Trends in radiotherapy for bone metastases, 2015–2017: Choosing Wisely in the era of complex radiation

Abstract

PURPOSE:

Guidelines recommend short-course (≤10 fractions) external-beam radiotherapy (EBRT) for bone metastases. Stereotactic body radiotherapy (SBRT) may also improve outcomes; however, routine use is not recommended outside clinical trials. We assess national radiotherapy trends in complex techniques for bone metastases and associated expenditures.

METHODS:

Using a claims-based Medicare dataset covering 84% of beneficiaries, we assessed the relative proportion of all radiation episodes represented by bone metastases. We then evaluated use of short-course and long-course (>10 fractions) EBRT, intensity-modulated radiotherapy (IMRT), and SBRT for bone metastases in hospital-affiliated outpatient (OPD) or freestanding (FREE) facilities. We assessed differences using chi-squared or Wilcoxon rank-sum tests for categorical and continuous variables, respectively. We identified associations with modality, fractionation, and expenditures using multivariable logistic/linear regression.

RESULTS:

Among 467,781 radiation episodes for 17 cancer diagnoses, the overall proportion of episodes dedicated to bone metastases (9.4%) was stable from 2015–2017, though treatments were increasing in the hospital-affiliated outpatient setting (P<0.005). We identified 40,993 episodes for bone metastases, of which 63% were short-course EBRT, 24% long-course EBRT, 7% SBRT, and 6% IMRT. Techniques more common in the hospital-affiliated outpatient setting included short-course EBRT (OPD, 69% vs. FREE, 56%) and SBRT (OPD, 9% vs. FREE, 5%). Techniques more common among freestanding centers included long-course EBRT (OPD, 19% vs. FREE, 31%) and IMRT (OPD, 4% vs. FREE, 9%). From 2015–2017, long-course EBRT decreased by an absolute 8%; short-course EBRT, SBRT, and IMRT increased by 4%, 2.5%, and 1%, respectively. SBRT/IMRT uptake did not differ by setting (P=0.4). Differences in expenditures between SBRT and short-course EBRT decreased by a relative 8% in professional and 12% in technical fees.

CONCLUSION:

Approximately 1-in-4 patients received long-course EBRT, with small reductions in use largely replaced by complex treatment modalities. However, expenditures for complex modalities also decreased over time. As alternative payment models take effect, quality metrics are needed to ensure appropriate, effective, and safe delivery of complex technologies.

INTRODUCTION

In patients with advanced cancer, bone metastases are a common cause of skeletal-related events, including pain and fracture, which can increase morbidity and decrease quality-of-life. Although the relative incidence of bone metastases varies by histology, recent estimates suggest a 10-year cumulative incidence of 28% across all cancer types (1). For decades, radiotherapy (RT) has been the mainstay for effective palliation of bone metastases, providing pain relief in up to 85% of patients, with at least 50% reporting complete pain response at 1–2 weeks (2). In particular, stereotactic body radiotherapy (SBRT)—a technique that delivers highly conformal radiation in ≤5 fractions to small targets throughout the body—is emerging as a safe, effective treatment for bone metastases (3). Recent prospective randomized trials report benefit with SBRT in terms of short-term palliation of pain (4,5) and progression-free survival for oligometastatic disease (6–8).

In 2013, the American Society of Radiation Oncology (ASTRO) released their Choosing Wisely guidelines, which specify avoidance of long-course (>10 fractions) external beam radiotherapy (EBRT) for bone metastases, and instead, advocate for short-course (≤10 fractions) regimens (9). In 2017, ASTRO updated their consensus guidelines further specifying that SBRT should only be employed in the setting of a clinical trial or prospective registry (10). Accordingly, use of long-course regimens decreased from 34% to 20% among patients receiving EBRT from 2011–2014 (11); at this time, SBRT comprised 4% of all treatments. A more recent analysis of Medicare claims data from 2016–2018 found that long-course regimens comprised 23% of EBRT treatments, suggesting limited improvement (12). However, neither study evaluated factors associated with use of complex modalities, including SBRT or IMRT, nor assessed trends over time including differences in expenditures by modality. Meanwhile, potential cost-effective solutions to delivering SBRT have been reported (13). Therefore, an updated comparative analysis of trends in practice and associated expenditures is warranted.

