial aspects), an increase of 10 or more points indicates improvement, whereas a decrease of 10 or more points indicates deterioration. Pain response was assessed according to the International Bone Metastases Consensus Endpoint Definitions16 (eTable 1 in the Supplement). Patients with CR and partial response (PR) were considered responders, whereas those with pain progression or a stable pain were considered nonresponders.16 Overall response was the sum of CR and PR.
Descriptive statistics were used to summarize baseline demographic information. Differences in demographics between responders and nonresponders at day 42 were compared using the χ2 test for categorical variables and 2-sample t test for continuous variables. The Wilcoxon rank sum test was used to test for differences in baseline, day 10, and day 42 QOL scores between responders and nonresponders. Similarly, the Wilcoxon rank sum test was used to test the differences between responders and nonresponders in terms of the changes of QOL scores between (1) baseline and day 10 and (2) baseline and day 42. Differences between the patients with and without clinically meaningful changes in QOL were compared using the χ2 test. A 2-sided P < .05 was considered statistically significant. All analyses were conducted using SAS software, version 9.2 (SAS Institute Inc).
A total of 298 patients were enrolled in the study (median age, 68.8 [range, 32-94] years at day 10 and 68.0 [range, 34-90] years at day 42). At day 10, 122 patients (40.9%) responded to radiotherapy (37 CRs and 85 PRs), and at day 42, 116 patients (38.9%) were responders (61 CRs and 55 PRs). Of the 298 patients, 215 (72.1%) had complete day 10 QOL data, and 187 (62.7%) had complete day 42 QOL data. For the 187 patients with pain response at day 42 with QOL data, prostate (54 [28.9%]), breast (51 [27.3%]), and lung (43 [23.0%]) were the 3 most common primary sites of cancer. Baseline worst pain scores ranged from 2 to 10, with 86 patients (46.0%) reporting a pain score of 7 to 10 at first site.
Of 187 patients with evaluable QOL data, 102 (54.5%) experienced a pain response at day 42. There were no statistically significant differences in baseline demographics (Table 1) or QOL scores between responders and nonresponders at day 42 except for primary cancer site, with patients with prostate cancer being more likely to respond to radiotherapy (40 [39.2%] of responders vs 14 [16.5%] of nonresponders, P = .007). Between included and excluded patients with no day 42 QOL data, there were no statistically significant differences in most baseline demographics except for primary cancer (54 included patients with prostate cancer [28.9%], 51 with breast cancer [27.3%], 43 with lung cancer [23.0%], and 39 with other cancer [20.9%] vs 20 excluded patients with prostate cancer [18.0%], 15 with breast cancer [13.5%], 41 with lung cancer [36.9%], and 35 with other cancer [31.5%]; P < .001) and Karnofsky Performance Status (scores of 40-60, 27 [14.4%] included patients vs 41 [36.9%] excluded patients; scores of 70-80, 106 [56.7%] included patients vs 59 [53.2%] excluded patients; scores of 90-100, 54 [28.9%] included patients vs 11 [9.9%] excluded patients; P < .001) (eTable 2 in the Supplement).
Comparing patients who had experienced a pain response on day 10 with those who did not, there was no statistically significant difference in baseline QOL scores. Patients with response reported better physical functioning (mean [SD] score, 76.1 [24.8] vs 69.1 [26.3]; P = .03), less pain (mean [SD] score, 43.9 [25.4] vs 61.0 [27.9]; P < .001), and improved constipation (mean [SD] score, 25 [29.3] vs 35.9 [34.2]; P = .02) compared with nonresponders.
On the QLQ-BM22, responders also reported fewer painful sites (mean [SD] score, 28.5 [16.5] vs 34.2 [18.5]; P = .03) and pain characteristics (mean [SD] score, 31.6 [20.6] vs 44.6 [24.4]; P < .001) and improved functional interference (mean [SD] score, 63.5 [20.6] vs 54.1 [23.9]; P = .007) compared with nonresponders. The greatest difference observed was in the pain item of the QLQ-C15-PAL, where responder-reported scores were a mean of 17.1 points lower than scores reported by nonresponders (mean scores, 43.9 in responders vs 61.0 in nonresponders; P < .001) (eTable 3 in the Supplement).
When observing the change in QOL scores from baseline to day 10, patients with a pain response had a significantly greater reduction in pain (mean reduction, 17.0 vs 1.8; P = .002) and pain characteristics (mean reduction, 12.8 vs 1.1; P = .002). This group also experienced greater improvements in functional interference (mean increase, 11.6 vs 3.6; P = .01) compared with nonresponders. A significant difference was observed in change in the psychosocial aspects, where responders improved by a mean of 1.2 points and nonresponders declined by a mean of 2.2 points (P = .04) (eTable 3 in the Supplement).
