- Email: [email protected]
Adoption of Hypofractionated Whole-Breast Irradiation for Early-Stage Breast Cancer: A National Cancer Data Base Analysis
International Journal of
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation
Adoption of Hypofractionated Whole-Breast Irradiation for Early-Stage Breast Cancer: A National Cancer Data Base Analysis Elyn H. Wang, BS,* Sarah S. Mougalian, MD,*,y,z Pamela R. Soulos, MPH,*,z Charles E. Rutter, MD,*,z,x Suzanne B. Evans, MD, MPH,*,z,x Bruce G. Haffty, MD,jj Cary P. Gross, MD,*,jj,{ and James B. Yu, MD*,z,x *Yale School of Medicine, Yale School of Medicine, New Haven, Connecticut; yDepartment of Medical Oncology, Yale School of Medicine, New Haven, Connecticut; zCancer Outcomes, Public Policy, and Effectiveness Research Center at Yale, New Haven, Connecticut; xDepartment of Therapeutic Radiology, Yale School of Medicine, New Haven, Connecticut; jjDepartment of Radiation Oncology, Rutgers Cancer Institute of New Jersey and Robert Wood Johnson Medical School, New Brunswick, New Jersey; and {Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut Received May 27, 2014, and in revised form Jun 12, 2014. Accepted for publication Jun 14, 2014.
Summary Hypofractionated wholebreast radiation therapy (hypofractionation) for early-stage breast cancer is supported by high-quality randomized trial evidence and clinical guidelines. We performed a retrospective study of breast cancer
Purpose: To evaluate the relationship of patient, hospital, and cancer characteristics with the adoption of hypofractionation in a national sample of patients diagnosed with early-stage breast cancer. Methods and Materials: We performed a retrospective study of breast cancer patients in the National Cancer Data Base from 2004-2011 who were treated with radiation therapy and met eligibility criteria for hypofractionation. We used logistic regression to identify factors associated with receipt of hypofractionation (vs conventional fractionation). Results: We identified 13,271 women (11.7%) and 99,996 women (88.3%) with earlystage breast cancer who were treated with hypofractionation and conventional fractionation, respectively. The use of hypofractionation increased significantly, with
Reprint requests to: James B. Yu, MD, Department of Therapeutic Radiology, Yale School of Medicine, HRT 138, 333 Cedar St, New Haven, CT 06520. Tel: (203) 785-5703; E-mail: [email protected] This work was supported by the National Institutes of Health (NIH; grant R01CA149045). J.B.Y. is also supported by Clinical Translational Science Awards grant KL2 TR000140 from the National Center for Advancing Translational Sciences, a component of the NIH, and NIH Roadmap for Medical Research. J.B.Y. and C.P.G. have research funding from 21st Century Oncology, LLC. C.P.G. also has research funding from Medtronic and Johnson & Johnson. The National Cancer Data Base is a joint project of the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society. The American College of Surgeons has executed a Int J Radiation Oncol Biol Phys, Vol. 90, No. 5, pp. 993e1000, 2014 0360-3016/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.06.038
Business Associate Agreement that includes a data use agreement with each of its CoC-accredited hospitals. The data used in the study are derived from a deidentified National Cancer Data Base file. The American College of Surgeons and the CoC have not verified and are not responsible for the analytic or statistical methodology used, or the conclusions drawn from these data by the investigator. The study sponsor (NIH) did not play a role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the manuscript for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Conflict of interest: J.B.Y. and C.P.G. have research funding from 21st Century Oncology, LLC.
994
International Journal of Radiation Oncology Biology Physics
Wang et al.
patients in the National Cancer Data Base from 2004-2011 who were treated with radiation therapy and met eligibility criteria for hypofractionation. We found that the use of hypofractionation is rising and is associated with increased travel distance and treatment at an academic center.
5.4% of patients receiving it in 2004 compared with 22.8% in 2011 (P<.001 for trend). Patients living 50 miles from the cancer reporting facility had increased odds of receiving hypofractionation (odds ratio 1.57 [95% confidence interval 1.44-1.72], P<.001). Adoption of hypofractionation was associated with treatment at an academic center (P<.001) and living in an area with high median income (P<.001). Hypofractionation was less likely to be used in patients with high-risk disease, such as increased tumor size (P<.001) or poorly differentiated histologic grade (P<.001). Conclusions: The use of hypofractionation is rising and is associated with increased travel distance and treatment at an academic center. Further adoption of hypofractionation may be tempered by both clinical and nonclinical concerns. Ó 2014 Elsevier Inc.
Introduction Randomized clinical trials in patients with early-stage breast cancer treated with breast-conserving surgery (BCS) have demonstrated that adjuvant whole-breast irradiation lowers the risk of tumor recurrence and improves survival (1). Historically, most studies have focused on the benefits of conventionally fractionated whole-breast irradiation (conventional fractionation), which targets the entire breast with 1.80-2.00 Gy once daily for a total dose of 45-50 Gy in 25-28 daily fractions over 5 to 6 weeks (2, 3). In modern practice, hypofractionated whole-breast irradiation (hypofractionation) has emerged as a viable alternative to conventional fractionation. Hypofractionation treats the entire breast with 2.5-3.20 Gy once daily for a total dose of 39-42.5 Gy in 13-16 fractions for 3-5 weeks (4-11). Depending on the scheme of choice, radiation can be delivered every day (4, 8-10) or delivered in 13 fractions over 35 days (7). Compared with conventional fractionation, hypofractionation involves a shorter treatment course, decreased total dose and number of fractions, and results in increased patient convenience and lower costs (12). In the past several years, randomized trials have tested hypofractionation against conventional fractionation after BCS for early-stage breast cancer patients (4, 7, 8, 10). These trials, originally reported in 2005, have shown that hypofractionation produces disease-free survival and overall survival outcomes that are equivalent to those for conventional fractionation (5-10, 13, 14) in properly selected patients. Furthermore, recently reported 10-year follow-up of these trials has demonstrated that the long-term toxicity profiles for both methods are comparable (10, 13, 14). As a result, clinical practice guidelines from the American Society for Radiation Oncology were recently published establishing hypofractionation as an appropriate therapeutic option for most patients with early breast cancer (15), and the American Society for Radiation Oncology has identified hypofractionation as one of its 5 Choosing Wisely recommendations, a national campaign focused on encouraging conversations between patients and providers about evidence-based treatment choices (13). Despite the strong evidence and guidelines supporting its use, there is some indication that physicians may be
slow to adopt hypofractionation: A recent survey study of physician practice patterns revealed that 56.2% of physicians do not use hypofractionation at all in the post-BCS setting, in conflict with patient preference (16). It is possible that physicians’ clinical concerns (that larger hypofractionated doses of radiation therapy [RT] per treatment are related to greater late side effects) are limiting usageddespite a lack of difference in toxicity seen in randomized trials with long follow-up (10). Physicians may also be concerned with evidence suggesting that hypofractionation for patients with high-grade disease is associated with inferior outcomes (10), although this was not seen in subsequent trials (8). Additionally, nonclinical factors such as financial pressure could also be impeding the adoption of hypofractionation, given that a less costly treatment also means less reimbursement to the physician and hospital. Given the clinical evidence, improved convenience, and decreased costs, understanding the adoption of hypofractionation is critical as radiation-related cancer costs are targeted as an area of potential health savings. We therefore evaluated the relationship of patient, hospital, and cancer characteristics with the adoption of hypofractionation in a national sample of patients diagnosed with early-stage breast cancer. We hypothesized that the use of hypofractionation has increased, particularly in patients for whom travel to and from radiation treatment would be difficult, owing either to distance, age, or comorbidity. We also hypothesized that factors that were indicative of less physician financial pressure (such as academic practice location) and lower-risk disease were associated with adoption of hypofractionation.
Methods and Materials Data source and study population We used National Cancer Data Base (NCDB) registry data from 2004 to 2011 to examine receipt of adjuvant wholebreast irradiation in patients eligible to receive both hypofractionated and conventional fractionation. The NCDB is a joint program of the Commission on Cancer of the
Volume 90 Number 5 2014
American College of Surgeons and the American Cancer Society (17-19). The NCDB, established in 1989, is a nationwide, facility-based, clinical surveillance data set that currently captures 70% of all newly diagnosed malignancies in the United States annually. Data reported to the NCDB are retrospective, and no patient or physician identifiers are collected (20). This study was granted an exemption from our institution’s Human Investigation Committee, because it was research involving existing data on unidentifiable research subjects. Patients with a first and only diagnosis of breast cancer whose data were entered in the NCDB from 2004 through 2011 were the subject of this study. To better evaluate factors independent of hypofractionation eligibility, we restricted the sample to early-stage female breast cancer patients who received adjuvant RT and who met the criteria for hypofractionation receipt according to guidelines and randomized trials. These criteria included (1) age 50 years or older at time of diagnosis; (2) pathologic stage T1-2 N0 according to the American Joint Committee on Cancer TNM staging classification; (3) treatment with BCS; and (4) no chemotherapy treatment (15). We limited our analysis to adjuvant beam radiation delivered to the intact breast as the primary target. We did not include patients who received regional nodal RT as reported by the NCDB. Brachytherapy, stereotactic RT, and treatments delivered using electron, neutron, or proton beams were excluded. Total dose of regional RT was limited to 40-66.4 Gy, and boost doses were limited to <21.6 Gy. Patients who received fewer than 10 or more than 50 fractions and RT treatments coded as palliative care were further excluded from the final sample.