In July 2019, the Centers for Medicare and Medicaid Services (CMS) introduced the Radiation Oncology Alternative Payment Model (RO-APM), which proposes bundled reimbursements for radiation treatments, in part to discourage use of extended fractionation and control costs (14). As part of this announcement, CMS released a limited claims-based dataset of Medicare beneficiaries who received RT between 2015–2017. With promising prospective data available for SBRT in bone metastases (3), and the first randomized trial reporting a progression-free survival benefit for SBRT in oligometastatic disease published in 2016 (15), we sought to evaluate early trends in uptake of complex modalities for bone metastases and associated expenditures over time using this dataset. We hypothesize that during the study period, the proportion of radiation oncology practice dedicated to bone metastases increased, thereby accompanying an increase in use of complex modalities (primarily among hospital-affiliated outpatient centers) and a decrease in associated expenditures.

PATIENTS AND METHODS

Study cohort

The CMS dataset released as part of the RO-APM proposal includes a total of 517,988 radiation treatment episodes among Medicare beneficiaries, including 17 different cancer diagnoses (including bone metastases), representative of 84% of radiation treatments from 2015–2017. Episodes among Medicare beneficiaries under the age of 65 were excluded. For detailed analysis of bone metastases episodes, proton therapy, brachytherapy, intraoperative RT, and/or episodes involving >1 modality were excluded (Figure 1). This study was approved by the Institutional Review Board at Memorial Sloan Kettering Cancer Center.

Figure 1. Episode Cohort Breakdown.

The Centers for Medicaid and Medicare Services (CMS) Radiation Oncology database consists of over 517,988 90-day episodes representative of 84% of Medicare beneficiaries who received radiation for treatment of one of 17 cancer diagnoses from 2015–2017. Our final study cohort was limited to episodes for patients aged 65 and older, who received radiation in the form of EBRT, IMRT, or SBRT for the treatment of bone metastases. Patients who received >1 radiation modality or were treated with PBT, brachytherapy or IORT were excluded from the study.

Abbreviations: EBRT, external beam radiotherapy; IMRT, intensity-modulated radiotherapy; IORT, intraoperative radiotherapy; PBT, proton beam therapy; SBRT, stereotactic body radiotherapy.

Covariables

In addition to diagnosis and modality, available covariables included gender, age group (i.e. 65–74, 75–84, ≥85), major procedures (yes/no), chemotherapy (yes/no), and treatment setting. While a list of all possible major procedures and chemotherapy regimens given is available online (16), no specific, episode-level information is available. Mortality was classified by time period (i.e. death within 1–30 days, 31–60 days, or 61–90 days of treatment; or no death within 90 days).

Outcomes

Primary outcomes were receipt of any radiation for bone metastases (compared to all episodes), as well as use of short-course (≤10 fractions) EBRT, long-course (>10 fractions) EBRT, IMRT, or SBRT among bone metastases episodes only. Secondary outcomes were winsorized payments for professional (PRO) and technical (TECH) fees. These expenditures were adjusted for inflation using the medical component of the Consumer Price Index.

Statistical Analysis

With a baseline rate of SBRT use at 5% (from prior studies), we deemed an absolute 1% increase to be clinically meaningful (as this represents a relative 20% increase in use). With approximately 14,000 bone metastases episodes per year and an alpha of 0.05, we had a 95.6% power to detect this difference between 2015–2017. Episodes were further grouped by treatment setting (hospital-affiliated outpatient, OPD, or freestanding centers, FREE), with baseline differences between groups assessed using Pearson’s chi-squared or Wilcoxon rank-sum tests for categorical and continuous variables, respectively. Time trends were depicted using stacked bar charts, stratified by treatment setting. Multivariable logistic regression models adjusting for select covariables were used to identify factors associated with short-course (vs. long-course RT) and SBRT (vs. EBRT/IMRT). Similarly, multivariable linear regression models were used to identify factors associated with treatment-related expenditures. All models tested the interaction between year and facility type to determine whether time trends varied by treatment setting. P-values <0.01 indicated statistical significance. All statistics were performed using R version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Bone metastases as a component of routine radiation oncology practice

