Abstract
Purpose:
American Society for Radiation Choosing Wisely guidelines recommend ≤10 fractions of radiation therapy (RT) for bone metastasis, with consideration for 1 fraction in patients with a poor prognosis. The purpose of this analysis was to evaluate characteristic differences in guideline concordance to fractionation regimens in a modern cohort of older patients with a diagnosis of bone metastasis.
Methods and materials
Medicare beneficiaries aged ≥65 years treated with RT for bone metastasis from 2012 to 2015 were identified. Guideline-concordant RT fractionation was defined in the entire cohort as ≤10 fractions. Utilization of 1 fraction versus ≥2 fractions was analyzed in deceased patients. Patient demographic, disease, and facility characteristics associated with shorter fractionation were analyzed.
Results
In 569 patients treated with RT, the median age at diagnosis was 73 years. The most common cancer types were lung (37%), genitourinary (26%), breast (15%), and gastrointestinal (10%). Among all patients, 34%, 30%, and 36% received 1 fraction, 2 to 10 fractions, and ≥11 fractions, respectively. In comparison with receipt of 1 to 10 fractions, receipt of ≥11 fractions was associated with a $1467 increase in per-patient cost to Medicare during the calendar quarter of RT. Almost two-thirds of patients who died within 30 days of RT completion were treated with >1 fraction.
Conclusions
Although guideline concordance was high overall, a large number of patients received longer courses of RT at the end of life. Strong consideration should be made for utilization of shorter courses, particularly in patients with a limited prognosis.
INTRODUCTION
Bone is a common site of metastasis in patients with advanced cancer. It is generally accepted that radiation therapy (RT) has the potential to provide palliative pain relief and minimize morbidity for patients with metastatic bony disease. With complete pain relief in up to 50% of patients and a response rate of 85%, palliative-intent RT can provide symptom relief within 1 to 2 weeks of treatment (1). Several randomized trials and meta-analyses have shown the efficacy of 1 or 5 versus ≥10 RT treatments for palliation of bone metastasis, with the caveat that 20% of patients may need repeat RT when treated with a single-fraction regimen 1, 2, 3, 4, 5. There is great variability in utilization of fractionation regimens around the world, with single-fraction RT rates of 65% in Canada, 75% in Australia, 37% in Sweden, 40% in the United Kingdom, and 19% in Poland (6). In the United States, the adoption rate has been lower, with common use of prolonged regimens.
The 2014 American Society for Radiation Oncology Choosing Wisely campaign issued guidelines recommending no more than 10 fractions, with consideration for 1 fraction in patients with a poor prognosis or prolonged geographic distance (7). The US Agency for Healthcare Research and Quality identified duration of RT for bone metastasis as the first quality metric to be evaluated in radiation oncology. Other professional societies have put forth stronger recommendations for 1-fraction courses of treatment. Choosing Wisely Canada and the American Academy of Hospice and Palliative Medicine do not recommend more than a single fraction of RT for uncomplicated bone metastasis 8, 9. This is evident in the extensive utilization of single-fraction regimens in British Columbia, where nearly half of patients were treated with a single-fraction regimen (10). Despite strong level I evidence supporting shorter courses of treatment, review of reported data in the United States reveals low uptake (11). At the time of this report, there are no recent US claims-based reports assessing the impact of guideline recommendations on the utilization of shorter courses of RT. The current analysis serves to assess uptake of recent guideline recommendations in a modern cohort of patients to fill this gap in the literature.
METHODS AND MATERIALS
This was a population-based analysis of 12 sites in 5 states using linked Medicare claims and tumor registry data for cancer patients within the University of Alabama Birmingham Health System Cancer Community Network (CCN) from 2012 to 2015 12, 13. The CCN comprises 12 academic and community cancer centers located in five southeastern states (Alabama, Georgia, Florida, Mississippi, Tennessee). Institutional review board approval was obtained at individual sites. Patients aged ≥65 years with stage IV disease and a diagnosis of bone metastasis who were treated with RT from 2012 to 2015 were identified (Table E1, available online at https://doi-org.ezproxy3.lhl.uab.edu/10.1016/j.ijrobp.2018.05.061). Those with continuous part A and B data and no health maintenance organization coverage were included. Patients with multiple myeloma or hematologic primaries were excluded. Receipt of RT was categorized into 1 fraction, 2 to 10 fractions, and ≥11 fractions. Guideline-concordant RT was defined as ≤10 fractions. Variables of interest were age at diagnosis (65–69 years, 70–74 years, ≥74 years), race (white, other), comorbidity index (0, 1, ≥2), treatment center volume (high [≥4000 cases/year], low), and primary cancer type (Table E2, available online at https://doi-org.ezproxy3.lhl.uab.edu/10.1016/j.ijrobp.2018.05.061) 14, 15, 16, 17. Patient disease, treatment, and center characteristics by survival status were analyzed using 2-sample t tests (continuous measures) and χ2 tests of independence (categorical measures). Mixed-effects models accounting for the within-site correlation of observations were used to compare total costs with Medicare per patient during the quarter of RT by concordance status. Models included a random intercept and slope. Analysis of utilization of shorter fractionation regimens (1 vs ≥2) was performed on the subset of study-eligible patients who died within the study period. Dates of death were only available through 2015. All analyses were performed using SAS software, version 9.4 (SAS Institute, Cary, NC). Results were considered statistically significant if the P value was <.05.