Study covariates and primary outcome Patient demographic and clinical information included as independent variables were age, race, type of primary health insurance, median income in zip code of residence by quartile, degree of rurality, and urban influence of patient’s residence using classification published by the US Department of Agriculture Economic Research Service, and the Charlson/Deyo comorbidity score. We also evaluated facility type (academic vs nonacademic) and geographic region (Northeast, Midwest, South, West). We created an indicator for whether patients had to travel a long distance to receive cancer care. Because the distance from the patients’ residence to the radiation treatment facility was not specifically recorded, we used distance to the cancer reporting facility as a proxy for distance travelled. We classified 50 miles as a long distance (21, 22). To check the validity of this assumption, analysis of distance traveled was performed for the subset of patients for whom the cancer reporting facility and radiation treatment facility were the same (10.5% of the sample). Cancer characteristics assessed included year of cancer diagnosis, tumor differentiation, size, estrogen receptor (ER) and human
Trends in hypofractionated breast
995
epidermal growth factor receptor 2 (HER2) assay results, and surgical margins. Unknown or missing values were entered as dummy variables of a separate category. The primary outcome was type of adjuvant RT received (hypofractionation vs conventional fractionation). Fraction dose was defined as total dose divided by the total number of treatment sessions administered during the first course of treatment. To take into account small inaccuracies in the recorded dose and number of fractions delivered, hypofractionation was broadly defined as fraction dose 2.2 and 4.0 Gy, and patients were required to have received at least 40 Gy. Conventional fractionation was broadly defined as fraction dose between 1.5 and 2.2 Gy.
Statistical analysis Descriptive statistics were used to summarize patient, facility, and cancer characteristics. Bivariate associations of patient, facility, and cancer variables were tested by the Pearson’s c2 test. Covariates associated with hypofractionation at P.10 were then included in a multivariable logistic regression model. A two-sided P value of <.05 was used to determine statistical significance. Stata SE version 13.0 (Stata Corp, College Station, TX) was used to perform all statistical analyses.
Results The total number of female patients with a first and only diagnosis of breast cancer reported in the NCDB increased from 139,300 patients in 2004 to 165,125 patients in 2009 and then decreased to 164,760 patients in 2011. This fluctuation was likely due to the changing number of facilities reporting cases to the NCDB, which also ranged from a low of 1298 in 2011 to a peak of 1329 in 2009. The number of early-stage breast cancer cases reported followed a similar trend as total breast cancer cases, with 57,984 patients in 2004 and a peak of 64,854 in 2009, followed by a decline in subsequent years. Despite this fluctuation, the overall number of early-stage breast cancer patients who received adjuvant RT (hypofractionation or conventional) who met our study’s criteria for hypofractionation increased steadily from 12,412 patients in 2004 to 16,685 patients in 2011, an increase of 34.4%. We identified 13,271 (11.7%) and 99,996 (88.3%) female early-stage breast cancer patients >50 years of age not receiving chemotherapy who were treated between 2004 and 2011 with hypofractionation and conventional fractionation (respectively) for whole-breast irradiation after BCS. The number of patients who met our study’s criteria for and underwent hypofractionation increased steadily from 677 patients (5.4% of patients undergoing whole-breast irradiation) in 2004 to 3809 patients (22.8%) in 2011 (P for trend, <.001) (Table 1). Furthermore, the rate of patients who underwent hypofractionation increased from 3.09 per 100 patients in 2004 to 13.05 in 2011,
996
International Journal of Radiation Oncology Biology Physics
Wang et al.
Table 1
Patient, facility, and cancer characteristics by RT modality Feature
Facility type Academic Community Comprehensive community Other Facility location South Northeast Midwest West Age in years 50-59 60-69 70-79 80 Race White non-Hispanic Black White Hispanic Other Primary payor Private insurance No insurance Medicaid Medicare Other government Unknown Median income quartile <$30,000 $30,000-$35,000 $35,000-$45,999 $46,000 Unknown Urbanization Metro Urban Rural Distance from cancer reporting facility <50 miles 50 miles Unknown Charlson-Deyo score 0 1 2 Year of diagnosis 2004 2005 2006 2007 2008 2009 2010 2011 Laterality Right-sided Left-sided
Conventional fractionation (nZ99,996)
Hypofractionation (nZ13,271)
P* <.001
23,307 12,156 63,488 1045
(82.8) (92.3) (89.7) (90.5)
4830 1018 7313 110
(17.3) (7.7) (10.3) (9.5)
29,634 23,848 29,060 17,454
(90.0) (88.7) (88.6) (84.5)
3279 3038 3747 3207
(10.0) (11.3) (11.4) (15.5)
29,959 37,369 25,366 7302
(91.2) (89.2) (86.4) (80.0)
2896 4544 4006 1825
(8.8) (10.8) (13.6) (20.0)
86,670 6715 2741 3870
(88.6) (89.0) (85.8) (83.3)
11,210 828 455 778
(11.5) (11.0) (14.2) (16.7)
49,153 983 2879 45,121 713 1147
(89.9) (88.2) (89.1) (86.6) (91.1) (86.6)
5545 131 354 6993 70 178
(10.1) (11.8) (10.9) (13.4) (8.9) (13.4)
8912 15,600 27,490 43,402 4592
(89.6) (90.1) (89.8) (86.6) (87.2)
1038 1712 3115 6729 677
(10.4) (9.9) (10.2) (13.4) (12.8)
<.001
<.001
<.001
<.001
<.001
<.001 79,523 (88.0) 13,886 (90.8) 6857 (87.1)
10,841 (12.0) 1416 (9.2) 1014 (12.9)
92,485 (88.6) 4038 (83.9) 3459 (86.6)
11,957 (11.5) 775 (16.1) 537 (13.4)
86,863 (88.3) 11,198 (88.1) 1935 (87.8)
11,495 (11.7) 1507 (11.9) 269 (12.2)
<.001
.65
<.001 11,735 11,969 12,668 12,834 12,638 12,892 12,384 12,876
(94.6) (94.5) (94.0) (93.0) (89.8) (86.0) (81.7) (77.2)
677 692 809 967 1440 2097 2780 3809
(5.4) (5.5) (6.0) (7.0) (10.2) (14.0) (18.3) (22.8) .15
49,874 (88.2) <50,045 (<88.4)
6674 (11.8) >6586 (>11.6) (continued on next page)
Volume 90 Number 5 2014
Trends in hypofractionated breast
997
Table 1 (continued ) Feature y
Bilateral/unknown Grade Well-differentiated Intermediate differentiation Poorly differentiated Undifferentiated Unknown Tumor size (mm) <10 10-29 30-49 50 Unknown Surgical margins No residual tumor Residual tumor Margins not evaluable ER-receptor Negative Positive Unknown Her2-result Negative Positive Unknown
Conventional fractionation (nZ99,996) >77 (>87.5)
Hypofractionation (nZ13,271) <11 (<12.5)
36,966 44,686 12,796 201 5347
(87.8) (88.2) (89.4) (90.1) (90.3)
5152 6006 1516 22 575
(12.2) (11.8) (10.6) (9.9) (9.7)
40,850 55,714 2681 328 423
(88.1) (88.4) (89.0) (88.7) (94.0)
5526 7343 333 42 27
(11.9) (11.6) (11.0) (11.3) (6.0)
P* <.001
.001
.77 96,245 (88.3) 3239 (88.0) 512 (87.7)
12,757 (11.7) 442 (12.0) 72 (12.3)
1286 (81.2) 23,438 (79.5) 75,272 (91.6)
296 (18.7) 6052 (20.5) 6923 (8.4)
23,510 (79.5) 1050 (82.2) 75,436 (91.5)
6071 (20.5) 227 (17.8) 6973 (8.5)
<.001
<.001
Abbreviations: ER Z estrogen receptor; RT Z radiation therapy. Values are number (percentage). * c2 test. y Actual values are suppressed owing to the American College of Surgeon’s prohibition against displaying cell sizes <11.
whereas the rate of patients undergoing conventional fractionation decreased from 53.55 per 100 patients to 44.11. Nonetheless, conventional fractionation predominated throughout the study period. We found that patients diagnosed at an academic treatment center were more likely to undergo hypofractionation. Among the 70,801 patients treated with adjuvant RT in nonacademic comprehensive community cancer programs, 10.3% received hypofractionation, whereas 17.3% of 28,137 RT-treated patients in academic centers received hypofractionation (odds ratio [OR] 0.51; 95% confidence interval [CI] 0.48-0.53) (Table 2). Of patients undergoing adjuvant RT at community cancer programs, only 7.7% underwent hypofractionation (OR 0.38 vs academic centers; 95% CI 0.35-0.42) (Fig. 1). Factors potentially associated with increasing difficulty of travel to and from radiation treatment were associated with hypofractionation, including age and distance to the cancer-reporting facility. For example, for patients >80 years of age, 20.0% of patients received hypofractionation, compared with 8.8% aged 50-59 years (Table 1) (OR 3.12; 95% CI 2.88-3.38) (Table 2). Patients living 50 miles away from the cancer-reporting facility were more likely to receive hypofractionation (OR 1.57; 95% CI 1.44-1.72). This remained true when analyzing
only patients who received radiation at the reporting facility (Table 3). Increased likelihood of hypofractionation was associated with residence in areas within the highest quartile of income (OR 1.25; 95% CI 1.16-1.35). Decreased likelihood of fractionation was associated with living in an urban area (OR 0.88; 95% CI 0.83-0.95). Tumor characteristics associated with more aggressive disease were associated with decreased odds of hypofractionation in bivariate and multivariate logistic regression. These characteristics included poor differentiation (OR 0.86 for poorly differentiated vs well-differentiated; 95% CI 0.80-0.91), increased size (OR 0.71 for 30-49 mm vs <10 mm; 95% CI 0.62-0.80), and ER-negativity (OR 1.15 for ER-positive vs ER-negative; 95% CI 1.00-1.32).
Discussion The use of hypofractionation has been steadily increasing over the past decade, in particular for patients treated in academic centers as well as for those for whom travel to daily treatment may be difficult or inconvenient. Despite this increase, a significant majority of patients who would be appropriate for hypofractionation were still receiving
998
International Journal of Radiation Oncology Biology Physics
Wang et al.
Table 2 Adjusted odds ratios for receipt of conventional fractionation versus hypofractionation Multivariate logistic regression Feature (reference)
Or (95% CI)
Facility type (academic) Community 0.38 Comprehensive community 0.51 Other 0.64 Facility location (South) Northeast 0.93 Midwest 1.04 West 1.69 Age in years (50-59) 60-69 1.22 70-79 1.77 80 3.12 Race (White non-Hispanic) Black 0.96 White Hispanic 1.23 Other 1.23 Primary payor (private insurance) No insurance 1.16 Medicaid 0.97 Medicare 1.05 Other government 0.81 Unknown 1.22 Median income quartile (<$30-000) $30,000-$35,000 0.95 $35,000-$45,999 0.97 $46,000 1.25 Unknown 1.04 Urbanization (Metro) Urban 0.88 Rural 1.10 Distance from cancer-reporting facility 50 miles 1.57 Unknown 1.13 Year of diagnosis (2004) 2005 1.03 2006 1.11 2007 1.31 2008 1.99 2009 2.88 2010 4.10 2011 5.48 Grade (well-differentiated) Intermediate differentiation 0.94 Poorly differentiated 0.86 Undifferentiated 1.03 Unknown 0.90 Tumor size (mm) (<10) 10-29 0.90 30-49 0.71 50 0.94 Unknown 0.63 ER-receptor (negative) Positive 1.15 Unknown 0.99
P
(0.35-0.42) (0.48-0.53) (0.52-0.78)
<.001 <.001 <.001
(0.88-0.99) (1.00-1.11) (1.59-1.78)
.02 .07 <.001
(1.16-1.29) (1.66-1.89) (2.88-3.38)
<.001 <.001 <.001
(0.89-1.05) (1.10-1.36) (1.13-1.34)
.41 <.001 <.001
(0.96-1.41) (0.86-1.10) (1.00-1.11) (0.63-1.05) (1.03-1.44)
.11 .71 .05 .12 .02
(0.87-1.04) (0.90-1.06) (1.16-1.35) (0.86-1.25)
.30 .54 <.001 .70
(0.83-0.95) (0.99-1.22) (<50 miles) (1.44-1.72) (0.91-1.40)
.001 .07 <.001 .27
(0.92-1.14) (1.00-1.24) (1.18-1.45) (1.80-2.19) (2.62-3.15) (3.62-4.64) (4.83-6.21)
.65 .05 <.001 <.001 <.001 <.001 <.001
(0.89-0.97) (0.80-0.91) (0.65-1.63) (0.82-0.99)
<.01 <.001 .89 .04
(0.87-0.94) (0.62-0.80) (0.67-1.32) (0.42-0.95)
<.001 <.001 .76 .03
(1.00-1.32) (0.82-1.19)
.05 .91
(continued)
Table 2
(continued ) Multivariate logistic regression
Feature (reference) Her2 result (negative) Positive Unknown
Or (95% CI)
P
0.86 (0.74-1.01) 1.16 (1.02-1.34)
.06 .03
Abbreviations: CI Z confidence interval; OR Z odds ratio. Other abbreviation as in Table 1.
conventional fractionation in 2011. In particular, many nonclinical factors were strongly related to receipt of hypofractionated therapy. Whether hypofractionation will become more thoroughly adopted in more recent years will be important to assess with future study, especially given the recently published evidence since 2011 (10, 11) and guidelines supporting hypofractionation (15, 23), potentially even as the preferred standard of care (13). The greater application of hypofractionation by academic practices compared with community practices may indicate a greater willingness to change practice patterns in light of new randomized evidence and published guidelines. This is juxtaposed against the greater adoption of less evidencebased, higher-cost technologies such as breast intensitymodulated RT and brachytherapy by free-standing facilities (24, 25). Though hospital-based centers are not necessarily academic, and vice versa, these trends are reflective of the nonuniform and complex pressures that drive current medical care in the United States. For example, in a referral-based practice like radiation oncology, a radiation oncologist’s reluctance to adopt breast brachytherapy could result not only in a more personally impactful loss of revenue (than academic/hospital-based practice) from that individual case, but also loss of a referral stream from the surgeon desirous of brachytherapy procedures. Academic and hospital-based practices may be more shielded from such pressure. Reflecting a concern for patient convenience, patients living a long distance from the cancer-reporting facilities were the most likely to undergo hypofractionation.
Fig. 1. Percentage of patients receiving hypofractionation by facility type.
Volume 90 Number 5 2014
Trends in hypofractionated breast
Table 3 Sensitivity analysis: all treatments at reporting facility (nZ11,913) Feature (reference)
Bivariate logistic regression Or (95% CI)
P
Multivariate logistic regression Or (95% CI)
P
Distance from cancer-reporting facility (<50 miles) 50 miles 2.75 (2.23-3.41) <.001 2.57 (2.04-3.25) <.001 Unknown 1.15 (0.9201.43) .22 0.86 (0.51-1.47) .60 Abbreviations as in Table 2.
