More than 58,000 new cases of kidney cancer were. Single-fraction stereotactic body radiotherapy for spinal metastases from renal cell carcinoma

Transcription

1 J Neurosurg Spine 17: , 2012 Single-fraction stereotactic body radiotherapy for spinal metastases from renal cell carcinoma Clinical article Ehsan H. Balagamwala, B.A., 1 Lilyana Angelov, M.D., 3,4 Shlomo A. Koyfman, M.D., 2 John H. Suh, M.D., 2,4 Chandana A. Reddy, M.S., 2 Toufik Djemil, Ph.D., 2 Grant K. Hunter, M.D., 2 Ping Xia, Ph.D., 2 and Samuel T. Chao, M.D. 2,4 1 Cleveland Clinic Lerner College of Medicine of Case Western Reserve University; and Departments of 2 Radiation Oncology and 3 Neurosurgery, and 4 Rosa Ella Burkhardt Brain Tumor and Neuro-Oncology Center, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio Object. Stereotactic body radiotherapy (SBRT) has emerged as an important treatment option for spinal metastases from renal cell carcinoma (RCC) as a means to overcome RCC s inherent radioresistance. The authors reviewed the outcomes of SBRT for the treatment of RCC metastases to the spine at their institution, and they identified factors associated with treatment failure. Methods. Fifty-seven patients (88 treatment sites) with RCC metastases to the spine received single-fraction SBRT. Pain relief was based on the Brief Pain Inventory and was adjusted for narcotic use according to the Radiation Therapy Oncology Group protocol Toxicity was scored according to Common Toxicity Criteria for Adverse Events version 4.0. Radiographic failure was defined as infield or adjacent (within 1 vertebral body [VB]) failure on follow-up MRI. Multivariate analyses were performed to correlate outcomes with the following variables: epidural, paraspinal, single-level, or multilevel disease (2 5 sites); neural foramen involvement; and VB fracture prior to SBRT. Kaplan-Meier analysis and Cox proportional hazards modeling were used for statistical analysis. Results. The median follow-up and survival periods were 5.4 months (range months) and 8.3 months (range months), respectively. The median time to radiographic failure and unadjusted pain progression were 26.5 and 26.0 months, respectively. The median time to pain relief (from date of simulation) and duration of pain relief (from date of treatment) were 0.9 months (range months) and 5.4 months (range months), respectively. Multivariate analyses demonstrated that multilevel disease (hazard ratio [HR] 3.5, p = 0.02) and neural foramen involvement (HR 3.4, p = 0.02) were correlated with radiographic failure; multilevel disease (HR 2.3, p = 0.056) and VB fracture (HR 2.4, p = 0.046) were correlated with unadjusted pain progression. One patient experienced Grade 3 nausea and vomiting; no other Grade 3 or 4 toxicities were observed. Twelve treatment sites (14%) were complicated by subsequent vertebral fractures. Conclusions. Stereotactic body radiotherapy for RCC metastases to the spine offers fast and durable pain relief with minimal toxicity. Stereotactic body radiotherapy seems optimal for patients who have solitary or few spinal metastases. Patients with neural foramen involvement are at an increased risk for failure. ( Key Words stereotactic body radiotherapy radiosurgery spinal metastasis renal cell carcinoma oncology More than 58,000 new cases of kidney cancer were diagnosed in the US in 2010, the vast majority of which were RCC. 11 In that year, kidney cancer accounted for more than 13,000 cancer-related deaths. The Abbreviations used in this paper: BPI = Brief Pain Inventory; CRT = conventional radiotherapy; CTCAE = Common Terminology Criteria for Adverse Events; CTV = clinical target volume; HR = hazard ratio; KPS = Karnofsky Performance Scale; RCC = renal cell carcinoma; RTOG = Radiation Therapy Oncology Group; SBRT = stereotactic body radiotherapy; VB = vertebral body. incidence of RCC is increasing steadily by 2% 3% per decade; however, the cause of this increase is not known. 6,17 Metastases to the spine are a common occurrence in patients with cancer, and the natural history of cancer includes spinal metastases in more than 40% of patients. 12 The VB is involved in the majority of cases, with subsequent involvement of the epidural space or posterior ele- This article contains some figures that are displayed in color on line but in black-and-white in the print edition. 556

2 Radiosurgery for spinal metastases from renal cell carcinoma ments as the disease progresses. Symptoms of metastasis to the spine can range from pain to motor dysfunction to severe neurological compromise due to spinal cord compression from progressive epidural disease. 8 Treatment of spinal metastases from RCC requires multispecialty management by medical, surgical, and radiation oncology teams. Patients with no neurological compromise benefit most from pain management using narcotic and/or nonnarcotic analgesics, as well as radiation therapy (CRT or SBRT). Surgical intervention is indicated for patients with neurological compromise or mechanical instability requiring tumor debulking, cord decompression, and/or spine stabilization. In the presence of spinal instability, both CRT and SBRT to the spine are contraindicated, and if the patient is a surgical candidate, spine surgery must be undertaken. The role of SBRT in cord compression remains controversial; at the Cleveland Clinic, we do not routinely use SBRT for the treatment of cord compression because of the inherent underdosing of the epidural disease to minimize the dose to the spinal cord. For patients who have neurological compromise but are not surgical candidates, urgent CRT can be used to treat cord compression. 8,22 Renal cell carcinoma has a historical reputation of being a radioresistant tumor. 24,25 It has long been believed that delivery of higher doses of radiation will yield better control of RCC metastases to the spine. However, the radiation tolerance of the spinal cord limits the dose that can be safely delivered to treat these metastases. Stereotactic body radiotherapy enables the delivery of highly conformal large doses of radiation with a sharp dose falloff outside the target volume. It is thought that the delivery of large radiation doses in a single fraction will provide the radiobiological advantage necessary to enhance the tumoricidal effects of radiation 1 and lead to greater relief for these patients. Hence, SBRT has become an important alternative to CRT for the treatment of metastases to the spine in patients without spinal cord compression. The aim of this study was to evaluate the rates of radiographic and pain progression after SBRT for spinal metastases from RCC at the Cleveland Clinic and to identify factors that predict treatment failure. Methods Patient Selection We retrospectively reviewed the charts of all patients with RCC metastases to the spine who underwent treatment with SBRT. Since the inception of the spine SBRT program at the Cleveland Clinic in February 2006, the modality has been performed in more than 300 patients. In this study, we included all patients with a histological diagnosis of primary RCC who had CT and MRI confirmation of spinal metastasis and were treated with SBRT. We use SBRT for the treatment of metastasis to the spine in patients who have a favorable life expectancy and are clinically stable. 4 Patients with diagnoses of 2 or more primary cancers and those who received multiple-fraction SBRT were excluded from this study. At our institution, patients with spinal instability or spinal cord compression are not routinely treated using SBRT, and thus these patients were not included in this study. Instead, these patients are evaluated by the neurosurgery team for possible spinal stabilization or tumor debulking prior to postoperative radiotherapy. Patients with neurological symptoms were included in this study as long as there was no evidence of cord compression. This study was approved by our institution s Institutional Review Board. Clinical and Toxicity Data Clinical records, including all consultation, completion, and follow-up notes were reviewed for each patient. At consultation, each patient was examined by a radiation oncologist and a neurosurgeon and underwent a complete history and physical examination. At consultation, and at each follow-up visit, each patient underwent a comprehensive neurological examination performed either by a neurosurgeon or a radiation oncologist. Each patient s neurological and pain symptoms were characterized. Pain was quantified using the BPI at initial consultation and at 1, 2, and 4 weeks, and then at each subsequent follow-up. Pain scores were adjusted for narcotic usage according to the RTOG protocol 0631 criteria ( Trials/ProtocolTable/StudyDetails.aspx?study=0631). Complete pain relief was recorded if the patient did not have any pain and had successfully discontinued pain medications. Patients with a decrease of at least 3 points on the BPI and/or with stable to decreased use of pain medications were recorded as having achieved partial pain relief. Patients who had an increase in pain medications were not scored as having achieved either partial or complete pain response regardless of a change in pain score. Pain progression was defined as either an increase in pain medications or an increase in pain score of at least 3 points on the BPI with stable use of pain medications. Radiographic failure was defined as infield or adjacent (within 1 VB) failure on follow-up MRI. All toxicity was scored according to the CTCAE version 4.0. Stereotactic Body Radiotherapy All patients were immobilized using a BodyFix stereotactic body fixation system (Elekta) and underwent CT simulation in the supine position and high-definition MRI (1.5-mm slices) of the spine at the involved vertebral level(s). The axial MRI studies were then fused to the planning CT to aid with spinal cord and tumor delineation. Treatment planning was performed using BrainScan (Brainlab) and later iplan (Brainlab), and image guidance was accomplished using the ExacTrac system (Brainlab) during treatment delivery. The CTV was defined as the entire VB and/or all posterior elements depending on tumor involvement. All paraspinal and epidural disease was included in the CTV. Although several grading systems exist to quantify the extent of epidural disease, we do not routinely measure epidural disease. 3,21 No additional margins were added to the CTV. The spinal cord was delineated on axial high-definition MRI studies and was contoured at the level of the CTV with 3-mm cranial and caudal extensions. The following dose constraints are used at our institution: 14-Gy maximum dose to the spi- 557

3 E. H. Balagamwala et al. nal cord with less than 10% of the volume of the spinal cord receiving 10 Gy, and 16-Gy maximum dose to the cauda equina with less than 10% of the volume of the cauda equina receiving 12 Gy. We attempt to achieve a minimum of 90% coverage of the CTV by the prescription dose as long as spinal cord and cauda equina tolerance is met. None of the patients included in this study required dose de-escalation due to inadequate coverage of the CTV by the prescription dose. Figure 1 illustrates a representative treatment plan for a patient treated for spinal metastasis using SBRT. Statistical Analysis Univariate and multivariate analyses were performed to correlate outcomes with the following variables: pres- ence of epidural, paraspinal, single-level, multilevel (2 5 vertebral levels), or diffuse (> 5 vertebral levels) disease; location (cervical/thoracic vs lumbar/sacral); neural foramen involvement; VB fracture prior to SBRT; SBRT dose; KPS score; oligometastatic disease (< 5 metastatic sites); control of systemic disease; and prior radiotherapy at the current site of treatment. Kaplan-Meier analysis and Cox proportional hazards modeling were used for statistical analysis. Analyses were performed using SAS version 9.2 (SAS Institute). Results Patient and Treatment Characteristics From more than 300 patients treated with spinal Fig. 1. Representative treatment plan for a patient with RCC metastases to the spine at T-10 treated using SBRT with a dose of 16 Gy in a single fraction. A: Axial planning CT of T-10 vertebral level. The green area represents the CTV; the red area, the spinal cord; the blue line, the 10-Gy isodose; and the red line, the 16-Gy isodose. B: Sagittal planning CT of the T-10 vertebral level. C: Dose volume histogram (DVH) of the CTV demonstrating that the prescription dose covers at least 90% of the CTV. D: Dose volume histogram of the spinal cord demonstrating that the clinical dose constraints (< 10% of the cord receives 10 Gy and maximum dose is < 14 Gy) are adequately met. 558

4 Radiosurgery for spinal metastases from renal cell carcinoma SBRT at the Cleveland Clinic, 57 patients (88 treatment sites) were eligible for this study based on the criteria outlined earlier. The median age, KPS score, and follow-up were 57.5 years, 80, and 5.4 months, respectively. Four treatment sites (5%) had received prior infield SBRT, and 14 sites (16%) had received prior infield CRT. Twenty-six treatment sites (30%) had a preexisting VB fracture at site of SBRT. The median spine SBRT dose was 15 Gy in 1 fraction (Table 1). Pain Relief After SBRT The BPI was used to assess pain control after SBRT. Pain relief was assessed either according to RTOG 0631 ( Adjusted Pain Relief ) or without adjusting for changes in narcotic usage ( Unadjusted Pain Relief ). Four lesions did not have follow-up data available. Prior to SBRT, at 17 treatment sites (19.3%) the patients were completely pain free. Two lesions (2.3%) that were treated for radiographic progression without any symptoms progressively caused pain at 1 and 3 months after SBRT. Of a total of 88 lesions, 67 (76%) presented with pain and were included in this analysis. The median time to pain relief and duration of pain relief were 0.9 months (range months) and 5.4 months (range months), respectively. TABLE 1: Characteristics of 57 patients undergoing 88 treatments* Parameter Value age (yrs) median 57.5 range KPS score at consult median 80 range sex male 72 (81.8) female 16 (18.2) comorbidities diabetes mellitus 12 (13.6) hypertension 57 (64.8) hyperlipidemia 27 (30.7) osteoporosis 1 (1.1) follow-up (mos) median 5.4 range VB fractures prior to SBRT 26 (29.5) SBRT dose (Gy) median 15 range 8 16 SBRT fractions 1 prior surgery 24 (27.3) prior SBRT 4 (4.5) prior CRT 14 (15.9) * Values represent the number of treatment sites (%) unless otherwise indicated. The rates of unadjusted and adjusted pain relief for patients who presented with pain are shown in Fig. 2. As expected, the rates of pain relief were lower when pain relief was adjusted for narcotic usage and was scored according to RTOG The trends, however, were similar between adjusted pain relief and unadjusted pain relief. The first trend apparent from this analysis is that the overall proportion of patients achieving pain relief peaks at 9 months after SBRT, after which it starts to decline. Similarly, the proportion of patients achieving complete pain relief increases quite dramatically from 15.4% at 2 weeks after SBRT to 66.7% by 9 months. Pain Progression and Radiographic Failure At last follow-up, 19 patients (33%) were alive, correlating with a median survival time of 11.7 months. The 1-year overall survival for all patients was 48.9% (95% CI 34.9% 62.9%). For all 88 metastases, the 1-year progression-free survival (either radiographic or pain progression) was 53.9% (95% CI 40.3% 67.5%) (Fig. 3A). We also individually examined radiographic and pain progression (both adjusted and unadjusted pain progression) after SBRT. The radiographic progression-free survival at 3, 6, 9, and 12 months was 92.3%, 83.3%, 78.1%, and 71.2%, respectively (Fig. 3B). At 3, 6, 9, and 12 months the unadjusted pain-free survival was 83.9%, 78.9%, 72.0%, and 62.6%, respectively (Fig. 3C), and the adjusted pain-free survival was 84.9%, 79.8%, 77.4%, and 67.7%, respectively (Fig. 3D). Univariate analysis demonstrated associations be- Fig. 2. Bar graphs showing patients achieving pain relief after SBRT. Upper: The proportion of patients achieving complete or partial pain relief after adjusting for analgesic pain medication usage. Lower: The proportion of patients achieving complete or partial pain relief without adjusting for analgesic pain medication usage. n = number of patients. 559

5 E. H. Balagamwala et al. Fig. 3. Kaplan-Meier curves. A: Progression-free survival. B: Radiographic progression-free survival. C: Unadjusted pain progression-free survival. D: Adjusted pain progression-free survival. m = months; ssbrt = spine SBRT. tween the occurrence of radiographic failure (infield and/or adjacent) and multilevel disease, neural foramen involvement, and longer time from primary diagnosis to spine SBRT. Multivariate analysis with these factors demonstrated that multilevel spinal disease (HR 3.48 [95% CI ], p = 0.02) and neural foramen involvement (HR 3.44 [95% CI ], p = 0.02) were strong predictors of radiographic failure, while longer duration from primary diagnosis (HR 1.01 [95% CI ], p = 0.01) was a weak predictor of radiographic failure (Table 2). Epidural disease (p = 0.24), paraspinal disease (p = 0.25), dose 16 Gy (p = 0.48), location (cervical/ thoracic vs lumbar/sacral, p = 0.09), single-level disease (p = 0.31), and systemic disease control (p = 0.56) were not associated with radiographic control. We also looked at the clinical variables that were associated with pain progression. Only larger disease volume (HR [95% CI ], p = 0.04) was weakly associated with adjusted pain progression. However, on univariate analysis we found associations between multilevel disease, preexisting VB fracture, and larger disease volume and the occurrence of unadjusted pain progression. Multivariate analysis demonstrated that multilevel spinal disease (HR 2.28 [95% CI ], p = 0.057) and VB fracture (HR 2.40 [95% CI ], p = 0.05) were strong predictors of unadjusted pain progression, whereas disease volume (HR [95% CI ], p = 0.10) was not a statistically significant predictor (Table 2). Epidural disease (p = 0.41), paraspinal disease (p = 0.11), dose 16 Gy (p = 0.14), location (cervical/thoracic vs lumbar/sacral, p = 0.42), single-level disease (p = 0.13), and systemic disease control (p = 0.29) were not associated with unadjusted pain progression. Toxicity and Patterns of Failure Clinical toxicities were graded according to the National Cancer Institute CTCAE version 4.0. Nineteen patients (33%) experienced some toxicity after SBRT. One patient experienced Grade 3 nausea and vomiting. No other Grade 3 or Grade 4 toxicities were observed in any category; specifically, no Grade 3 or 4 neurological toxicities occurred in this cohort of patients. The 2 most common toxicities were Grade 1 fatigue (in 6 treatments) and Grade 1 sensory deficits (in 5 treatments). Table 3 summarizes the other Grade 1 and Grade 2 toxicities observed. Of 88 treatments, 20 treatments (22.7%) were associated with subsequent infield and/or adjacent failure. Six treatments (6.8%) had subsequent infield failure, 2 treatments (2.3%) had adjacent failure, and 12 treatments (13.6%) had both infield and adjacent failure. Of those treatments that had subsequent failure, 16 failed (80%) in the epidural space. Table 4 details the pattern of failure in each of the 20 treatments that failed. Discussion This study presents the results of one of the largest reported cohort of patients with spinal metastases from RCC who underwent treatment with SBRT. In this study, we evaluated our radiographic and pain control outcomes for SBRT in patients with spinal metastases from RCC and identified factors associated with treatment failure. In the majority of SBRT cases, the objective of the treatment is to achieve superior pain control. We present the adjusted and unadjusted pain relief rates to explore 560

6 Radiosurgery for spinal metastases from renal cell carcinoma TABLE 2: Univariate and multivariate analyses* Univariate Analysis Multivariate Analysis Parameter HR 95% CI p Value HR 95% CI p Value radiographic failure epidural disease paraspinal disease dose 16 Gy location (cervical/thoracic vs lumbar/sacral) single-level disease multilevel disease NFI systemic disease control time from diagnosis to treatment pain (unadjusted) progression epidural disease paraspinal disease dose 16 Gy location (cervical/thoracic vs lumbar/sacral) single-level disease multilevel disease vol of disease systemic disease control VB fracture * Values in boldface are significant. Abbreviation: NFI = neural foramen involvement. whether correcting for narcotic pain medications and using a stricter and more objective definition for pain relief has an effect on the overall results for pain relief (Fig. 2). Overall, unadjusted pain response rates were higher than adjusted pain response rates because several patients had an increase in narcotic pain medications after SBRT with a concomitant decrease in pain scores. We also show that the peak pain relief occurs in most patients by 9 months before pain relief rates start declining. In this study, at 9 months after SBRT, the overall unadjusted pain relief rate TABLE 3: Toxicity profile No. of Patients (%) Parameter Grade 1 Grade 2 Grade 3 general fatigue 6 (10.5) 2 (3.5) 0 neurological sensory neuropathy 5 (8.8) 0 0 motor neuropathy 2 (3.5) 1 (1.8) 0 gastrointestinal nausea 3 (5.3) 3 (5.3) 1 (1.8) vomiting 2 (3.5) 1 (1.8) esophagitis 0 1 (1.8) 0 No. of Treatments (%) pain flare* 7 (8.0) * Pain flare was not graded. was 86.6% and the proportion of patients who achieved complete pain relief was 73.3%. These results compare favorably with those of Gerzsten et al., 5 who reported an overall 89% unadjusted pain relief rate, and those of Nguyen et al., 18 who reported a complete unadjusted pain relief rate of 64.3% at 9 months (Table 5). Interestingly, our series showed a slower rate of unadjusted complete pain response 2 weeks after SBRT than the study of Nguyen et al. (20.5% vs 39.0%). This is likely because we assessed pain relief from the date of simulation rather than date of treatment, which is an important consideration since many of these patients are also candidates for CRT, which can usually be started within a day after simulation. For a more detailed analysis of the outcomes of patients with RCC metastases to the spine that were treated with SBRT or CRT, we refer the interested reader to the analysis of the topic by Hunter et al. 10 In summary, the authors demonstrated that SBRT offers more durable pain relief than CRT. Furthermore, our SBRT pain response rates compare favorably with those of RTOG 9714, 9 which showed a 65% unadjusted pain relief rate with 8 Gy/1 fraction and 66% unadjusted pain relief with 30 Gy/10 fractions. Importantly, RTOG 9714 studied breast and prostate cancers, which are more radiosensitive than RCC, the histology in this study. Similarly, Lee et al. 16 conducted a prospective Phase II trial in patients with bone metastases from RCC. They showed an 83% pain relief rate, and 48% of their patients did not have a concomitant increase in analgesic medications. With SBRT for bone metastases from RCC, we show that 561

7 E. H. Balagamwala et al. TABLE 4: Patterns of failure* Tumor No. Treatment Site Level TTF (mos) Type of Failure Pattern of Failure 1 VB, PS sacrum 5.85 infield VB, NF 2 VB, PS T infield & adjacent VB, epi, PE 3 VB, PE, PS T11 L infield VB, epi 4 VB T infield & adjacent VB, epi 5 VB, PS L adjacent VB, epi, PS 6 VB, PS L adjacent VB 7 VB, PS sacrum infield VB, NF 8 VB T infield & adjacent VB, PE, PS, NF 9 VB, PE, PS L infield & adjacent VB, PE, epi, PS 10 VB, PE T infield & adjacent VB, PE, epi 11 VB T infield epi 12 VB, PE, PS L infield VB, PE, epi 13 VB, PS L infield & adjacent VB, epi 14 VB, PE, PS T infield & adjacent epi, PS, NF 15 VB, PE C infield & adjacent epi 16 VB, PE T infield & adjacent VB, epi 17 VB, PE, PS T infield & adjacent epi 18 VB L infield VB, epi 19 VB, PS L infield & adjacent VB, epi, PS 20 VB, PE T infield & adjacent VB, epi * epi = epidural; NF = neural foramen; PE = posterior elements; PS = paraspinal; TTF = time to failure. the adjusted complete pain relief rate at 9 months was 66.7%, indicating that a large proportion of patients are able to discontinue pain medications and achieve complete pain relief. To better select patients for SBRT for spinal metastases, we were interested in studying which patient and treatment variables predicted radiographic failure or pain progression. Univariate and multivariate analyses indicated that multilevel disease and neural foramen involvement are strong predictors of radiographic failure, and VB fractures and multilevel disease are predictors of unadjusted pain progression. None of the variables studied correlated with adjusted pain progression. Unsurprisingly, neural foramen involvement was found to be a predictor of infield and adjacent radiographic failure. This is likely because it is a proxy for epidural failure, which has been shown to be an important site of failure due to difficulty in adequately dosing epidural disease. Although the extent of epidural disease is not evaluated routinely in our practice, we intend to study whether the extent of epidural disease (as defined by formal grading systems 3,21 ) is associated with neural foramen involvement. Furthermore, patients with multilevel disease were at higher risk of having radiographic progression and pain failure after SBRT. This is