Among all 467,781 eligible episodes of radiation delivered between 2015–2017 among Medicare beneficiaries age ≥65 years old, 43,196 episodes were for the treatment of bone metastases. Although the total number of episodes for bone metastases increased slightly during the study period (Figure 2), the proportion of episodes for bone metastases relative to all cancer diagnoses remained stable, comprising 9.4% of all radiation episodes (OR 0.97, 95% CI: 0.93–1.01 in both years, P=0.2). On multivariable analysis comparing episodes for bone metastases to episodes for all cancer diagnoses (controlling for gender, age, major procedures, chemotherapy, and treatment year), treatment for bone metastases was more likely to be performed in freestanding centers compared to hospital-based facilities (OR 0.73, 95% CI: 0.70–0.75). However, there was an increasing trend over time in the proportion of bone metastases treated in hospital-affiliated outpatient facilities relative to freestanding (OR 1.06, 95% CI: 1.01–1.11 in 2016 and 1.09, 1.03–1.14 in 2017; P=0.005 for the interaction between year and facility type; Supplementary Table 4).

Figure 2. Radiotherapy Trends by Cancer Diagnosis, 2015–2017^†.

^† Note: Please refer to Supplementary Tables 5a–c for exact numbers and percentages by year.

Bone metastases cohort characteristics

In total, 40,993 episodes met criteria for study inclusion, including 58% of episodes occurring in hospital-affiliated outpatient centers and 42% in freestanding centers. Cohort characteristics by treatment setting are summarized in Table 1. Overall, 43% of patients were female, 52% of patients were in the youngest age category (65–74 years) and 74% had received chemotherapy within the 90-day episode, with statistically significant but small (and likely not clinically meaningful) differences based on facility type. There was also no significant difference in overall rate of death within 90 days of treatment between facility types (P=0.6).

Table 1.

Cohort Characteristics, All Bone Metastases

	Statistic*	ALL N=40,993	FREE N=17,307	OPD N=23,686	P-value†
Demographics
Gender
Female	N (%)	17480 (43%)	7190 (42%)	10290 (43%)	<0.001
Male	N (%)	23513 (57%)	10117 (58%)	13396 (57%)
Age group
65–74	N (%)	21193 (52%)	8739 (50%)	12454 (53%)	<0.001
75–84	N (%)	14963 (37%)	6459 (37%)	8504 (36%)
85+	N (%)	4837 (12%)	2109 (12%)	2728 (12%)
Treatment Characteristics
Major procedure(s)
Yes	N (%)	12010 (29%)	4872 (28%)	7138 (30%)	<0.001
No	N (%)	28,983 (71%)	12,435 (72%)	16,548 (70%)
Chemotherapy
Yes	N (%)	30,492 (74%)	13188 (76%)	17304 (73%)	<0.001
No	N (%)	1,0501 (26%)	4,119 (24%)	6,382 (27%)
Radiation modality
SBRT	N (%)	2840 (7%)	818 (5%)	2022 (9%)	<0.001
EBRT	N (%)	35697 (87%)	14986 (87%)	20711 (87%)
IMRT	N (%)	2456 (6%)	1503 (9%)	953 (4%)
Course length
1–10 services	N (%)	29,939 (73%)	11118 (64%)	18821 (79%)	<0.001
11+ services	N (%)	11,054 (27%)	6189 (36%)	4865 (21%)
Expenditures
Provider fees (USD, $)	Median (IQR)	1234 (948, 1655)	1348 (1025, 1785)	1158 (900, 1554)	<0.001
Technology fees (USD, $)	Median (IQR)	4429 (3245, 6094)	4528 (3522, 6268)	4332 (3104, 5942)	<0.001
Clinical Outcomes
Death during 90-day episode
No death	N (%)	32382 (79%)	13705 (79%)	18677 (79%)	0.6
Within 1–30 days	N (%)	1993 (5%)	846 (5%)	1147 (5%)
Within 31–60 days	N (%)	3586 (9%)	1477 (9%)	2109 (9%)
Within 61–90 days	N (%)	3032 (7%)	1279 (7%)	1753 (7%)

Abbreviations: 95% CI, 95% confidence interval; EBRT, external beam radiotherapy; FREE, freestanding; IMRT, intensity modulated radiotherapy; OPD, hospital-affiliated outpatient; OR, odds ratio; SBRT, stereotactic body radiotherapy.