Furthermore, patients who received hypofractionation were more likely to be elderly (defined as >80 years), possibly reflecting difficulty in travel, as well as a perception of lessaggressive disease by physicians. This may also be a reflection of clinicians’ reluctance to offer a protracted treatment that has been shown to be of small absolute benefit in elderly patients (26). A final nonclinical factor associated with increased likelihood of hypofractionation was residence in a high-income area. This is concerning because women living in lowerincome areas and thus conceivably having fewer financial resources may be disproportionately impacted by unnecessarily prolonged treatment. Better education may be needed, and improved communication may help in understanding these treatment preferences and alleviating patient concerns that may serve as barriers to providing optimal care. In addition to these nonclinical factors, there was lower likelihood of receiving hypofractionation with higher-risk disease. For example, we found that patients treated with hypofractionation had ER-positive, smaller tumors on average, with a higher proportion of well-differentiated tumors relative to their conventionally treated counterparts. This perhaps represents initial concerns from randomized trials regarding the comparative efficacy of hypofractionated treatment on high-grade disease, or a lack of robust long-term follow-up (4, 8). However, the suitability of hypofractionation for patients with grade 3 cancer originally questioned in early clinical trials has since then been refuted by evidence from other trials (27, 28). Whether rate of use of hypofractionation for high-risk breast cancer will change with further reporting of long-term results, such as from the United Kingdom START trials, which showed no difference in outcomes between low- and high-grade disease, will require further study (11). Our observed hypofractionation rates of 5 to 23% from 2004 through 2011 stand in marked contrast to reported rates from British Columbia, which ranged from 76% to 85% from 2003 through 2012, and from Ontario, which ranged from 66% to 69% from 2002 through 2008 (29-31). The investigators cite the fact that the British Columbia Cancer Agency’s long-standing commitment to consensus-based practice guidelines may explain the increased use of hypofractionation in their province, especially given the explicit recommendation by current British Columbia guidelines stating shorter fractionation to be the standard dose for whole-breast irradiation (32). The more centralized structure of the Canadian health care system may
999
also explain the large difference in hypofractionation rates observed in our study compared with those observed in Canada. Our study has several limitations to consider. First, there remain important clinical variables that we could not measure because of the use of an observational database. For example, we could not assess patient breast size, which can influence dose heterogeneity with tangent-based wholebreast irradiation, perhaps influencing the appropriateness of a patient for hypofractionation. We were also not able to directly assess possible regional differences in patient and provider preferences, rates of obesity, or reimbursement patterns, which could account for the geographic variations found in our study. Our results regarding the association of hormone receptor status and hypofractionation should be interpreted with caution owing to the large number of missing values in the database. There were also important provider-level variables that are not recorded in NCDB, such as for-profit facility status, years of practice, and board certification, which could influence provider attitudes toward hypofractionation. Finally, there was likely regionally variable underascertainment of RT after BCS in the NCDB, as has been shown in other national and state cancer registries (33). It is unclear what impact this underascertainment had on our analysis, though it is unlikely it impacted one form of fractionation significantly more than another, or significantly varied over time. Finally, our study represents the pattern of care as recently as 2011. However, since then, there have been several important developments that have further supported the routine use of hypofractionation. First, the UK START trials (11) reported their long-term follow-up indicating continued safety and efficacy of hypofractionated RT, adding further evidence in addition to the Canadian randomized trial (10). As a result, editorials (13) and guidelines (23) have further supported the use of hypofractionated RT. Over time these national guidelines and recommendations may alter community practice, where radiation oncologists may be less likely to adopt new strategies until they are more broadly embraced with national guidelines. In conclusion, we found increasing but incomplete adoption of hypofractionation and identified nonclinical factors that play an important role in hypofractionation utilizationdin particular whether a patient was treated at an academic facility and the distance from the patient’s residence to the cancer-reporting facility. Socioeconomic differences in hypofractionation utilization were also evident in our analysis. Future study is needed to determine whether recent evidence has increased the use of hypofractionation, and efforts should continue to encourage wider adoption of this efficacious and more cost-effective treatment.
References 1. Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: An overview of the randomised trials. Lancet 2005;366:2087-2106.
1000 Wang et al.
International Journal of Radiation Oncology Biology Physics
2. Overgaard M, Hansen PS, Overgaard J, et al. Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial. N Engl J Med 1997;337:949-955. 3. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med 2002;347:1233-1241. 4. Whelan T, MacKenzie R, Julian J, et al. Randomized trial of breast irradiation schedules after lumpectomy for women with lymph nodenegative breast cancer. J Natl Cancer Inst 2002;94:1143-1150. 5. Yarnold J, Ashton A, Bliss J, et al. Fractionation sensitivity and dose response of late adverse effects in the breast after radiotherapy for early breast cancer: Long-term results of a randomised trial. Radiother Oncol 2005;75:9-17. 6. Owen JR, Ashton A, Bliss JM, et al. Effect of radiotherapy fraction size on tumour control in patients with early-stage breast cancer after local tumour excision: Long-term results of a randomised trial. Lancet Oncol 2006;7:467-471. 7. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START) Trial A of radiotherapy hypofractionation for treatment of early breast cancer: A randomised trial. Lancet Oncol 2008;9:331-341. 8. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: A randomised trial. Lancet 2008;371:1098-1107. 9. Whelan TJ, Pignol J, Julian J, et al. Long-term results of a randomized trial of accelerated hypofractionated whole breast irradiation following breast conserving surgery in women with node-negative breast cancer. Int J Radiat Oncol Biol Phys 2008;72:S28. 10. Whelan TJ, Pignol JP, Levine MN, et al. Long-term results of hypofractionated radiation therapy for breast cancer. N Engl J Med 2010; 362:513-520. 11. Haviland JS, Owen JR, Dewar JA, et al. The UK Standardisation of Breast Radiotherapy (START) trials of radiotherapy hypofractionation for treatment of early breast cancer: 10-year follow-up results of two randomised controlled trials. Lancet Oncol 2013;14:1086-1094. 12. Suh WW, Pierce LJ, Vicini FA, et al. A cost comparison analysis of partial versus whole-breast irradiation after breast-conserving surgery for early-stage breast cancer. Int J Radiat Oncol Biol Phys 2005;62: 790-796. 13. Haffty BG, Buchholz TA. Hypofractionated breast radiation: Preferred standard of care? Lancet Oncol 2013;14:1032-1034. 14. Izano M, Satariano WA, Tammemagi MC, et al. Long-term outcomes among African-American and white women with breast cancer: What is the impact of comorbidity? J Geriatr Oncol 2014 Mar 5. pii: S18794068(14)00039-3. 15. Smith BD, Bentzen SM, Correa CR, et al. Fractionation for whole breast irradiation: An American Society for Radiation Oncology (ASTRO) evidence-based guideline. Int J Radiat Oncol Biol Phys 2011;81:59-68.
16. Hoopes DJ, Kaziska D, Chapin P, et al. Patient preferences and physician practice patterns regarding breast radiotherapy. Int J Radiat Oncol Biol Phys 2012;82:674-681. 17. American College of Surgeons. National Cancer Data Base. Available at: http://www.facs.org/cancer/ncdb/index.html. Accessed May 3, 2014. 18. Menck HR, Cunningham MP, Jessup JM, et al. The growth and maturation of the National Cancer Data Base. Cancer 1997;80:2296-2304. 19. Partridge EE. The National Cancer Data Base: Ten years of growth and commitment. CA Cancer J Clin 1998;48:131-133. 20. American College of Surgeons. National Cancer Data Base PUF. Available at: http://ncdbpufbeta.facs.org/. Accessed May 3, 2014. 21. Onega T, Duell EJ, Shi X, et al. Geographic access to cancer care in the U.S. Cancer 2008;112:909-918. 22. Massarweh NN, Chiang YJ, Xing Y, et al. Association between travel distance and metastatic disease at diagnosis among patients with colon cancer. J Clin Oncol 2014;32:942-948. 23. Volpp KG, Loewenstein G, Asch DA. Choosing wisely: Low-value services, utilization, and patient cost sharing. JAMA 2012;308:1635-1636. 24. Roberts KB, Soulos PR, Herrin J, et al. The adoption of new adjuvant radiation therapy modalities among Medicare beneficiaries with breast cancer: Clinical correlates and cost implications. Int J Radiat Oncol Biol Phys 2013;85:1186-1192. 25. Sen S, Soulos PR, Herrin J, et al. For-profit hospital ownership status and use of brachytherapy after breast-conserving surgery. Surgery 2014;155:776-788. 26. Hughes KS, Schnaper LA, Bellon JR, et al. Lumpectomy plus tamoxifen with or without irradiation in women age 70 years or older with early breast cancer: Long-term follow-up of CALGB 9343. J Clin Oncol 2013;31:2382-2387. 27. Haviland JS, Yarnold JR, Bentzen SM. Hypofractionated radiotherapy for breast cancer. N Engl J Med 2010;362:1843. author reply 1843e1844. 28. Herbert C, Nichol A, Olivotto I, et al. The impact of hypofractionated whole breast radiotherapy on local relapse in patients with Grade 3 early breast cancer: A population-based cohort study. Int J Radiat Oncol Biol Phys 2012;82:2086-2092. 29. Ashworth A, Kong W, Whelan T, et al. A population-based study of the fractionation of postlumpectomy breast radiation therapy. Int J Radiat Oncol Biol Phys 2013;86:51-57. 30. Berrang TS, Truong PT, Tyldesley S, et al. A population-based study of the fractionation of postlumpectomy breast radiation therapy. In regard to Ashworth, et al. Int J Radiat Oncol Biol Phys 2013;87:632-633. 31. Ashworth A, Kong W, Whelan T, et al. In Reply to Berrang, et al. Int J Radiat Oncol Biol Phys 2013;87:633. 32. Stage I and II Invasive Breast Cancer RT Section: BC Cancer Agency Cancer Management Guidelines. Available at: http://www.bccancer. bc.ca/HPI/CancerManagementGuidelines/Breast/Management/15. þStageþIþandþIIþInvasiveþBreastþCancerþRTþSection.htm. Accessed May 3, 2014. 33. Jagsi R, Abrahamse P, Hawley ST, et al. Underascertainment of radiotherapy receipt in Surveillance, Epidemiology, and End Results registry data. Cancer 2012;118:333-341.