When patients were classified by clinically meaningful changes in QOL (change ≥10 points), a greater portion of responders also had a QOL improvement in pain characteristics (64.0% vs 42.6%, P = .002) and psychosocial aspects (37.0% vs 20.9%, P = .009) when compared with nonresponders (Table 2).
Responders and nonresponders as determined on day 42 did not have statistically significant differences in baseline QOL scores on the QLQ-C15-PAL or the QLQ-BM22. Responders reported better QOL scores in every item of the QLQ-C15-PAL except for dyspnea (mean [SD] score, 18.6 [20.8] vs 28.2 [30.6]; P = .06) and insomnia (mean [SD] score, 24.5 [25.7] vs 33.3 [32.3]; P = .09). In all domains of the QLQ-C15-PAL (physical, emotional, and global), mean scores reported by responders were at least 10 points greater than those reported by nonresponders, representing significant differences. The greatest difference between responders and nonresponders was their pain score (29.9 for responders and 52.0 for nonresponders, P < .001). Similarly, responders reported better QOL in all domains of the QLQ-BM22, except the psychosocial aspects (mean [SD] score, 56.8 [18.6] vs 50.3 [20.0]; P = .07) (eTable 4 in the Supplement).
Comparing changes in QOL from baseline to day 42, responders had significantly greater improvements in the physical (mean increase, 6.2; P < .001), emotional (mean increase, 12.3; P < .001), and global domains (mean increase, 10.3; P < .001) of the QLQ-C15-PAL compared with nonresponders. Responders also experienced greater reductions in pain (mean reduction, 30.1 vs 10.4; P < .001), fatigue (mean reduction, 7.5 vs 9.5 increase; P < .001), appetite (mean reduction, 4.6 vs 5.9 increase; P = .03), and constipation (mean reduction, 8.5 vs 4.4 increase; P = .006) on the QLQ-C15-PAL and greater improvements in painful sites (mean reduction, 13.5 vs 3.2; P < .001), pain characteristics (mean reduction, 21.4 vs 10.6; P = .01), functional interference (mean increase, 18.8 vs 9.2; P = .001), and psychosocial aspects (mean increase, 6.2 vs −0.8; P = .01) (eTable 4 in the Supplement) on the QLQ-BM22. The eFigure in the Supplement includes an illustration of the changes in scores from baseline to day 42 in responders and nonresponders.
When comparing the proportion of responders vs nonresponders who had clinically meaningful changes in QOL (QOL change of ≤10 or ≥10), responders were better in the physical (41 [40.6%] vs 19 [22.4%], P = .008), emotional (51 [50.0%] vs 26 [30.6%], P = .007), and global (56 [56.0%] vs 22 [25.9%], P < .001) domains and pain (71 [71.7%] vs 43 [50.6%], P = .003), fatigue (43 [42.2%] vs 17 [20.2%], P = .002), and constipation (34 [33.3%] vs 16 [19.3%], P = .03) items of the QLQ-C15-PAL. Responders had a significantly better response in all items of the QLQ-BM22 (painful sites, 59 [57.8%] vs 31 [36.5%], P = .004; pain characteristics, 74 [72.5%] vs 44 [51.8%], P = .003; functional interference, 66 [64.7%] vs 38 [44.7%], P = .006; psychosocial aspects, 43 [42.2%] vs 23 [27.1%], P = .03) (Table 3).
The international consensus end points take into account the pain score and analgesic consumption,16 which lowers the CR and PR rates when compared with the pain-only end points.19 Despite that, by intent-to-treat analysis, 40% of patients experienced pain reduction at day 10 after radiotherapy using the international consensus end points in our study. To our knowledge, this study is the first to report significant improvements in certain domains of QOL in those who responded to radiotherapy as early as 10 days after treatment for painful bone metastases using the international consensus end points and QLQ-BM22. At day 10, responders had significantly greater reductions in pain and pain characteristics and better improvements in functional interference and psychosocial aspects compared with nonresponders.
Zeng et al12 described 79 patients receiving radiotherapy for painful bone metastases. Patients were asked to complete the QLQ-BM22 at baseline and month 1 after treatment. The authors found that although responders to radiotherapy had a trend toward improved psychosocial aspects at month 1, such improvements were not statistically significant. However, their study is limited by a small sample size, and the overall response rate was only 35%, whereas it was 55% in the current study. It is likely that the greater proportion of responders in our study is responsible for the significant results. Our response rates are more comparable to those cited in the literature.