^*Statistics presented: N (%); Median (IQR)

^†Statistical tests performed: chi-square test of independence; Wilcoxon rank-sum test

Trends in radiation fractionation and modality for bone metastases

Trends in RT for bone metastases are summarized in Figure 3. EBRT accounted for 87% of all treatments, with short-course and long-course EBRT comprising 63% and 24% of these episodes, respectively. From 2015–2017, long-course EBRT decreased by an absolute 8% (from 28% to 20%) and was more common among freestanding facilities (OPD, 19% vs. FREE, 31%). Meanwhile, short-course EBRT increased by an absolute 4% (from 61% to 65%) and was more common in hospital-affiliated outpatient facilities (OPD, 69% vs. FREE, 56%). On multivariable logistic regression, patient characteristics associated with higher odds of receiving short-course RT included being older (age >75), not undergoing major procedures, and death within 90 days of treatment (all P<0.01; Table 2).

Figure 3. Radiotherapy Trends by Provider Type, 2015–2017 *.

Abbreviations: EBRT, external beam radiotherapy; IMRT, intensity-modulated radiotherapy; Long, 11+ fractions; SBRT, stereotactic body radiotherapy; Short, 1–10 fractions.

* Note: Please refer to Supplementary Table 1a–c for exact numbers and percentages by year.

Table 2.

Multivariable Logistic Regression, Probability of ≤10 fractions ^*

Observations used: All bone metastases (subset ≤10 fractions) = 40,993 (29,939)
	OR	95% CI	P-value
Demographics
Gender
Female	—	—	0.2
Male	1.03	0.98, 1.08
Age group
65–74	—	—	<0.001
75–84	1.08	1.03, 1.14
85+	1.32	1.22, 1.43
Treatment Characteristics
Major procedure(s)
No	—	—	0.007
Yes	0.93	0.89, 0.98
Chemotherapy
No	—	—	0.016
Yes	1.07	1.01, 1.13
Radiation modality
SBRT	—	—	<0.001
EBRT	0.01	0.01, 0.02
IMRT	0.00	0.00, 0.01
Year
2015	—	—	<0.001
2016	1.20	1.11, 1.30
2017	1.37	1.27, 1.49
Facility Type
FREE	—	—	<0.001
OPD	1.93	1.79, 2.09
Year * Provider type
2016 * OPD	1.03	0.92, 1.15	0.10
2017 * OPD	1.13	1.01, 1.26
Clinical Outcomes
Death during 90-Day episode
Did not die within 90 days	—	—	<0.001
Death within 1–30 days	14.4	11.3, 18.7
Death within 31–60 days	2.06	1.88, 2.26
Death within 61–90 days	1.31	1.20, 1.43

Abbreviations: EBRT, external beam radiotherapy; FREE, freestanding; IMRT, intensity-modulated radiotherapy; OPD, hospital-affiliated outpatient; OR, odds ratio; SBRT, stereotactic body radiotherapy.

^*Versus long-course RT (11+ fractions)

The remaining 13% of episodes involved complex radiation modalities: 7% SBRT and 6% IMRT. From 2015–2017, there was an absolute 3.5% increase in the collective use of complex modalities (from 12% to 15%), representative of an absolute 2.5% and 1% increase in SBRT and IMRT, respectively, as well as a significant association between later years and higher odds of complex radiation modality use compared to 2015 (SBRT and IMRT: P<0.001 and P=0.033, respectively; Table 3). Notably, 52% of IMRT episodes were long-course (>10 fractions), compared to 27% of EBRT episodes. Similar to EBRT, long-course regimens accounted for a larger percentage of IMRT episodes in the freestanding setting (OPD, 40% vs. FREE, 59%).

Table 3.