RESULTS
Among patients with a diagnosis of bone metastasis and fee-for-service Medicare coverage, we identified 569 patients treated with RT (Fig. 1). The median age at diagnosis was 73 years (interquartile range [IQR], 8.4 years), 89% of patients were white, and 53% had a Charlson Comorbidity Index of 2 or more (Table 1). The most common primary sites were lung (37%), genitourinary (26%), breast (15%), and gastrointestinal (10%). Overall guideline concordance was 64%: 1 fraction in 34% and 2 to 10 fractions in 30%. In patients with nonconcordant RT (≥11 fractions), the median number of fractions was 16 (IQR, 8). On bivariate analysis, treatment center volume was associated with guideline-concordant fractionation; centers with large volumes were more likely to deliver guideline-concordant fractionation regimens (69% vs 55%, P = .002). In comparison with receipt of 1 to 10 fractions, receipt of ≥11 fractions was associated with an estimated $1467 increase in per-patient cost to Medicare during the calendar quarter of RT. This increase in cost was not driven by utilization of systemic agents. No patient or disease characteristics were associated with use of shorter fractionation. Larger-volume centers were more likely to treat patients with shorter fractionation regimens (Table 2).
Fig. 1.
Study population exclusion cascade. Abbreviations: CCN = Cancer Community Network; HMO = health maintenance organization; met = metastasis; UAB = University of Alabama Birmingham.
Table 1.
Patient characteristics by vital status (N = 569)
| Empty Cell | All (N = 569) | Alive (n = 139) | Dead (n = 430) | P value |
|---|---|---|---|---|
| Median age at diagnosis (IQR), y | 72.8 (8.4) | 72.1 (8.6) | 73.1 (8.7) | .03 |
| Age at diagnosis, n (%) | ||||
| 65–69 y | 187 (32.9) | 54 (38.9) | 133 (30.9) | .19 |
| 70–74 y | 182 (32.0) | 43 (30.9) | 139 (32.3) | |
| ≥75 y | 200 (35.2) | 42 (30.2) | 158 (36.7) | |
| Sex, n (%) | .01 | |||
| Male | 329 (58.0) | 68 (49.0) | 261 (61.0) | |
| Female | 238 (42.0) | 71 (51.1) | 167 (39.0) | |
| Race, n (%) | .84 | |||
| White | 505 (88.8) | 124 (89.2) | 381 (88.6) | |
| Nonwhite | 64 (11.3) | 15 (10.8) | 49 (11.4) | |
| Primary site, n (%) | <.001 | |||
| Lung | 209 (36.7) | 31 (22.3) | 178 (41.4) | |
| GU | 147 (25.8) | 42 (30.2) | 105 (24.4) | |
| Breast | 88 (14.7) | 44 (31.7) | 44 (10.2) | |
| GI | 58 (10.2) | * | * | |
| Other | 67 (11.8) | * | * | |
| Fractionation, n (%) | .60 | |||
| 1 | 193 (33.9) | 45 (32.4) | 148 (34.4) | |
| 2–10 | 171 (30.1) | 39 (28.1) | 132 (30.7) | |
| ≥11 | 205 (36.0) | 55 (39.6) | 150 (34.9) | |
| Comorbidity index, n (%) | <.001 | |||
| 0 | 134 (23.6) | 52 (37.4) | 82 (19.1) | |
| 1 | 135 (23.7) | 35 (25.2) | 100 (23.3) | |
| ≥2 | 300 (52.7) | 52 (37.4) | 248 (57.7) | |
| Treatment center volume, n (%) | .13 | |||
| High† | 367 (64.5) | 97 (69.8) | 270 (62.8) | |
| Low | 202 (35.5) | 42 (30.2) | 160 (37.2) |
Abbreviations: GI = gastrointestinal; GU = genitourinary; IQR = interquartile range.