Caissie et al11 investigated 178 patients who completed the QLQ-C15-PAL at baseline, weeks 1 and 2 after the initiation of radiotherapy, and months 1 and 2 after the initiation of radiotherapy. At day 7, improvements were only seen in fatigue and pain symptoms for responders; no improvements were found in nonresponders. Our study similarly found a significant difference between responders and nonresponders for pain at day 10, such that responders had reduced pain, whereas nonresponders did not. Differences in fatigue were not observed between responders and nonresponders in our study; however, their study did not use the EORTC bone-specific module, the QLQ-BM22, and patient questionnaire completion rates were low (38% at week 1, 37% at week 2, 68% at month 1, and 40% at month 2).11
The study with the largest patient population exploring QOL changes after radiotherapy to painful bone metastases was a recently published secondary analysis of the Dutch Bone Metastasis Study.13 Patients were asked to complete several questionnaires, including a pain scale, medication list, the Rotterdam Symptom Checklist, and a visual analog scale for general health weekly for 12 weeks and monthly thereafter until 2 years, death, or study closure. A total of 956 patients accrued between 1996 and 1998 were eligible for analysis. Responders and nonresponders were observed to have significant differences in QOL at week 12 after radiotherapy: psychological distress, physical symptom distress, visual analog scale for general health score, and overall evaluation of life score improved in responders and deteriorated in nonresponders. In addition, activity-level impairment deteriorated in nonresponders but remained stable in responders.13 The authors report that these differences were minimal immediately after radiotherapy and increased with time until 12 weeks after treatment, when significance was evaluated. Although their study was robust, our study found significant differences at earlier time points, namely, at day 10 after treatment for certain components of QOL and at day 42 after treatment for most QOL domains.
Of interest, when patients with a pain response after radiotherapy also experienced improvements in psychosocial aspects at day 10, nonresponders experienced deterioration in the same domain. Finally, the day 42 results of our study confirm previous findings that a greater proportion of those who have a pain response after radiotherapy will also have a clinically meaningful response in many QOL domains.11,13 In this study, these items were pain, fatigue, constipation, painful sites, pain characteristics, physical functioning, emotional functioning, functional interference, psychosocial aspects, and global QOL. Because the main purpose of radiotherapy was pain relief, the most significantly improved QOL outcomes in the present study were pain-related outcomes, such as pain characteristics, number of painful sites, and functional interference. The QOL items that were not expected to improve with pain reduction, including dyspnea and insomnia, did not significantly improve in responders. Thus, physicians should use other more appropriate treatment modalities to address these symptoms separately.
After rescaling the pain severity to 0 to 100 as that of the pain item score in QOL, at day 10, a mean 20-point reduction in pain severity and an 8-point improvement in QOL were observed. At day 42, the magnitude in mean QOL improvement increased to 20 points for a 20-point reduction in pain severity. In addition, more QOL items were improved at day 42 compared with day 10. Therefore, although improvements in pain may occur by day 10, QOL may require more time to improve because it encompasses all aspects of well-being.
Our study is limited by the attrition rates in follow-up at day 42, leading to a high exclusion rate as in most palliative trials. In addition, the study patients may not be fully representative of all patients with short life expectancy.
Our study reports clinically meaningful improvements in pain and QOL as early as 10 days after radiotherapy and more extensive improvements in QOL at 42 days after treatment. These results confirm that radiotherapy for painful bone metastases can improve pain and QOL quickly after treatment and therefore should be offered even for those with a limited expected survival. Our evaluation time points (days 10 and 42) should be used in future studies that involve similar patient populations because they are more relevant than evaluating those with poor expected survival at 2 or even 3 months after treatment.
Corresponding Author: Edward Chow, MBBS, MSc, PhD, FRCPC, Department of Radiation Oncology, Odette Cancer Centre, Sunnybrook Health Sciences Centre, Toronto, ON M4N 3M5, Canada (firstname.lastname@example.org).
Accepted for Publication: November 23, 2016.
Published Online: February 9, 2017. doi:10.1001/jamaoncol.2016.6770
Author Contributions: Dr Chow had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: McDonald, Ding, Brundage, Meyer, Mahmud, Fairchild, Wu, DeAngelis, Wong, Chow.
Acquisition, analysis, or interpretation of data: McDonald, Ding, Brundage, Meyer, Nabid, Chabot, Coulombe, Ahmed, Kuk, Dar, Wilson, Dennis, DeAngelis, Wong, Zhu, Chow.
Drafting of the manuscript: McDonald, Meyer, Mahmud, Wu, DeAngelis, Zhu, Chow.
Critical revision of the manuscript for important intellectual content: Ding, Brundage, Meyer, Nabid, Chabot, Coulombe, Ahmed, Kuk, Dar, Fairchild, Wilson, Dennis, DeAngelis, Wong, Chow.
Statistical analysis: Ding, Meyer, Zhu.
Obtained funding: Meyer, Wu, DeAngelis, Wong.
Administrative, technical, or material support: McDonald, Brundage, Meyer, Kuk, Fairchild, Wilson, Wu, Dennis, DeAngelis, Wong.
Study supervision: Kuk, DeAngelis, Wong, Chow.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by NCIC Clinical Trials Group programmatic grants from the Canadian Cancer Society Research Institute.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Meeting Presentation: This study was presented at 21st International Congress on Palliative Care; October 19, 2016; Montreal, Quebec, Canada.
Additional Contributions: We appreciate the participation of all the patients and research teams.
Create a personal account or sign in to: