Stereotactic body radiation therapy (SBRT) for metastatic renal cell carcinoma: A multi-institutional experience

Abstract

Objectives: Examine local control(LC), overall survival(OS), and toxicity following stereotactic body radiation therapy(SBRT) for patients with metastatic renal cell carcinoma(mRCC).

Methods: A multi-institutional registry was queried. Potential predictive factors of LC and OS were evaluated with a Cox-proportional hazards model for multivariate analysis(MVA).

Results: We identified 115 mRCC patients with 181 lesions. Median biologically effective dose (BED₇) was 72.9 Gy₇ (range: 42.9–231.4 Gy₇) with a median dose/fraction of 10 Gy (range: 5-24 Gy). Utilizing both Karnofsky Performance Score (KPS) and presence of osseous metastatic disease as prognostic indicators, estimated 2-year OS rates were 67.7% (95% CI: 49.9-89.5%), 31.8% (95% CI: 19.0-45.3%), and 20% (95% CI: 1.4-54.7%; p=0.0012). One- and 2-year LC rates were 88.2% and 82.7%, respectively, with no prognostic factors identified. Roughly 13% of patients reported toxicities with one Grade 3-5 toxicity.

Conclusion: SBRT was well-tolerated with promising LC. Both KPS and osseous metastatic disease should be considered in determining which patients with mRCC may preferentially benefit from SBRT.

Introduction

Renal cell carcinoma (RCC) is an aggressive malignancy with a tendency to metastasize to the lungs, bone, liver, spine, and brain [1]. Metastases from renal cell carcinoma (mRCC) have long been considered to be radioresistant on the spectrum of malignant diseases, the role of conventional radiotherapy (RT) for mRCC has been limited often to palliative intent treatments with modest local control (LC) [2-4]. However, over the last two decades, stereotactic radiosurgery (SRS) and stereotactic body radiation therapy (SBRT) have surfaced as effective treatment modalities for metastatic renal cell carcinoma (mRCC) [5-7]. Early studies successfully demonstrated the efficacy of SBRT for treating mRCC with LC rates ranging from 87-98% and pain alleviation rates of roughly 90% [5-7]. Dose and fractionation schedules varied among early studies, ranging from 26-40 Gy/3-6 fractions dependent on the location of metastatic disease [5-6]. Particularly for mRCC of the spine, single fraction SBRT with a mean marginal dose of 20 Gy (range: 17.5 to 25 Gy) have been reported[7]. More recent investigations into the efficacy of SBRT for mRCC have also shown promising 1-year LC rates of nearly 100% and noted LC benefits to dose escalation with biologically effective doses (BEDs) greater than 100 Gy for various sites of mRCC and 130 Gy specifically for pulmonary metastatic disease treated with SBRT [8,9].

Despite these advances, the optimal dose and fractionation schedules for mRCC to maximize LC and reduce morbidity remain unclear, with most investigations limited to single-institutional series. As the role of SBRT in the management of patients with mRCC continues to grow, a consensus regarding the ideal dose and fractionation selection for SBRT is paramount as is the identification of other prognostic factors for patients with mRCC treated with SBRT [10]. The RSSearch Patient Registry (RSSPR) is a multi-institutional, international database created for gathering clinical information in patients treated with SRS or SBRT at both academic and community-based radiotherapy centers. Containing data for over 25,000 patients, the RSSPR has been repeatedly demonstrated as a successful tool in the analysis of patient outcomes for numerous types of cancer treated with SBRT, providing results comparable to prior literature [11-13]. Thus, we sought to utilize the RSSPR to examine overall survival (OS), LC, and treatment-related toxicities for mRCC patients treated with SBRT across a wide range of clinical practice settings to help further guide and optimize dose selection.

Materials and Methods

The RSSearch Patient Registry (Clinicaltrials.gov Identifier: NCT01885299) was searched to identify patients with mRCC treated with SBRT from May 2006 to February 2019. Inclusion criteria required that there was thorough documentation available regarding patient characteristics, lesion features, treatment planning, OS, and patient follow-up. Initial gross tumor volume (GTV) and LC were evaluated either by positron-emission tomography (PET)/computed tomography (CT), magnetic resonance imaging (MRI), or CT alone.

All centers utilizing SBRT as a modality for treating patients are offered and encouraged to voluntarily participate in RSSearch, and no compensation is provided either to the patients nor the participating centers. Local Institutional Review Board/Ethics Committee (IRB/EC) approval is required at all participating centers. Prior to data entry, informed consent was obtained from all patients, as required by individual IRB/ECs. Radiotherapy planning was performed per individual institutional guidelines using inverse planning on the MultiPlanSystem^® (Accuray Incorporated, Sunnyvale, CA), and all patients were treated using the CyberKnife^® Robotic Radiosurgery System. Both planning of SBRT and treatment delivery differed per respective institutional preferences.

Statistical summaries of relevant patient, treatment, and lesion characteristics were performed with descriptive analyses. Potential predictive factors of LC and OS (including patient age, prescription doses and fractionation schedules, initial Karnofsky Performance Scores (KPS), gross tumor volume (GTV), gender, tumor location, and BED were evaluated using time-to-event analysis and log-rank t-tests with the Kaplan-Meier method, as well as a Cox-proportional hazards model for multivariate analysis (MVA) with a forward entry parsimonious method including variables identified to be significant on UVA with a p < 0.05 considered statistically significant.

The effect of dose escalation was examined at cutoffs of BED₇ ≥ 85 Gy₇, 100 Gy₇, and 125 Gy₇ among all fractionation schedules based on prior literature [23]. BED was calculated using a linear quadratic model with an assumed alpha-beta ratio of 7 define the BED₇ based on prior reports that have utilized an alpha-beta ratio of 7 for means of comparison [23]. For purposes of OS analysis, if the patient had multiple lesions treated and any were treated with a BED₇ < 100, then this was coded as < 100 Gy₇. If the patient had multiple lesions treated and all were treated with a BED₇ ≥ 100, then this was coded as ≥ 100 Gy₇. We also examined the impact of dose-per-fraction as an alternative means of assessing the impact of dose escalation, with cutoffs of 10 Gy based on the median dose per fraction among our cohort as well as 15 Gy based on prior literature [17]. The correlation of side-effect incidence (with toxicities graded based on Common Terminology Criteria for Adverse Events (CTCAE) guidelines) and dose escalation was evaluated via logistic regression.

Results

Patient, treatment, and lesion characteristics

A summary of the cohort’s demographic data, radiotherapy planning, treatment, and lesion characteristics can be found in Table 1. A total of 115 patients with 181 treated lesions met inclusion criteria, and 71 of these patients also had LC evaluated. The median age of patients was 66 years (range: 33-84) with 40 female patients and 75 male patients. Eighty-six percent of patients in the cohort were of Caucasian ethnicity, with a median pre-treatment KPS of 90% (range: 50-100%). The majority of patients treated with SBRT had disease located in the spine (42 patients, 36.5%), lungs (29 patients, 25.2%), or non-spinal osseous metastases (26 patients; 22.6%). Median GTV was 26.3cc (range: 0.41-110cc) and the median time to local progression in the cohort was 9.15 months (range: 4.18–41.68 months). At patients’ last follow-up, the most common radiographic responses following SBRT were no growth in the size of the lesion treated with SBRT or change in distant disease (stable disease; 36.7% of lesions treated) or LC following SBRT with progression of distant metastatic disease (29.6% of lesions treated).

Table 1.

Summary of patient and lesion characteristics and radiotherapy planning

Variable
Gender	Female – 40 patients Male – 75 patients
Median Age (years) (range)	66 (33 – 84)
Race	Caucasian – 100 patients African-American – 8 patients Asian – 1 patient Multiracial – 2 patients Unknown – 4 patients
Median Initial KPS (range)	90% (50%-100%)
Median Initial GTV (cc) (range)	26.3 (0.41-110)
Location	Spine – 42 patients Lung – 29 patients Non-spinal bone metastases – 26 patients Adrenal Gland – 5 patients Lymph nodes – 4 patients Liver and Intrahepatic Bile Ducts – 2 patient Other – 7 patients
Median number of fractions (range)	3 (1 – 5)
Median dose per fraction (Gy) (range)	10 (5-24)
Median Prescription Dose (Gy) (range)	1 fraction (n = 17): 18 (15-26) 2 fractions (n = 3): 27 Gy (26-28) 3 fractions (n = 54): 31.5 (21 – 60) 4 fractions (n = 6): 49 (48 – 60) 5 fractions (n = 35): 31 (25 – 45)
Median BED7 (Gy7) (range)	Entire cohort (n = 115): 72.9 (42.9–231.4) 1 fraction (n = 17): 64.3 (52.6–122.6) 2 fractions (n = 3): 79.1 (74.3–84) 3 fractions (n = 54): 72.9 (42–231.4) 4 fractions (n = 6): 134.8 (130.3–188.6) 5 fractions (n = 35): 58.4 (42.9 – 121.4)
Fractionation by lesion location	1 fraction (4 lung, 4 spine, 6 bone, 3 others) 2 fractions (1 spine,1 bone, 1 other) 3 fractions (15 lung, 23 spine, 9 bone, 7 other) 4 fractions (3 lung,1 bone, 2 others) 5 fractions (6 lung, 14 spine, 9 bone, 6 others)
Treatment response at last radiographic follow-up	Complete Response: 11.3% (8 patients) Partial Response: 12.7% (9 patients) Stable Disease: 36.7% (26 patients) LC with Distant Metastasis: 29.6% (21 patients) Locally Progressive Disease: 9.9% (7 patients)
Median time to local progression (n = 7) (range)	9.15 (4.18 – 41.68)

Fractionation schedules varied, with the median dose and fractionation among the cohort being 36 Gy in 3 fractions. The cohort’s median BED₇ was 72.9 Gy₇ (range: 42.9–231.4 Gy₇) with a median dose per fraction of 10 Gy (range: 5-24 Gy). Seventeen patients were treated with 1 fraction (median dose of 18 Gy (range:15-26 Gy)), and 3 patients were treated with 2 fractions (median dose: 27 Gy, (range: 26-28 Gy)). Fifty-four patients were treated with 3 fractions (median dose of 31.5 Gy (range: 21-60 Gy)), 6 patients were treated with 4 fractions (median dose of 49 Gy (range: 48-60 Gy)), and 35 patients were treated with 5 fractions [median dose of 31 Gy (range: 25-45 Gy)].

Overall Survival

Table 2 demonstrates the relationship between variables of interest and 1-year OS rates as well as median OS for patients in our cohort. Prognostic factors evaluated via Kaplan-Meier analysis included BED₇, location of metastases, initial KPS, GTV, patient age, fractionation scheme, dose/fraction, and lesion location. Following SBRT, 1- and 2-year OS rates were 69.1% (95% CI: 59.3-76.9%) and 45.2% (95% CI: 34.6-55.1%), respectively.

Both initial KPS and location of treated metastases were found to be significantly correlated with OS on UVA. Patients with KPS ≥ 80% demonstrated superior median OS compared to those with a KPS of < 80% (20.3 months vs. 6.6 months; p = 0.007). Following MVA, patients with KPS ≥ 80% were not found to have significantly improved prognosis with respect to OS (hazard ratio (HR) = 0.50 (95% CI: 0.24-1.05; p = 0.068)). Patients with osseous metastases demonstrated a significantly lower median OS compared to patients with non-osseous metastases (15.1 months vs. 29.3 months; p = 0.0023). This finding was also significant on MVA with poorer OS noted in patients with osseous metastases (HR = 2.16 (95% CI: 1.01–4.62); p = 0.046). When examining the prognostic significance of both osseous metastases and KPS < 80%, patients with combined scores of 0 (no osseous metastases and KPS ≥ 80%), 1 (either osseous metastases or KPS < 80%), and 2 (both osseous metastases and KPS < 80%), had 2-year OS rates of 67.7% (95% CI: 49.9-89.5%), 31.8% (95% CI: 19.0-45.3%), and 20% (95% CI: 1.4-54.7%) (Figure 1; p = 0.0012).

Figure 1.

Kaplan-Meier curves examining OS by both presence of osseous metastases and KPS.

The impact of radiation therapy planning on median OS is also displayed in Table 2. Patients treated to BED₇ ≥ 100 Gy₇ on univariate analysis had higher 1-year and median OS rates (p = 0.052), though this was not significant on MVA (p = 0.285). Those treated with SBRT of doses per fraction of 15 Gy or greater had higher median OS (30.5 vs. 15.1 months; p = 0.009) on univariate analysis that was not found to be significant on MVA (HR = 0.57 (95% CI: 0.28-1.17); p = 0.127)). Patients treated with single fraction SBRT (n= 16) demonstrated a higher 1-year OS rate of 82.5% (95% CI: 46.1–95.3%) compared to hypofractionated SBRT (67.3% (95% CI: 56.8 – 75.8%)), though again this was non-significant on MVA (p = 0.57).

Table 2.

Kaplan-Meier Analysis of potential predictive factors for OS following SBRT

Variable	Number of Patients	1-year OS Rate (95% CI)	Median OS (months)	p-value
BED7
< 85 Gy7	72	66.4% (53.5-76.5%)	17.8	0.15
≥ 85 Gy7	43	72.3% (55.3-83.7%)	27.9
< 100 Gy7	78	65.0% (52.7-74.9%)	17.6	0.052
≥ 100 Gy7	37	75.7% (56.9-87.1%)	28.4
< 125 Gy7	91	66.4% (55.0-75.4%)	17.6	0.062
≥ 125 Gy7	24	77.0% (53.1-89.8%)	29.5
Fractionation				0.53
Single Fraction	17	82.5% (46.1-95.3%)	30.5
Multiple Fractions	98	67.3% (56.8-75.8%)	18.3
Dose per fraction
< 10 Gy	44	64.5% (48.2-76.9%)	15.8	0.06
≥ 10 Gy	71	71.4% (58.4-81.0%)	28.4
< 15 Gy	80	63.2% (51.3-73.0%)	15.1	0.009
≥ 15 Gy	35	82.5% (62.8-92.3%)	30.5
Location
Osseous	68	61.6% (48.3 – 72.4%)	15.1	0.002
Non-Osseous	47	78.7% (62.9 – 88.3%)	29.5
Lung	29	80.2 % (58.9 – 91.3%)	15.8
Non-Spinal Bone	26	73.9% (50.1% – 87.3%)	12.4
Age				0.29
< 65 years	53	70.0% (55.1 – 80.7%)	17.6
≥ 65 years	62	68.5% (54.6 – 78.9%)	21.5
GTV				0.65
< 25 cc	37	66.3% (48.3 – 79.3%)	29.1
≥ 25 cc	78	70.5% (58.3 – 79.7%)	17.6
Initial KPS				0.007
< 80%	13	41.7% (15.3% - 66.5%)	6.6
≥ 80%	102	72.4% (62.2 – 80.4%)	20.3

Local Control

Table 3 depicts the impact of predictive factors on LC. The 1- and 2-year LC rates for the cohort were 88.2% (95% CI: 73.7-94.9%) and 82.7% (95% CI: 63.1-92.5%), respectively. When investigating the effect of various prognostic factors on LC rates, there was no difference in 1-year LC among patients with osseous metastases (89.2%) as compared to non-osseous metastases (86.7%) on UVA (Figure 2; p = 0.93). Patients aged < 65 years old had a 1-year LC rate of 100%, whereas patients ≥ 65 years old had a 1-year LC rate of 84.4% (p = 0.054; Figure 3), though this was not significant on MVA (p = 0.226). GTVs ≥ 25 cc were not found to be associated with inferior LC (p = 0.51).

Figure 2.

Kaplan-Meier curves examining difference in LC for mRCC patients with osseous vs. non-osseous metastases following SBRT.

Figure 3.

Kaplan-Meier curves examining difference in LC for mRCC patients based on age.

With regards to the impact of radiation therapy planning, we did not identify a correlation between dose or fractionation schedule and LC. No statistically significant dose response was noted for BED₇ for the entire cohort at dose cutoffs of 85 Gy₇, 100 Gy₇, or 125 Gy₃ on univariate analysis (p = 0.99, p = 0.75, and p = 0.44, respectively; Figure 4) or on MVA. Similarly, dose per fraction at cutoffs of either 10 Gy (p = 0.84) or 15 Gy (p = 0.40) had no significant impact on LC based on univariate and MVA. With regards to fractionation schedule, no difference in LC was noted when comparing single and hypofractionated SBRT (p = 0.83).

Figure 4.

Kaplan-Meier curves examining difference in LC for patients treated with BED₇ > 125 Gy.

Toxicities

Following SBRT, the incidence of acute and late toxicities was relatively low. Acute and late toxicities were reported by 13.04% of patients (15 patients). Of these, 66.7% were acute toxicities and 33.3% late toxicities. All toxicities were either Grade 1-2 (66.7% were Grade 1 (10 patients), 26.7% were Grade 2 (4 patients)) other than one reported Grade 3 toxicity of vomiting following SBRT to a spinal metastasis. With regards to Grade 2 toxicities, the majority were fatigue (3/4) with one case of Grade 2 nausea following treatment of an adrenal metastasis. Dose escalation with BED₇ ≥85 Gy₇ was not found to be significantly correlated with toxicity incidence (p = 0.07) nor were higher doses per fraction at cutoffs of 10 Gy (p = 0.64) or 15 Gy (p = 0.53).

Table 3.

Kaplan-Meier Analysis of potential predictive factors on LC following SBRT

Variable	Number of Lesions	1-year LC rate (95%CI)	p-value
BED7
< 85 Gy7	42	92.2% (71.5-98.0%)	0.99
≥ 85 Gy7	29	82.3% (55.8-94.2%)
< 100 Gy7	47	92.3% (73.7-98.2%)	0.75
≥ 100 Gy7	24	80.1% (51.1-93.4%)
< 125 Gy7	55	87.2% (69.0-95.1%)	0.44
≥ 125 Gy7	16	91.7% (53.9-98.8%)
Fractionation			0.83
Single Fraction	15	87.5% (38.7-98.1%)
Multiple Fractions	56	88.5% (71.8-95.6%)
Dose per fraction
< 10 Gy	31	95.8% (73.9-99.4%)	0.84
≥ 10 Gy	40	82.4% (59.5-93.0%)
< 15 Gy	42	87.3% (64.8-95.8%)	0.40
≥ 15 Gy	29	89.6% (64.3-97.3%)
Location
Osseous Metastases	42	89.2% (69.8-96.5%)	0.93
Non-Osseous Metastases	29	86.7% (56.4%-96.5%)
Bone (Excluding Spine)	24	91.3% (69.5-97.8%)
Lung	21	83.5% (64.4-92.9%)
Age			0.054
< 65 years	31	100% (N/A)
≥ 65 years	40	76.1% (51.0-89.5%)
GTV			0.51
< 25 cc	18	100% (N/A)
≥ 25 cc	53	84.4% (66.2 – 93.3%)
Initial KPS			0.43
< 80%	13	100%
≥ 80%	58	86.9% (71.0-94.3%)

Discussion

The role of SBRT in the setting of metastatic disease is growing, particularly for patients with oligometastatic disease given prior studies suggesting improved progression-free survival as well as OS for a variety of primary sites [14-16]. While studies examining the role of SBRT strictly in the setting of mRCC are limited, further investigation is warranted given the theoretical radioresistance of RCC and the potential of SBRT to provide superior LC as compared to conventionally RT. As such, we aimed to examine LC outcomes following SBRT for mRCC, with a particular interest in determining if dose-escalation results in superior LC.

Among our cohort, 1- and 2-year LC rates of 88.2% for and 82.7%, respectively. These findings are comparable with the results of numerous studies. For example, Grossman et al., noted a 1-year LC rate of 94.7% for 16 patients with mRCC oligometastatic lesions treated with SBRT (median dose of 50 Gy in 5 fractions) [17]. Similarly, one of the first reported experiences of SBRT for mRCC by Ranck, et al., reported on outcomes for 18 patients with 39 oligometastatic sites (12/18 patients treated to all sites of disease) with 2-year LC and OS rates of 91.4% and 85% among their cohort, respectively [18]. Notably, a recent systematic review and meta-analysis examining outcomes for patients with mRCC treated with SBRT noted 1-year LC rates ranging from 72.4% to 99.4% with an overall 1-year LC estimate of 89.1% [19]. Lower LC rates have been seen among series examining mRCC patients treated with SBRT to sites of spinal metastases, with poorer LC thought to be due to limitations in dose escalation secondary to spinal cord dose constraints [19-22].

When examining the effects of BED on LC and OS, we found patients treated to a BED₇ of ≥ 100 Gy₇ trended towards increased OS compared to patients treated to < 100 Gy₇ (p = 0.052) with no significant correlation found with LC on univariate analysis with no correlation between BED₇ and either LC or OS on MVA. We did note that patients treated with doses of 15 Gy or greater per fraction had higher median OS (30.5 vs. 15.1 months) on UVA that was not significant on MVA (HR = 0.57 (95% CI: 0.27-1.17); p = 0.127). In comparing our findings to other series, Altoos et al. examined outcomes among patients with mRCC treated with SBRT (36 lesions treated to a median dose of 50 Gy in 5 fractions and a median BED of 216.67 Gy with 75% of lesions being thoracic) and found a significant improvement in LC rates upon univariate analysis when lesions were treated to a BED ≥ 100 Gy (HR = 0.048; 95 % CI, 0.006–0.382; p=0.005), though this finding was not significant on MVA [8]. Amini et al. also illustrated the value of BED₇ as a predictive factor of LC in treating 48 patients with 95 total lesions from mRCC with either SBRT (median dose of 27 Gy in 3 fractions) or RT with BED₇ ≥80 Gy₇ associated with superior LC under univariate and multivariate analysis (HR = 0.14; 95% CI, 0.025-0.787; p = 0.026) [23]. Grossman, et al., have previously noted a relationship between total SBRT dose and OS (HR = 0.94 (95% CI: 0.90-0.99)) but did not find a significant relationship between dose/fraction and OS [17]. Similarly, one of the largest single institution experiences to date on these use of SBRT for 84 mRCC patients with 175 lesions treated by Wang, et al., did find that patients treated to a BED < 115 Gy had higher rates of local failures (HR = 3.45; p = 0.0254) with BED > 98.7 Gy received by 99% of the target volume found to be significant on MVA (HR = 0.12; p = 0.0014) [24] with an overall 1-year LC rate of 91.2% [23]. Notably, among their cohort they also found that patients with spinal metastases had significantly higher rates of local failure (HR = 5.36; p = 0.0041) similar to prior experiences [24]. With regards to fractionation, a prior prospective trial by Ghia, et al., examining outcomes following SBRT for 43 mRCC patients with 47 lesions noted inferior LC on MVA with multiple fractions as compared to single fraction SBRT (HR = 5.26; p = 0.033) [25].

In our analysis, the 1-year and 2-year OS rates were 69.1% and 45.2%, respectively. These results correlate with Thibault, et al., who among 37 mRCC patients with spinal metastases treated with SBRT (median dose of 24 Gy in 2 fractions) noted a similar 1-year OS rate of 64% and found on MVA that oligometastatic disease was associated with improved OS [26]. In their systemic review of mRCC SBRT literature, Zaorsky, et al., noted 1-year OS among all studies examining mRCC patients treated with SBRT ranging from 48.9%-100% with an overall 1-year OS estimate of 86.8% (95% CI: 62.0-99.8%) [19]. Among our cohort, patients with a KPS < 80% had a median OS rate of 6.6 months as compared to 20.3 months for patients with a KPS ³ 80% that was significant on UVA and trended towards significant on MVA. This supports the findings of Motzer et al., and highlights the importance of the inclusion of KPS in the current Motzer score in determining the prognosis for patients with mRCC [27]. Grossman, et al., have previously also noted that the absence of synchronous brain metastases as well as higher initial KPS of 90-100% were associated with improved OS [17].

Also, patients with osseous metastases had inferior median OS (29.5 vs. 15.1 months; p = 0.0023) on both UVA as well as on MVA (HR = 2.16 (95% CI: 1.01–4.62); p = 0.046). The presence of osseous metastases has been previously cited as poor prognostic factor in mRCC as compared to other sites of metastases [28, 29]. A recently published experience on SBRT for 47 oligometastatic RCC patients (with only extracranial metastases) with 88 lesions noted favorable 1- and 2-year OS rates of 93.1% and 84.8% [28]. Similar to our own findings, patients with non-osseous metastases had improved prognosis and had a significantly a longer period of stable disease not requiring systemic therapy compared to those with osseous metastases (HR = 2.21; p = 0.04) [30].

Our results demonstrate a favorable incidence of acute and late toxicities (13.04% of patients total). Of these toxicities, the majority were acute toxicities (66.7%) versus late toxicities (33.3%). Furthermore, all toxicities were Grade 1-2 aside from a one incidence of a Grade 3 toxicity of vomiting. These results are consistent with prior literature, including studies by Ranck et al. as well as Altoos et al., who both found that treatment of mRCC with SBRT was well tolerated with Grade 3 toxicities limited to mucositis and no Grade 4-5 toxicities reported [8,18]. Prior large single institution experiences have also noted favorable acute and late Grade 3 toxicity rates of 1.7% and 2.9%, respectively, and among oligometastatic patients have noted no Grade 3 or greater toxicities [24, 29]. Similarly, the previously cited systematic review and meta-analysis by Zaorsky, et al., found that incidences of Grade 3-4 toxicity for mRCC patients ranged from 0-4% with an estimate of 0.7% (95% CI: 0-2.1%) [19]. However, we did not find that higher BEDs were associated with higher incidences of toxicity, potentially due to under-reporting of incidences of toxicity in the registry as well as heterogeneous treatment sites.

There are notable limitations to this study which merit attention. First, the relatively small sample size of our study limits the generalizability of our findings, especially when analyzing prognostic factors of. Further, no information was available in the registry regarding the extent of disease at the time SBRT was offered (i.e. oligometastatic or polymetastatic), whether one site or how many sites of disease were treated with SBRT, whether patients were treated for symptoms (i.e palliative intent SBRT to the spine for pain) or for durable LC, or whether systemic treatments had been utilized prior to, during, or following SBRT, which severely confounds any conclusions with respect to OS due to the inability to control for the previously mentioned data. As such, any analyses with respect to OS are severely limited Also, dosimetric information of interest such as prescription isodose line, volume of the GTV receiving certain doses, and mean dose to the GTV were unavailable, which limited dosimetric evaluations of LC. Next, as with any multi-institutional retrospective studies, there are numerous avenues by which data may have been improperly obtained or recorded from the various institutions. Similar to other analyses of the RSSPR, toxicity information has the potential to be under-reported and bias the estimate of toxicity to be lower, particularly given the finding that dose escalation was not associated with higher rates of toxicity. Information available regarding LC was limited and only available for select patients. Another limitation is the absence of information on other variables of interest that have previously been established as poor prognostic factors, such as whether patients had synchronous brain metastases. Finally, given the lack of local failures/events (less than 10%) among the cohort, analyses aimed at determining potential benefits from dose escalation were limited.

Conclusion

Our findings demonstrate the efficacy of SBRT in the treatment of mRCC, with favorable patient outcomes and low toxicity rates across a wide variety of clinical settings. Notably, mRCC patients with osseous metastases had significantly poorer OS following SBRT as compared to other sites as did those with poor performance statuses. No significant LC benefit was found with dose escalation with no difference in LC based on treated site. Prospective randomized clinical trials are warranted to further elucidate the role of SBRT for patients with mRCC, with data presented herein highlighting the prognostic significance of non-osseous metastases and good performance statuses (KPS ≥ 80%) as a more favorable cohort of mRCC patients for which trials may want to focus on or at a minimum stratify by.

Ethics: This work has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. Informed consent was obtained for all patients that participated in the RSSPR.