likely because their disease may inherently be more resistant. Additionally, since SBRT is targeted only to the involved vertebrae, micrometastases in adjacent vertebral levels are missed and, hence, may contribute to the higher radiographic failure and pain progression rates in patients with multilevel disease. Recently, we studied factors that TABLE 5: Comparison of studies evaluating the use of SBRT for the treatment of RCC metastases to the spine* Authors & Year Gerszten et al., 2005 Nguyen et al., 2010 Type of Study No. of Patients/ No. of Tumors Prescription Dose/Fx retrospective 48/ Gy at 80% isodose (mean 16 Gy)/1 Fx retrospective 48/55 30 Gy/5 Fx (13 cases), 27 Gy/3 Fx (34 cases), 24 Gy/1 Fx (8 cases) current study retrospective 57/ Gy at 100% isodose (median 15 Gy)/1 Fx Median Follow-Up/ Median Survival (mos) Outcomes 37/NA 1-mo pain improvement in 95%; long-term pain improvement in 89% 13.1/22 1-yr actuarial PFS in 82.1%; 44% pain free at 1 mo; 52% pain free at 12 mos 5.4/8.3 actuarial median time to radiographic failure: 26.5 mos; actuarial median time to pain progression: 26.0 mos; 73.3% pain free at 9 mos; max pain relief at 9 mos * Fx = fraction; max = maximum; NA = not available; PFS = progression-free survival. 562

8 Radiosurgery for spinal metastases from renal cell carcinoma predicted recurrence at one VB above or below the index lesion after SBRT. 13 We found that paraspinal disease (HR 2.82, p = 0.02) and dose < 16 Gy (HR 2.69, p = 0.05) were significant predictors. This study shows that multilevel disease and neural foramen involvement are also strong predictors of radiographic failure (defined in the current study as either infield or adjacent VB failure). The strengths of this study include that it reports one of the largest cohort of patients with RCC metastases to the spine who underwent treatment with SBRT. Additionally, we only included patients who were treated with single-fraction SBRT for the spinal metastases, and we studied variables that have not been reported previously in the literature. This study also has several limitations. Because it is a retrospective review, it is not only subject to selection bias but also to missing data points. However, we do monitor these patients closely, and at times, we call patients to conduct pain surveys. Furthermore, given the poor prognosis of patients with spinal metastases, only a fraction of patients lived 6 months after SBRT (21.6%). This may have introduced a bias in our pain relief rates as patients with better pain relief may have more sensitive disease and hence had longer overall survival. Renal cell carcinoma is historically recognized as a radio- and chemoresistant tumor, 24,25 and patients with RCC tend to have a poor prognosis. Approximately onethird of patients with RCC present initially with bone metastases. 26 Studies have shown that patients with bone metastases from RCC 2,15,20 and poor performance status 26 have a shorter overall survival. Survival is especially short in patients who have vertebral and extraosseous metastases. 14 The introduction of tyrosine kinase inhibitors in the last few years has not only increased the overall survival of patients with RCC, but it has also decreased the incidence of brain metastases. 23 Early studies showed that dose escalation for the treatment of radioresistant spine metastases may be useful. 7,19 The introduction of SBRT in the past few years has raised the possibility of utilizing the radiobiological advantages of radiosurgery in achieving better pain and tumor growth control, 1 especially for radioresistant tumors. Recent data from our center have shown that a dose-response relationship exists for spine metastases from radiosensitive cancers; however, this dose-response relationship is counterbalanced when both radiosensitive and radioresistant metastases are included in the analysis (unpublished data). This raises the interesting possibility of dose escalation for the treatment of radioresistant spine metastases using SBRT, and we have recently started treating select patients with RCC spine metastasis using 18 Gy in a single fraction. Conclusions In this study, we demonstrate that SBRT for RCC metastases to the spine is safe and provides fast and durable pain