Radiation Oncology biology
physics
www.redjournal.org
Clinical Investigation
Adoption of Hypofractionated Whole-Breast Irradiation for Early-Stage Breast Cancer: A National Cancer Data Base Analysis Elyn H. Wang, BS,* Sarah S. Mougalian, MD,*,y,z Pamela R. Soulos, MPH,*,z Charles E. Rutter, MD,*,z,x Suzanne B. Evans, MD, MPH,*,z,x Bruce G. Haffty, MD,jj Cary P. Gross, MD,*,jj,{ and James B. Yu, MD*,z,x *Yale School of Medicine, Yale School of Medicine, New Haven, Connecticut; yDepartment of Medical Oncology, Yale School of Medicine, New Haven, Connecticut; zCancer Outcomes, Public Policy, and Effectiveness Research Center at Yale, New Haven, Connecticut; xDepartment of Therapeutic Radiology, Yale School of Medicine, New Haven, Connecticut; jjDepartment of Radiation Oncology, Rutgers Cancer Institute of New Jersey and Robert Wood Johnson Medical School, New Brunswick, New Jersey; and {Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut Received May 27, 2014, and in revised form Jun 12, 2014. Accepted for publication Jun 14, 2014.
Summary Hypofractionated wholebreast radiation therapy (hypofractionation) for early-stage breast cancer is supported by high-quality randomized trial evidence and clinical guidelines. We performed a retrospective study of breast cancer
Purpose: To evaluate the relationship of patient, hospital, and cancer characteristics with the adoption of hypofractionation in a national sample of patients diagnosed with early-stage breast cancer. Methods and Materials: We performed a retrospective study of breast cancer patients in the National Cancer Data Base from 2004-2011 who were treated with radiation therapy and met eligibility criteria for hypofractionation. We used logistic regression to identify factors associated with receipt of hypofractionation (vs conventional fractionation). Results: We identified 13,271 women (11.7%) and 99,996 women (88.3%) with earlystage breast cancer who were treated with hypofractionation and conventional fractionation, respectively. The use of hypofractionation increased significantly, with
Reprint requests to: James B. Yu, MD, Department of Therapeutic Radiology, Yale School of Medicine, HRT 138, 333 Cedar St, New Haven, CT 06520. Tel: (203) 785-5703; E-mail: [email protected] This work was supported by the National Institutes of Health (NIH; grant R01CA149045). J.B.Y. is also supported by Clinical Translational Science Awards grant KL2 TR000140 from the National Center for Advancing Translational Sciences, a component of the NIH, and NIH Roadmap for Medical Research. J.B.Y. and C.P.G. have research funding from 21st Century Oncology, LLC. C.P.G. also has research funding from Medtronic and Johnson & Johnson. The National Cancer Data Base is a joint project of the Commission on Cancer (CoC) of the American College of Surgeons and the American Cancer Society. The American College of Surgeons has executed a Int J Radiation Oncol Biol Phys, Vol. 90, No. 5, pp. 993e1000, 2014 0360-3016/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ijrobp.2014.06.038
Business Associate Agreement that includes a data use agreement with each of its CoC-accredited hospitals. The data used in the study are derived from a deidentified National Cancer Data Base file. The American College of Surgeons and the CoC have not verified and are not responsible for the analytic or statistical methodology used, or the conclusions drawn from these data by the investigator. The study sponsor (NIH) did not play a role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the manuscript for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Conflict of interest: J.B.Y. and C.P.G. have research funding from 21st Century Oncology, LLC.
994
International Journal of Radiation Oncology Biology Physics
Wang et al.
patients in the National Cancer Data Base from 2004-2011 who were treated with radiation therapy and met eligibility criteria for hypofractionation. We found that the use of hypofractionation is rising and is associated with increased travel distance and treatment at an academic center.
5.4% of patients receiving it in 2004 compared with 22.8% in 2011 (P<.001 for trend). Patients living 50 miles from the cancer reporting facility had increased odds of receiving hypofractionation (odds ratio 1.57 [95% confidence interval 1.44-1.72], P<.001). Adoption of hypofractionation was associated with treatment at an academic center (P<.001) and living in an area with high median income (P<.001). Hypofractionation was less likely to be used in patients with high-risk disease, such as increased tumor size (P<.001) or poorly differentiated histologic grade (P<.001). Conclusions: The use of hypofractionation is rising and is associated with increased travel distance and treatment at an academic center. Further adoption of hypofractionation may be tempered by both clinical and nonclinical concerns. Ó 2014 Elsevier Inc.
Introduction Randomized clinical trials in patients with early-stage breast cancer treated with breast-conserving surgery (BCS) have demonstrated that adjuvant whole-breast irradiation lowers the risk of tumor recurrence and improves survival (1). Historically, most studies have focused on the benefits of conventionally fractionated whole-breast irradiation (conventional fractionation), which targets the entire breast with 1.80-2.00 Gy once daily for a total dose of 45-50 Gy in 25-28 daily fractions over 5 to 6 weeks (2, 3). In modern practice, hypofractionated whole-breast irradiation (hypofractionation) has emerged as a viable alternative to conventional fractionation. Hypofractionation treats the entire breast with 2.5-3.20 Gy once daily for a total dose of 39-42.5 Gy in 13-16 fractions for 3-5 weeks (4-11). Depending on the scheme of choice, radiation can be delivered every day (4, 8-10) or delivered in 13 fractions over 35 days (7). Compared with conventional fractionation, hypofractionation involves a shorter treatment course, decreased total dose and number of fractions, and results in increased patient convenience and lower costs (12). In the past several years, randomized trials have tested hypofractionation against conventional fractionation after BCS for early-stage breast cancer patients (4, 7, 8, 10). These trials, originally reported in 2005, have shown that hypofractionation produces disease-free survival and overall survival outcomes that are equivalent to those for conventional fractionation (5-10, 13, 14) in properly selected patients. Furthermore, recently reported 10-year follow-up of these trials has demonstrated that the long-term toxicity profiles for both methods are comparable (10, 13, 14). As a result, clinical practice guidelines from the American Society for Radiation Oncology were recently published establishing hypofractionation as an appropriate therapeutic option for most patients with early breast cancer (15), and the American Society for Radiation Oncology has identified hypofractionation as one of its 5 Choosing Wisely recommendations, a national campaign focused on encouraging conversations between patients and providers about evidence-based treatment choices (13). Despite the strong evidence and guidelines supporting its use, there is some indication that physicians may be
slow to adopt hypofractionation: A recent survey study of physician practice patterns revealed that 56.2% of physicians do not use hypofractionation at all in the post-BCS setting, in conflict with patient preference (16). It is possible that physicians’ clinical concerns (that larger hypofractionated doses of radiation therapy [RT] per treatment are related to greater late side effects) are limiting usageddespite a lack of difference in toxicity seen in randomized trials with long follow-up (10). Physicians may also be concerned with evidence suggesting that hypofractionation for patients with high-grade disease is associated with inferior outcomes (10), although this was not seen in subsequent trials (8). Additionally, nonclinical factors such as financial pressure could also be impeding the adoption of hypofractionation, given that a less costly treatment also means less reimbursement to the physician and hospital. Given the clinical evidence, improved convenience, and decreased costs, understanding the adoption of hypofractionation is critical as radiation-related cancer costs are targeted as an area of potential health savings. We therefore evaluated the relationship of patient, hospital, and cancer characteristics with the adoption of hypofractionation in a national sample of patients diagnosed with early-stage breast cancer. We hypothesized that the use of hypofractionation has increased, particularly in patients for whom travel to and from radiation treatment would be difficult, owing either to distance, age, or comorbidity. We also hypothesized that factors that were indicative of less physician financial pressure (such as academic practice location) and lower-risk disease were associated with adoption of hypofractionation.
Methods and Materials Data source and study population We used National Cancer Data Base (NCDB) registry data from 2004 to 2011 to examine receipt of adjuvant wholebreast irradiation in patients eligible to receive both hypofractionated and conventional fractionation. The NCDB is a joint program of the Commission on Cancer of the
Volume 90 Number 5 2014
American College of Surgeons and the American Cancer Society (17-19). The NCDB, established in 1989, is a nationwide, facility-based, clinical surveillance data set that currently captures 70% of all newly diagnosed malignancies in the United States annually. Data reported to the NCDB are retrospective, and no patient or physician identifiers are collected (20). This study was granted an exemption from our institution’s Human Investigation Committee, because it was research involving existing data on unidentifiable research subjects. Patients with a first and only diagnosis of breast cancer whose data were entered in the NCDB from 2004 through 2011 were the subject of this study. To better evaluate factors independent of hypofractionation eligibility, we restricted the sample to early-stage female breast cancer patients who received adjuvant RT and who met the criteria for hypofractionation receipt according to guidelines and randomized trials. These criteria included (1) age 50 years or older at time of diagnosis; (2) pathologic stage T1-2 N0 according to the American Joint Committee on Cancer TNM staging classification; (3) treatment with BCS; and (4) no chemotherapy treatment (15). We limited our analysis to adjuvant beam radiation delivered to the intact breast as the primary target. We did not include patients who received regional nodal RT as reported by the NCDB. Brachytherapy, stereotactic RT, and treatments delivered using electron, neutron, or proton beams were excluded. Total dose of regional RT was limited to 40-66.4 Gy, and boost doses were limited to <21.6 Gy. Patients who received fewer than 10 or more than 50 fractions and RT treatments coded as palliative care were further excluded from the final sample.