Numbers are too small to report.
High treatment center volume is defined as ≥4000 cases/year.
Table 2.
Patient characteristics by fractionation (N = 569)
| 1 fraction (n = 193) | 2–10 fractions (n = 171) | ≥11 fractions (n = 205) | P value | |
|---|---|---|---|---|
| Median age at diagnosis (IQR), y | 72.9 (8.3) | 72.8 (8.4) | 72.6 (8.8) | .82 |
| Age at diagnosis, n (%) | ||||
| 65–69 y | 59 (30.6) | 56 (32.8) | 72 (35.1) | .53 |
| 70–74 y | 63 (32.6) | 61 (35.7) | 58 (28.3) | |
| ≥75 y | 71 (36.8) | 54 (31.6) | 75 (36.6) | |
| Race, n (%) | .69 | |||
| White | 172 (89.1) | 154 (90.1) | 179 (87.3) | |
| Nonwhite | 21 (10.9) | 17 (9.9) | 26 (12.7) | |
| Primary site, n (%) | .23 | |||
| Lung | 76 (39.4) | 63 (36.8) | 70 (34.2) | |
| GU | 46 (23.8) | 40 (23.4) | 61 (29.8) | |
| Breast | 32 (16.6) | 24 (14.0) | 32 (15.6) | |
| GI | 19 (9.8) | 20 (11.7) | 19 (9.3) | |
| Other | 20 (10.4) | 24 (14.0) | 23 (11.2) | |
| Comorbidity index, n (%) | .62 | |||
| 0 | 48 (24.9) | 44 (25.7) | 42 (20.5) | |
| 1 | 49 (25.4) | 36 (21.1) | 50 (24.4) | |
| ≥2 | 96 (49.7) | 91 (53.2) | 113 (55.1) | |
| Treatment center volume, n (%) | .007 | |||
| High* | 135 (70.0) | 117 (68.4) | 115 (56.1) | |
| Low | 58 (30.1) | 54 (31.6) | 90 (43.9) |
Abbreviations: GI = gastrointestinal; GU = genitourinary; IQR = interquartile range.
High treatment center volume is defined as ≥4000 cases/year.
In a subset analysis of 430 patients with known dates of death, the results were similar to those reported earlier. The median time from the last fraction of RT to death was 168 days (IQR, 314 days), with 21% and 36% treated for ≤30 days and ≤90 days, respectively. The proportion of patients treated with shorter courses of RT did not change in those who died within 30 or 90 days of completion of RT (Table 3).
Table 3.
Fractionation in deceased patients (n = 430)
| Empty Cell | 1 fraction (n = 148) | 2–10 fractions (n = 132) | ≥11 fractions (n = 150) |
|---|---|---|---|
| All deceased | 148 (34.4) | 132 (30.7) | 150 (34.9) |
| Deceased ≤30 d from RT | 32 (34.8) | 33 (35.9) | 27 (29.4) |
| Deceased ≤90 d from RT | 54 (34.6) | 56 (35.9) | 46 (30.7) |
Abbreviation: RT = radiation therapy.
DISCUSSION
The purpose of Choosing Wisely recommendations is to encourage discourse about the value provided for health care services rendered. In this analysis, compliance with American Society for Radiation Oncology Choosing Wisely recommendations was high overall (64%). Despite our hypothesis that more patients would be treated with shorter courses in the last 30 days of life, there was no change in this trend over survival time. A recent systematic review of RT use for any indication in the last 30 days of life revealed great heterogeneity, from 0% to 59%, regarding the use of single-fraction regimens (18). This is likely multifactorial in etiology. First, it is generally accepted that survival estimates by physicians are difficult to establish, particularly during the final months and weeks of life (19). This is made even more difficult by the multitude of newer, targeted agents that can prolong survival in patients with particular molecular subtypes, such as those seen for lung cancer in recent years. Second, physician comfort level with utilization of single-fraction regimens likely plays a role in determining fractionation regimens. An international survey of physician patterns of practice showed great variability in utilization of different fractionation regimens, ranging from 3 Gy in 1 fraction to 60 Gy in 20 fractions, with 30 Gy in 10 fractions being the most common regimen (20). In a more recent review of the literature looking at patterns of practice, continued variability was noted throughout the world (6). In Canada, there was early adoption of single fractionation, with more than 25% and 33% of patients treated with a single fraction from 1984 to 1986 and from 1999 to 2001, respectively. There has been a further increase in the uptake of shorter regimens to encompass half to two-thirds of patients within the past decade. Uptake of shorter fractions has been more heterogeneous in Europe, ranging from 11% in Germany to almost 50% in Spain. Rates of single-fraction utilization were the lowest in the United States, however, often in the single digits. Although it is routinely cited that up to 20% of patients may require more re-treatment if treated with a single course of therapy, it should be noted that these data did not factor in the importance of physician reluctance for re-treatment in longer-course regimens in comparison with single-fraction treatment. It has been shown that a physician’s culture of practice can influence the fractionation regimen (10). In addition, physician training in and comfort level with the use of single-fraction regimens have likely contributed to the low uptake of single-fraction regimens 5, 10, 21, 22.
To our knowledge, this is the most recent population-based analysis in the United States showing a clinically relevant increase in the utilization of single-fraction RT compared with historical controls. However, this study has several limitations inherent to a claims-based analysis. Medicare is a federal health insurance program primarily for people aged ≥65 years. Although it allows for large-scale claims-based analysis, its primary design was not for research purposes. Hence, it is subject to coding errors that may have occurred as part of routine care. These limitations are commonly known and accepted, as Medicare data alone or in combination with other datasets provide for a rich appreciation of patterns of practice in the modern era. Second, it is not known whether bone metastases were due to complicated events, such as pathologic fracture or cord compression. Complicated bone metastases were not included in the majority of clinical trials; hence the role of single-fractionation treatment is less clear. Although it is likely that all patients would not be candidates for single-fraction treatment, it may be reasonable to estimate that 60% may have such an option (23). Third, a history of RT could necessitate utilization of longer fractionation schemes to avoid additional morbidity associated with higher doses administered during shorter courses. Lack of knowledge about clinical circumstances surrounding the choice of fractionation regimens and no knowledge of pain response to the RT regimen are further limitations. Finally, this dataset is lacking information about physician and patient preferences that may have played a part in fractionation regimens. These clinical factors were likely important in the decision-making process, but it is unlikely that more than two-thirds of these patients required re-treatment using longer courses in the last 30 days of life because of such factors. Possible solutions may be better prognostication tools, increasing involvement of palliative care services with goals-of-care discussion, and utilization of clinical pathways 24, 25.
In conclusion, this is a modern US-based population analysis showing a clinically relevant increase in the utilization of single-fraction RT for bone metastasis compared with historical controls. However, nearly two-thirds of patients received >1 fraction of RT in the last 30 days of life. Practitioners should consider utilization of shorter fractions, particularly in patients with a poor prognosis.
Supplementary Material
REFERENCES
- 1.Coleman R, Body JJ, Aapro M, et al. Bone health in cancer patients: ESMO clinical practice guidelines Ann Oncol, 25 (Suppl. 3) (2014), pp. iii124-iii137 [DOI] [PubMed] [Google Scholar]
- 2.Steenland E, Leer JW, van Houwelingen H, et al. The effect of a single fraction compared to multiple fractions on painful bone metastases: A global analysis of the Dutch Bone Metastasis Study Radiother Oncol, 52 (1999), p. 101. [DOI] [PubMed] [Google Scholar]
- 3.Van der Linden YM, Lok JJ, Steenland E, et al. Single fraction radiotherapy is efficacious: A further analysis of the Dutch Bone Metastasis Study controlling for the influence of retreatment Int J Radiat Oncol Biol Phys, 59 (2004), p. 528. [DOI] [PubMed] [Google Scholar]
- 4.Hartsell WF, Scott CB, Bruner DW, et al. Randomized trial of short- versus long-course radiotherapy for palliation of painful bone metastases J Natl Cancer Inst, 97 (2005), p. 798. [DOI] [PubMed] [Google Scholar]
- 5.Chow E, Harris K, Fan G, et al. Palliative radiotherapy trials for bone metastases: A systematic review J Clin Oncol, 25 (2007), pp. 1423–1436 [DOI] [PubMed] [Google Scholar]
- 6.McDonald R, Chow E, Lam H, et al. International patterns of practice in radiotherapy for bone metastases: A review of the literature J Bone Oncol, 3 (2014), pp. 96–102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hahn C, Kavanagh B, Bhatnagar A, et al. Choosing wisely: The American Society for Radiation Oncology’s top 5 list Pract Radiat Oncol, 4 (2014), pp. 349–355 [DOI] [PubMed] [Google Scholar]
- 8.Oncology: Ten things physicians and patients should question. Available at: https://choosingwiselycanada.org/oncology/. Accessed April 23, 2018.