Multivariable Logistic Regression, Probability of Complex Radiation Modalities

	SBRT1			IMRT2
Observations used:	All episodes (SBRT) = 40,993 (2,835)			All episodes (IMRT) = 40,993 (2,456)
	OR	95% CI	P-value	OR	95% CI	P-value
Demographics
Gender
Female	—	—	0.001	—	—	0.02
Male	1.14	1.05, 1.23		1.10	1.02, 1.20
Age group
65–74	—	—	<0.001	—	—	0.3
75–84	0.88	0.81, 0.95		0.94	0.86, 1.03
85+	0.60	0.52, 0.69		0.91	0.80, 1.04
Treatment characteristics
Major procedure(s)
No	—	—	<0.001	—	—	<0.001
Yes	1.22	1.13, 1.33		0.74	0.67, 0.81
Chemotherapy
No	—	—	<0.001	—	—	0.5
Yes	0.58	0.53, 0.63		0.97	0.88, 1.07
Year
2015	—	—	<0.001	—	—	0.03
2016	1.30	1.08, 1.56		1.01	0.88, 1.15
2017	1.56	1.31, 1.86		1.17	1.02, 1.33
Facility Type
FREE	—	—	<0.001	—	—	<0.001
OPD	2.01	1.71, 2.36		0.44	0.38, 0.51
Year * Provider type
2016 * OPD	0.86	0.69, 1.07	0.4	0.96	0.77, 1.19	0.8
2017 * OPD	0.90	0.73, 1.11		1.03	0.84, 1.26
Clinical outcomes
Death during 90-Day episode
Did not die within 90 days	—	—	<0.001	—	—	<0.001
Death within 1–30 days	0.23	0.17, 0.31		0.62	0.49, 0.77
Death within 31–60 days	0.28	0.23, 0.35		0.78	0.66, 0.91
Death within 61–90 days	0.37	0.30, 0.45		0.72	0.60, 0.86

Abbreviations: 95% CI, 95% confidence interval; EBRT, external beam radiotherapy; FREE, freestanding; IMRT, intensity modulated radiotherapy; OPD, hospital-affiliated outpatient; OR, odds ratio; SBRT, stereotactic body radiotherapy.

¹Probability of SBRT (versus EBRT or IMRT alone)

²Probability of IMRT (versus EBRT or SBRT alone).

On multivariable logistic regression, treatment in hospital-affiliated outpatient facilities was associated with higher odds of SBRT (OPD, 9% vs. FREE, 5%; OR 2.01, 95% CI: 1.71–2.36) and lower odds of IMRT (OPD, 4% vs. FREE 9%; OR 0.44, 95% CI: 0.38–0.51; both P<0.001). Although SBRT was increasing at a slower rate among hospital-affiliated outpatient facilities compared to freestanding facilities, this difference was not statistically significant (Interaction test: Year*Facility Type, P=0.4). Patient characteristics associated with higher odds of SBRT included being male and younger (age <75), while lower odds of SBRT were associated with receipt of chemotherapy and death within 90 days of treatment (all P<0.001). Similar characteristics were associated with receipt of IMRT (Table 3).

Trends in radiation treatment expenditures for bone metastases

Compared to 2015, overall expenditures significantly increased in 2016 and 2017, both in terms of professional fees and technical fees (both P<0.001). On multivariable linear regression, EBRT and IMRT were significantly associated with lower expenditures compared to SBRT (both P<0.001; Supplementary Table 2). Other factors associated with increased expenditures included long-course (>10 fractions) regimens and treatment in freestanding facilities (both P<0.001). For episodes with ≤10 fractions, SBRT averaged $1,755 in professional and $10,240 in technical fees; IMRT averaged $1,708 in professional and $5,991 in technical fees; and EBRT averaged $1,151 in professional and $3,538 in technical fees.

Notably, from 2015 to 2017, the differences in expenditures between complex modalities and short-course EBRT were decreasing. For SBRT and short-course EBRT, differences in expenditures decreased by a relative 8% in professional fees (from $721 to $664) and 12% in technical fees (from $7,026 to $6,168). For IMRT and short-course EBRT, differences in expenditures decreased by a relative 19% in professional fees (from $591 to $479) and 28% in technical fees (from $2,908 to $2,095). Further details characterizing expenditure trends by modality, fractionation, and treatment setting are summarized in Supplementary Table 3.

Lastly, while technical fees did not differ by setting (P=0.5), they were increasing at a significantly faster rate over time among hospital-affiliated outpatient facilities on multivariable analysis controlling for both modality and fractionation (Interaction test: Year*Facility Type, P<0.001; Supplementary Table 2).

DISCUSSION

To our knowledge, this nationally comprehensive, claims-based analysis of >40,000 radiation episodes for bone metastases from 2015–2017 is the largest contemporary study evaluating national trends in use of complex radiation technologies and associated expenditures. Collectively, our results demonstrate that long-course EBRT for bone metastases decreased over time (from 28% in 2015 to 20% in 2017), in accordance with Choosing Wisely recommendations, potentially demonstrating an improvement from prior estimates ranging from 24–36% (11,12,17). However, while there has been a small concomitant increase in the adoption of short-course EBRT (from 61% to 65%), a nearly equal rise in the use of complex modalities was also observed (from 12% to 15%). Expenditures associated with complex treatments did, nonetheless, decrease over time.