Study covariates and primary outcome Patient demographic and clinical information included as independent variables were age, race, type of primary health insurance, median income in zip code of residence by quartile, degree of rurality, and urban influence of patient’s residence using classification published by the US Department of Agriculture Economic Research Service, and the Charlson/Deyo comorbidity score. We also evaluated facility type (academic vs nonacademic) and geographic region (Northeast, Midwest, South, West). We created an indicator for whether patients had to travel a long distance to receive cancer care. Because the distance from the patients’ residence to the radiation treatment facility was not specifically recorded, we used distance to the cancer reporting facility as a proxy for distance travelled. We classified 50 miles as a long distance (21, 22). To check the validity of this assumption, analysis of distance traveled was performed for the subset of patients for whom the cancer reporting facility and radiation treatment facility were the same (10.5% of the sample). Cancer characteristics assessed included year of cancer diagnosis, tumor differentiation, size, estrogen receptor (ER) and human
Trends in hypofractionated breast
995
epidermal growth factor receptor 2 (HER2) assay results, and surgical margins. Unknown or missing values were entered as dummy variables of a separate category. The primary outcome was type of adjuvant RT received (hypofractionation vs conventional fractionation). Fraction dose was defined as total dose divided by the total number of treatment sessions administered during the first course of treatment. To take into account small inaccuracies in the recorded dose and number of fractions delivered, hypofractionation was broadly defined as fraction dose 2.2 and 4.0 Gy, and patients were required to have received at least 40 Gy. Conventional fractionation was broadly defined as fraction dose between 1.5 and 2.2 Gy.
Statistical analysis Descriptive statistics were used to summarize patient, facility, and cancer characteristics. Bivariate associations of patient, facility, and cancer variables were tested by the Pearson’s c2 test. Covariates associated with hypofractionation at P.10 were then included in a multivariable logistic regression model. A two-sided P value of <.05 was used to determine statistical significance. Stata SE version 13.0 (Stata Corp, College Station, TX) was used to perform all statistical analyses.
Results The total number of female patients with a first and only diagnosis of breast cancer reported in the NCDB increased from 139,300 patients in 2004 to 165,125 patients in 2009 and then decreased to 164,760 patients in 2011. This fluctuation was likely due to the changing number of facilities reporting cases to the NCDB, which also ranged from a low of 1298 in 2011 to a peak of 1329 in 2009. The number of early-stage breast cancer cases reported followed a similar trend as total breast cancer cases, with 57,984 patients in 2004 and a peak of 64,854 in 2009, followed by a decline in subsequent years. Despite this fluctuation, the overall number of early-stage breast cancer patients who received adjuvant RT (hypofractionation or conventional) who met our study’s criteria for hypofractionation increased steadily from 12,412 patients in 2004 to 16,685 patients in 2011, an increase of 34.4%. We identified 13,271 (11.7%) and 99,996 (88.3%) female early-stage breast cancer patients >50 years of age not receiving chemotherapy who were treated between 2004 and 2011 with hypofractionation and conventional fractionation (respectively) for whole-breast irradiation after BCS. The number of patients who met our study’s criteria for and underwent hypofractionation increased steadily from 677 patients (5.4% of patients undergoing whole-breast irradiation) in 2004 to 3809 patients (22.8%) in 2011 (P for trend, <.001) (Table 1). Furthermore, the rate of patients who underwent hypofractionation increased from 3.09 per 100 patients in 2004 to 13.05 in 2011,
996
International Journal of Radiation Oncology Biology Physics
Wang et al.
Table 1
Patient, facility, and cancer characteristics by RT modality Feature
Facility type Academic Community Comprehensive community Other Facility location South Northeast Midwest West Age in years 50-59 60-69 70-79 80 Race White non-Hispanic Black White Hispanic Other Primary payor Private insurance No insurance Medicaid Medicare Other government Unknown Median income quartile <$30,000 $30,000-$35,000 $35,000-$45,999 $46,000 Unknown Urbanization Metro Urban Rural Distance from cancer reporting facility <50 miles 50 miles Unknown Charlson-Deyo score 0 1 2 Year of diagnosis 2004 2005 2006 2007 2008 2009 2010 2011 Laterality Right-sided Left-sided
Conventional fractionation (nZ99,996)
Hypofractionation (nZ13,271)
P* <.001
23,307 12,156 63,488 1045
(82.8) (92.3) (89.7) (90.5)
4830 1018 7313 110
(17.3) (7.7) (10.3) (9.5)
29,634 23,848 29,060 17,454
(90.0) (88.7) (88.6) (84.5)
3279 3038 3747 3207
(10.0) (11.3) (11.4) (15.5)
29,959 37,369 25,366 7302
(91.2) (89.2) (86.4) (80.0)
2896 4544 4006 1825
(8.8) (10.8) (13.6) (20.0)
86,670 6715 2741 3870
(88.6) (89.0) (85.8) (83.3)
11,210 828 455 778
(11.5) (11.0) (14.2) (16.7)
49,153 983 2879 45,121 713 1147
(89.9) (88.2) (89.1) (86.6) (91.1) (86.6)
5545 131 354 6993 70 178
(10.1) (11.8) (10.9) (13.4) (8.9) (13.4)
8912 15,600 27,490 43,402 4592
(89.6) (90.1) (89.8) (86.6) (87.2)
1038 1712 3115 6729 677
(10.4) (9.9) (10.2) (13.4) (12.8)
<.001
<.001
<.001
<.001
<.001
<.001 79,523 (88.0) 13,886 (90.8) 6857 (87.1)
10,841 (12.0) 1416 (9.2) 1014 (12.9)
92,485 (88.6) 4038 (83.9) 3459 (86.6)
11,957 (11.5) 775 (16.1) 537 (13.4)
86,863 (88.3) 11,198 (88.1) 1935 (87.8)
11,495 (11.7) 1507 (11.9) 269 (12.2)
<.001
.65
<.001 11,735 11,969 12,668 12,834 12,638 12,892 12,384 12,876
(94.6) (94.5) (94.0) (93.0) (89.8) (86.0) (81.7) (77.2)
677 692 809 967 1440 2097 2780 3809
(5.4) (5.5) (6.0) (7.0) (10.2) (14.0) (18.3) (22.8) .15
49,874 (88.2) <50,045 (<88.4)
6674 (11.8) >6586 (>11.6) (continued on next page)
Volume 90 Number 5 2014
Trends in hypofractionated breast
997
Table 1 (continued ) Feature y
Bilateral/unknown Grade Well-differentiated Intermediate differentiation Poorly differentiated Undifferentiated Unknown Tumor size (mm) <10 10-29 30-49 50 Unknown Surgical margins No residual tumor Residual tumor Margins not evaluable ER-receptor Negative Positive Unknown Her2-result Negative Positive Unknown
Conventional fractionation (nZ99,996) >77 (>87.5)
Hypofractionation (nZ13,271) <11 (<12.5)
36,966 44,686 12,796 201 5347
(87.8) (88.2) (89.4) (90.1) (90.3)
5152 6006 1516 22 575
(12.2) (11.8) (10.6) (9.9) (9.7)
40,850 55,714 2681 328 423
(88.1) (88.4) (89.0) (88.7) (94.0)
5526 7343 333 42 27
(11.9) (11.6) (11.0) (11.3) (6.0)
P* <.001
.001
.77 96,245 (88.3) 3239 (88.0) 512 (87.7)
12,757 (11.7) 442 (12.0) 72 (12.3)
1286 (81.2) 23,438 (79.5) 75,272 (91.6)
296 (18.7) 6052 (20.5) 6923 (8.4)
23,510 (79.5) 1050 (82.2) 75,436 (91.5)
6071 (20.5) 227 (17.8) 6973 (8.5)
<.001
<.001
Abbreviations: ER Z estrogen receptor; RT Z radiation therapy. Values are number (percentage). * c2 test. y Actual values are suppressed owing to the American College of Surgeon’s prohibition against displaying cell sizes <11.