- 9.Choosing Wisely: Five things physicians and patients should question in hospice and palliative medicine. Available at: http://aahpm.org/outreach/choosing-wisely. Accessed April 23, 2018. [DOI] [PubMed]
- 10.Olson RA, Tiwana MS, Barnes M, et al. Use of single- versus multiple-fraction palliative radiation therapy for bone metastases: Population-based analysis of 16,898 courses in a Canadian province Int J Radiat Oncol Biol Phys, 89 (2014), pp. 1092–1099 [DOI] [PubMed] [Google Scholar]
- 11.Bekelman JE, Epstein AJ, Emanuel EJ Single- vs multiple-fraction radiotherapy for bone metastases from prostate cancer JAMA, 310 (2013), pp. 1501–1502 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Rocque GB, Partridge EE, Pisu M, et al. The Patient Care Connect Program: Transforming health care through lay navigation J Oncol Pract, 12 (2016), pp. e633–e642 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Rocque GB, Pisu M, Jackson BE, et al. Resource use and Medicare costs during lay navigation for geriatric patients with cancer JAMA Oncol, 3 (2017), pp. 817–825 View PDF This article is free to access. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation J Chronic Dis, 40 (1987), pp. 373–383 [DOI] [PubMed] [Google Scholar]
- 15.Deyo RA, Cherkin DC, Ciol MA Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases J Clin Epidemiol, 45 (1992), pp. 613–619 [DOI] [PubMed] [Google Scholar]
- 16.Romano PS, Roos LL, Jollis JG Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: Differing perspectives J Clin Epidemiol, 46 (1993), pp. 1075–1079 discussion 1081–1090 [DOI] [PubMed] [Google Scholar]
- 17.Klabunde CN, Potosky AL, Legler JM, et al. Development of a comorbidity index using physician claims data J Clin Epidemiol, 53 (2000), pp. 1258–1267 [DOI] [PubMed] [Google Scholar]
- 18.Park KR, Lee CG, Tseng YD, et al. Palliative radiation therapy in the last 30 days of life: A systematic review Radiother Oncol, 125 (2017), pp. 193–199 [DOI] [PubMed] [Google Scholar]
- 19.Glare P, Virik K, Jones M, et al. A systematic review of physicians’ survival predictions in terminally ill cancer patients BMJ, 327 (2003), pp. 195–198 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Fairchild A, Barnes E, Ghosh S, et al. International patterns of practice in palliative radiotherapy for painful bone metastases: Evidence-based practice? Int J Radiat Oncol Biol Phys, 75 (2009), pp. 1501–1510 [DOI] [PubMed] [Google Scholar]
- 21.Crellin AM, Marks A, Maher EJ Why don’t British radiotherapists give single fractions of radiotherapy for bone metastases? Clin Oncol, 1 (1989), pp. 63–66 [DOI] [PubMed] [Google Scholar]
- 22.Roos DE Continuing reluctance to use single fractions of radiotherapy for metastatic bone pain: An Australian and New Zealand practice survey and literature review Radiother Oncol, 56 (2000), pp. 315–322 [DOI] [PubMed] [Google Scholar]
- 23.Tiwana MS, Barnes M, Yurkowski E, et al. Incidence and treatment patterns of complicated bone metastases in a population-based radiotherapy program Radiother Oncol, 118 (2016), pp. 552–556 [DOI] [PubMed] [Google Scholar]
- 24.Olson RA, Tiwana M, Barnes M, et al. Impact of using audit data to improve the evidence-based use of single-fraction radiation therapy for bone metastases in British Columbia Int J Radiat Oncol Biol Phys, 94 (2016), pp. 40–47 [DOI] [PubMed] [Google Scholar]
- 25.Gebhardt BJ, Rajagopalan MS, Gill BS, et al. Impact of dynamic changes to a bone metastases pathway in a large, integrated, National Cancer Institute-designated comprehensive cancer center network Pract Radiat Oncol, 5 (2015), pp. 398–405 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.