Importantly, there was no difference in rates of SBRT uptake between hospital-affiliated (more often academic) centers and freestanding (more often community-based) clinics, thus suggesting increasing use in routine practice, despite ASTRO guidelines specifying that highly conformal modalities should be limited to clinical trials and prospective registries (18). This may be supported by randomized trial data available as early as 2016, which at the time, suggested potentially improved outcomes with SBRT in well-selected patients (15), and have since become more robust between 2018–2020 (4,7,19). Without detailed clinical information in the current study, we cannot assess the appropriateness of increasing use nationally, but can draw attention to a potential gap between contemporary practice patterns and the availability of clinical practice guidelines for complex technologies.

Although the overall incidence of cancer in the U.S. has remained stable over time, cancer spending continues to account for approximately 10% of total healthcare expenditures (20), with current estimates suggesting that cancer-related expenditures are projected to continue their disproportionate increase compared to other health expenditures in the next decade (21). Our results show that although expenditures for radiation of bone metastases are increasing overall, absolute differences in expenditures between conventional (short-course EBRT) and complex (SBRT or IMRT) treatments are decreasing over time. This finding could reflect both decreasing reimbursement as well as more efficient processes within radiation clinics. In a recent prospective study, STAT RAD, an expedited workflow for single-fraction SBRT actually reduced expenditures compared to 10-fraction conventional RT for bone metastases (13).

Meanwhile, in 2014, the American Society for Clinical Oncology (ASCO) recommended moving toward bundled payments in an effort to standardize high-quality, value-based cancer care (22). Then, in 2019, with radiation oncology identified as a specialty exceeding the sustainable growth rate (23), the CMS proposed a new bundled payment model (RO-APM) (14). Of note, while bundled payments may decrease expenditures by reducing persistent use of extended fractionation schemes that are unsupported by evidence, they may also reduce use of complex radiation modalities which may, in select circumstances, improve patient outcomes. To combat this, the Final Rule announced by CMS mandates reporting of patient characteristics and established quality metrics. First, all episodes of bone metastases will require reporting of Clinical Data Elements including ECOG Performance Status, prior radiation to the site (yes/no), and spinal location (yes/no) (24), which is consistent with prior reports (Gillespie et al, submitted), though may warrant greater specification in national guidelines. Second, quality measures will include a plan of care for pain (CMW Quality ID #144), communication of the radiation treatment summary, and documentation of an advanced care plan (CMS Quality ID #47) (16). To further optimize outcomes in the setting of bundled payments, research is needed to establish benchmarks for appropriate use of complex treatments (especially SBRT) and identify patient-centered outcomes (such as skeletal-related events and toxicity) that could ensure avoidance of under-treatment.

There are several limitations to this study. First, our cohort includes Medicare recipients age ≥65, thereby excluding younger patients, those with private insurance, and those who pay out-of-pocket, thus limiting generalizability. Second, the dataset lacks clinical details which may have influenced choice of modality and fractionation including performance status, number of metastases, life expectancy, history of prior radiation, and the presence of complicated bone metastases, i.e. lesions with impending or existing pathologic fracture, neurologic compromise, and/or retreatment (25). Third, we could not account for the potential dependence between episodes from the same patient because the dataset does not include a patient identifier. Fourth, hospital-based outpatient facilities can encompass both tertiary centers and hospital-affiliated satellite sites (26). As such, we are unable to evaluate practice patterns at satellite centers separately, though such an investigation may be of future interest given their expanding role in cancer care (27). Lastly, the dataset does not specify the exact number of fractions delivered, thereby precluding assessment of single-fraction RT.

CONCLUSION

Despite multiple published guidelines, still 1-in-4 patients treated for bone metastases from 2015–2017 received long-course conventional EBRT. Fortunately, long-course EBRT is decreasing, though it is largely being replaced by complex treatments. This is occurring in both hospital-affiliated outpatient and freestanding settings, and therefore, likely reflects increasing use in routine practice rather than strictly on-trial, as current guidelines specify. As evidence emerges in support of complex radiation technologies, guidance on appropriate use in routine care is needed. Meanwhile, with systems-level changes in reimbursement forthcoming, metrics to better evaluate appropriate use and sustain high quality care are also needed.