whereas the rate of patients undergoing conventional fractionation decreased from 53.55 per 100 patients to 44.11. Nonetheless, conventional fractionation predominated throughout the study period. We found that patients diagnosed at an academic treatment center were more likely to undergo hypofractionation. Among the 70,801 patients treated with adjuvant RT in nonacademic comprehensive community cancer programs, 10.3% received hypofractionation, whereas 17.3% of 28,137 RT-treated patients in academic centers received hypofractionation (odds ratio [OR] 0.51; 95% confidence interval [CI] 0.48-0.53) (Table 2). Of patients undergoing adjuvant RT at community cancer programs, only 7.7% underwent hypofractionation (OR 0.38 vs academic centers; 95% CI 0.35-0.42) (Fig. 1). Factors potentially associated with increasing difficulty of travel to and from radiation treatment were associated with hypofractionation, including age and distance to the cancer-reporting facility. For example, for patients >80 years of age, 20.0% of patients received hypofractionation, compared with 8.8% aged 50-59 years (Table 1) (OR 3.12; 95% CI 2.88-3.38) (Table 2). Patients living 50 miles away from the cancer-reporting facility were more likely to receive hypofractionation (OR 1.57; 95% CI 1.44-1.72). This remained true when analyzing
only patients who received radiation at the reporting facility (Table 3). Increased likelihood of hypofractionation was associated with residence in areas within the highest quartile of income (OR 1.25; 95% CI 1.16-1.35). Decreased likelihood of fractionation was associated with living in an urban area (OR 0.88; 95% CI 0.83-0.95). Tumor characteristics associated with more aggressive disease were associated with decreased odds of hypofractionation in bivariate and multivariate logistic regression. These characteristics included poor differentiation (OR 0.86 for poorly differentiated vs well-differentiated; 95% CI 0.80-0.91), increased size (OR 0.71 for 30-49 mm vs <10 mm; 95% CI 0.62-0.80), and ER-negativity (OR 1.15 for ER-positive vs ER-negative; 95% CI 1.00-1.32).
Discussion The use of hypofractionation has been steadily increasing over the past decade, in particular for patients treated in academic centers as well as for those for whom travel to daily treatment may be difficult or inconvenient. Despite this increase, a significant majority of patients who would be appropriate for hypofractionation were still receiving
998
International Journal of Radiation Oncology Biology Physics
Wang et al.
Table 2 Adjusted odds ratios for receipt of conventional fractionation versus hypofractionation Multivariate logistic regression Feature (reference)
Or (95% CI)
Facility type (academic) Community 0.38 Comprehensive community 0.51 Other 0.64 Facility location (South) Northeast 0.93 Midwest 1.04 West 1.69 Age in years (50-59) 60-69 1.22 70-79 1.77 80 3.12 Race (White non-Hispanic) Black 0.96 White Hispanic 1.23 Other 1.23 Primary payor (private insurance) No insurance 1.16 Medicaid 0.97 Medicare 1.05 Other government 0.81 Unknown 1.22 Median income quartile (<$30-000) $30,000-$35,000 0.95 $35,000-$45,999 0.97 $46,000 1.25 Unknown 1.04 Urbanization (Metro) Urban 0.88 Rural 1.10 Distance from cancer-reporting facility 50 miles 1.57 Unknown 1.13 Year of diagnosis (2004) 2005 1.03 2006 1.11 2007 1.31 2008 1.99 2009 2.88 2010 4.10 2011 5.48 Grade (well-differentiated) Intermediate differentiation 0.94 Poorly differentiated 0.86 Undifferentiated 1.03 Unknown 0.90 Tumor size (mm) (<10) 10-29 0.90 30-49 0.71 50 0.94 Unknown 0.63 ER-receptor (negative) Positive 1.15 Unknown 0.99
P
(0.35-0.42) (0.48-0.53) (0.52-0.78)
<.001 <.001 <.001
(0.88-0.99) (1.00-1.11) (1.59-1.78)
.02 .07 <.001
(1.16-1.29) (1.66-1.89) (2.88-3.38)
<.001 <.001 <.001
(0.89-1.05) (1.10-1.36) (1.13-1.34)
.41 <.001 <.001
(0.96-1.41) (0.86-1.10) (1.00-1.11) (0.63-1.05) (1.03-1.44)
.11 .71 .05 .12 .02
(0.87-1.04) (0.90-1.06) (1.16-1.35) (0.86-1.25)
.30 .54 <.001 .70
(0.83-0.95) (0.99-1.22) (<50 miles) (1.44-1.72) (0.91-1.40)
.001 .07 <.001 .27
(0.92-1.14) (1.00-1.24) (1.18-1.45) (1.80-2.19) (2.62-3.15) (3.62-4.64) (4.83-6.21)
.65 .05 <.001 <.001 <.001 <.001 <.001
(0.89-0.97) (0.80-0.91) (0.65-1.63) (0.82-0.99)
<.01 <.001 .89 .04
(0.87-0.94) (0.62-0.80) (0.67-1.32) (0.42-0.95)
<.001 <.001 .76 .03
(1.00-1.32) (0.82-1.19)
.05 .91
(continued)
Table 2
(continued ) Multivariate logistic regression
Feature (reference) Her2 result (negative) Positive Unknown
Or (95% CI)
P
0.86 (0.74-1.01) 1.16 (1.02-1.34)
.06 .03
Abbreviations: CI Z confidence interval; OR Z odds ratio. Other abbreviation as in Table 1.
conventional fractionation in 2011. In particular, many nonclinical factors were strongly related to receipt of hypofractionated therapy. Whether hypofractionation will become more thoroughly adopted in more recent years will be important to assess with future study, especially given the recently published evidence since 2011 (10, 11) and guidelines supporting hypofractionation (15, 23), potentially even as the preferred standard of care (13). The greater application of hypofractionation by academic practices compared with community practices may indicate a greater willingness to change practice patterns in light of new randomized evidence and published guidelines. This is juxtaposed against the greater adoption of less evidencebased, higher-cost technologies such as breast intensitymodulated RT and brachytherapy by free-standing facilities (24, 25). Though hospital-based centers are not necessarily academic, and vice versa, these trends are reflective of the nonuniform and complex pressures that drive current medical care in the United States. For example, in a referral-based practice like radiation oncology, a radiation oncologist’s reluctance to adopt breast brachytherapy could result not only in a more personally impactful loss of revenue (than academic/hospital-based practice) from that individual case, but also loss of a referral stream from the surgeon desirous of brachytherapy procedures. Academic and hospital-based practices may be more shielded from such pressure. Reflecting a concern for patient convenience, patients living a long distance from the cancer-reporting facilities were the most likely to undergo hypofractionation.
Fig. 1. Percentage of patients receiving hypofractionation by facility type.
Volume 90 Number 5 2014
Trends in hypofractionated breast
Table 3 Sensitivity analysis: all treatments at reporting facility (nZ11,913) Feature (reference)
Bivariate logistic regression Or (95% CI)
P
Multivariate logistic regression Or (95% CI)
P
Distance from cancer-reporting facility (<50 miles) 50 miles 2.75 (2.23-3.41) <.001 2.57 (2.04-3.25) <.001 Unknown 1.15 (0.9201.43) .22 0.86 (0.51-1.47) .60 Abbreviations as in Table 2.
Furthermore, patients who received hypofractionation were more likely to be elderly (defined as >80 years), possibly reflecting difficulty in travel, as well as a perception of lessaggressive disease by physicians. This may also be a reflection of clinicians’ reluctance to offer a protracted treatment that has been shown to be of small absolute benefit in elderly patients (26). A final nonclinical factor associated with increased likelihood of hypofractionation was residence in a high-income area. This is concerning because women living in lowerincome areas and thus conceivably having fewer financial resources may be disproportionately impacted by unnecessarily prolonged treatment. Better education may be needed, and improved communication may help in understanding these treatment preferences and alleviating patient concerns that may serve as barriers to providing optimal care. In addition to these nonclinical factors, there was lower likelihood of receiving hypofractionation with higher-risk disease. For example, we found that patients treated with hypofractionation had ER-positive, smaller tumors on average, with a higher proportion of well-differentiated tumors relative to their conventionally treated counterparts. This perhaps represents initial concerns from randomized trials regarding the comparative efficacy of hypofractionated treatment on high-grade disease, or a lack of robust long-term follow-up (4, 8). However, the suitability of hypofractionation for patients with grade 3 cancer originally questioned in early clinical trials has since then been refuted by evidence from other trials (27, 28). Whether rate of use of hypofractionation for high-risk breast cancer will change with further reporting of long-term results, such as from the United Kingdom START trials, which showed no difference in outcomes between low- and high-grade disease, will require further study (11). Our observed hypofractionation rates of 5 to 23% from 2004 through 2011 stand in marked contrast to reported rates from British Columbia, which ranged from 76% to 85% from 2003 through 2012, and from Ontario, which ranged from 66% to 69% from 2002 through 2008 (29-31). The investigators cite the fact that the British Columbia Cancer Agency’s long-standing commitment to consensus-based practice guidelines may explain the increased use of hypofractionation in their province, especially given the explicit recommendation by current British Columbia guidelines stating shorter fractionation to be the standard dose for whole-breast irradiation (32). The more centralized structure of the Canadian health care system may
999
also explain the large difference in hypofractionation rates observed in our study compared with those observed in Canada. Our study has several limitations to consider. First, there remain important clinical variables that we could not measure because of the use of an observational database. For example, we could not assess patient breast size, which can influence dose heterogeneity with tangent-based wholebreast irradiation, perhaps influencing the appropriateness of a patient for hypofractionation. We were also not able to directly assess possible regional differences in patient and provider preferences, rates of obesity, or reimbursement patterns, which could account for the geographic variations found in our study. Our results regarding the association of hormone receptor status and hypofractionation should be interpreted with caution owing to the large number of missing values in the database. There were also important provider-level variables that are not recorded in NCDB, such as for-profit facility status, years of practice, and board certification, which could influence provider attitudes toward hypofractionation. Finally, there was likely regionally variable underascertainment of RT after BCS in the NCDB, as has been shown in other national and state cancer registries (33). It is unclear what impact this underascertainment had on our analysis, though it is unlikely it impacted one form of fractionation significantly more than another, or significantly varied over time. Finally, our study represents the pattern of care as recently as 2011. However, since then, there have been several important developments that have further supported the routine use of hypofractionation. First, the UK START trials (11) reported their long-term follow-up indicating continued safety and efficacy of hypofractionated RT, adding further evidence in addition to the Canadian randomized trial (10). As a result, editorials (13) and guidelines (23) have further supported the use of hypofractionated RT. Over time these national guidelines and recommendations may alter community practice, where radiation oncologists may be less likely to adopt new strategies until they are more broadly embraced with national guidelines. In conclusion, we found increasing but incomplete adoption of hypofractionation and identified nonclinical factors that play an important role in hypofractionation utilizationdin particular whether a patient was treated at an academic facility and the distance from the patient’s residence to the cancer-reporting facility. Socioeconomic differences in hypofractionation utilization were also evident in our analysis. Future study is needed to determine whether recent evidence has increased the use of hypofractionation, and efforts should continue to encourage wider adoption of this efficacious and more cost-effective treatment.
References 1. Clarke M, Collins R, Darby S, et al. Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: An overview of the randomised trials. Lancet 2005;366:2087-2106.
1000 Wang et al.
International Journal of Radiation Oncology Biology Physics
2. Overgaard M, Hansen PS, Overgaard J, et al. Postoperative radiotherapy in high-risk premenopausal women with breast cancer who receive adjuvant chemotherapy. Danish Breast Cancer Cooperative Group 82b Trial. N Engl J Med 1997;337:949-955. 3. Fisher B, Anderson S, Bryant J, et al. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. N Engl J Med 2002;347:1233-1241. 4. Whelan T, MacKenzie R, Julian J, et al. Randomized trial of breast irradiation schedules after lumpectomy for women with lymph nodenegative breast cancer. J Natl Cancer Inst 2002;94:1143-1150. 5. Yarnold J, Ashton A, Bliss J, et al. Fractionation sensitivity and dose response of late adverse effects in the breast after radiotherapy for early breast cancer: Long-term results of a randomised trial. Radiother Oncol 2005;75:9-17. 6. Owen JR, Ashton A, Bliss JM, et al. Effect of radiotherapy fraction size on tumour control in patients with early-stage breast cancer after local tumour excision: Long-term results of a randomised trial. Lancet Oncol 2006;7:467-471. 7. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START) Trial A of radiotherapy hypofractionation for treatment of early breast cancer: A randomised trial. Lancet Oncol 2008;9:331-341. 8. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: A randomised trial. Lancet 2008;371:1098-1107. 9. Whelan TJ, Pignol J, Julian J, et al. Long-term results of a randomized trial of accelerated hypofractionated whole breast irradiation following breast conserving surgery in women with node-negative breast cancer. Int J Radiat Oncol Biol Phys 2008;72:S28. 10. Whelan TJ, Pignol JP, Levine MN, et al. Long-term results of hypofractionated radiation therapy for breast cancer. N Engl J Med 2010; 362:513-520. 11. Haviland JS, Owen JR, Dewar JA, et al. The UK Standardisation of Breast Radiotherapy (START) trials of radiotherapy hypofractionation for treatment of early breast cancer: 10-year follow-up results of two randomised controlled trials. Lancet Oncol 2013;14:1086-1094. 12. Suh WW, Pierce LJ, Vicini FA, et al. A cost comparison analysis of partial versus whole-breast irradiation after breast-conserving surgery for early-stage breast cancer. Int J Radiat Oncol Biol Phys 2005;62: 790-796. 13. Haffty BG, Buchholz TA. Hypofractionated breast radiation: Preferred standard of care? Lancet Oncol 2013;14:1032-1034. 14. Izano M, Satariano WA, Tammemagi MC, et al. Long-term outcomes among African-American and white women with breast cancer: What is the impact of comorbidity? J Geriatr Oncol 2014 Mar 5. pii: S18794068(14)00039-3. 15. Smith BD, Bentzen SM, Correa CR, et al. Fractionation for whole breast irradiation: An American Society for Radiation Oncology (ASTRO) evidence-based guideline. Int J Radiat Oncol Biol Phys 2011;81:59-68.
16. Hoopes DJ, Kaziska D, Chapin P, et al. Patient preferences and physician practice patterns regarding breast radiotherapy. Int J Radiat Oncol Biol Phys 2012;82:674-681. 17. American College of Surgeons. National Cancer Data Base. Available at: http://www.facs.org/cancer/ncdb/index.html. Accessed May 3, 2014. 18. Menck HR, Cunningham MP, Jessup JM, et al. The growth and maturation of the National Cancer Data Base. Cancer 1997;80:2296-2304. 19. Partridge EE. The National Cancer Data Base: Ten years of growth and commitment. CA Cancer J Clin 1998;48:131-133. 20. American College of Surgeons. National Cancer Data Base PUF. Available at: http://ncdbpufbeta.facs.org/. Accessed May 3, 2014. 21. Onega T, Duell EJ, Shi X, et al. Geographic access to cancer care in the U.S. Cancer 2008;112:909-918. 22. Massarweh NN, Chiang YJ, Xing Y, et al. Association between travel distance and metastatic disease at diagnosis among patients with colon cancer. J Clin Oncol 2014;32:942-948. 23. Volpp KG, Loewenstein G, Asch DA. Choosing wisely: Low-value services, utilization, and patient cost sharing. JAMA 2012;308:1635-1636. 24. Roberts KB, Soulos PR, Herrin J, et al. The adoption of new adjuvant radiation therapy modalities among Medicare beneficiaries with breast cancer: Clinical correlates and cost implications. Int J Radiat Oncol Biol Phys 2013;85:1186-1192. 25. Sen S, Soulos PR, Herrin J, et al. For-profit hospital ownership status and use of brachytherapy after breast-conserving surgery. Surgery 2014;155:776-788. 26. Hughes KS, Schnaper LA, Bellon JR, et al. Lumpectomy plus tamoxifen with or without irradiation in women age 70 years or older with early breast cancer: Long-term follow-up of CALGB 9343. J Clin Oncol 2013;31:2382-2387. 27. Haviland JS, Yarnold JR, Bentzen SM. Hypofractionated radiotherapy for breast cancer. N Engl J Med 2010;362:1843. author reply 1843e1844. 28. Herbert C, Nichol A, Olivotto I, et al. The impact of hypofractionated whole breast radiotherapy on local relapse in patients with Grade 3 early breast cancer: A population-based cohort study. Int J Radiat Oncol Biol Phys 2012;82:2086-2092. 29. Ashworth A, Kong W, Whelan T, et al. A population-based study of the fractionation of postlumpectomy breast radiation therapy. Int J Radiat Oncol Biol Phys 2013;86:51-57. 30. Berrang TS, Truong PT, Tyldesley S, et al. A population-based study of the fractionation of postlumpectomy breast radiation therapy. In regard to Ashworth, et al. Int J Radiat Oncol Biol Phys 2013;87:632-633. 31. Ashworth A, Kong W, Whelan T, et al. In Reply to Berrang, et al. Int J Radiat Oncol Biol Phys 2013;87:633. 32. Stage I and II Invasive Breast Cancer RT Section: BC Cancer Agency Cancer Management Guidelines. Available at: http://www.bccancer. bc.ca/HPI/CancerManagementGuidelines/Breast/Management/15. þStageþIþandþIIþInvasiveþBreastþCancerþRTþSection.htm. Accessed May 3, 2014. 33. Jagsi R, Abrahamse P, Hawley ST, et al. Underascertainment of radiotherapy receipt in Surveillance, Epidemiology, and End Results registry data. Cancer 2012;118:333-341.