relief. We show that SBRT may be more optimal for patients with solitary or few spine metastases, and those with neural foramen involvement are more likely to experience radiographic failure and pain progression. It appears that SBRT for spinal metastases from RCC provides optimal progression-free survival, especially in patients who have a long life expectancy, corroborating the results of the SBRT recursive partitioning analysis index by Chao et al. 4 We believe that future studies evaluating pain response after radiotherapy should correct for narcotic analgesic use. Since the rates of pain relief were similar in our study in comparison with other studies with more radiosensitive tumors, we believe that cases of RCC should be included in future clinical trials evaluating the efficacy of SBRT. Disclosure Dr. Angelov receives honoraria from Brainlab, Dr. Suh is a consultant for Abbott Oncology, and Dr. Djemil receives support of non study related clinical or research effort from Brainlab. Author contributions to the study and manuscript preparation include the following. Conception and design: Chao, Balagamwala, Angelov, Koyfman, Suh, Xia. Acquisition of data: Balagamwala, Hunter. Analysis and interpretation of data: Chao, Balagamwala, Koyfman, Reddy, Hunter, Xia. Drafting the article: Balagamwala. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Chao. Statistical analysis: Balagamwala, Reddy. Administrative/technical/material support: Chao, Suh, Djemil, Xia. Study supervision: Chao, Suh, Xia. References 1. Balagamwala EH, Chao ST, Suh JH: Principles of radiobiology of stereotactic radiosurgery and clinical applications in the central nervous system. Technol Cancer Res Treat 11:3 13, Beuselinck B, Oudard S, Rixe O, Wolter P, Blesius A, Ayllon J, et al: Negative impact of bone metastasis on outcome in clearcell renal cell carcinoma treated with sunitinib. Ann Oncol 22: , Bilsky MH, Laufer I, Fourney DR, Groff M, Schmidt MH, Varga PP, et al: Reliability analysis of the epidural spinal cord compression scale. Clinical article. J Neurosurg Spine 13: , Chao ST, Koyfman SA, Woody N, Angelov L, Soeder SL, Reddy CA, et al: Recursive partitioning analysis index is predictive for overall survival in patients undergoing spine stereotactic body radiation therapy for spinal metastases. Int J Radiat Oncol Biol Phys 82: , Gerszten PC, Burton SA, Ozhasoglu C, Vogel WJ, Welch WC, Baar J, et al: Stereotactic radiosurgery for spinal metastases from renal cell carcinoma. J Neurosurg Spine 3: , Gupta K, Miller JD, Li JZ, Russell MW, Charbonneau C: Epidemiologic and socioeconomic burden of metastatic renal cell carcinoma (mrcc): a literature review. Cancer Treat Rev 34: , Halperin EC, Harisiadis L: The role of radiation therapy in the management of metastatic renal cell carcinoma. Cancer 51: , Harel R, Angelov L: Spine metastases: current treatments and future directions. Eur J Cancer 46: , Hartsell WF, Scott CB, Bruner DW, Scarantino CW, Ivker RA, Roach M III, et al: Randomized trial of short- versus longcourse radiotherapy for palliation of painful bone metastases. J Natl Cancer Inst 97: , Hunter GK, Balagamwala EH, Koyfman SA, Bledsoe T, Sheplan LJ, Reddy CA, et al: The efficacy of external beam radiotherapy and stereotactic body radiotherapy for painful spinal metastases from renal cell carcinoma. Practi Radiati Oncol [epub ahead of print], Jemal A, Siegel R, Xu J, Ward E: Cancer statistics, CA Cancer J Clin 60: ,

9 E. H. Balagamwala et al. 12. Klimo P Jr, Schmidt MH: Surgical management of spinal metastases. Oncologist 9: , Koyfman SA, Djemil T, Burdick MJ, Woody N, Balagamwala EH, Reddy CA, et al: Marginal recurrence requiring salvage radiotherapy after stereotactic body radiotherapy for spinal metastases. Int J Radiat Oncol Biol Phys 83: , Kume H, Kakutani S, Yamada Y, Shinohara M, Tominaga T, Suzuki M, et al: Prognostic factors for renal cell carcinoma with bone metastasis: who are the long-term survivors? J Urol 185: , Kusuda Y, Miyake H, Behnsawy HM, Fukuhara T, Inoue TA, Fujisawa M: Prognostic prediction in patients with metastatic renal cell carcinoma treated with sorafenib based on expression levels of potential molecular markers in radical nephrectomy specimens. Urol Oncol [epub ahead of print], Lee J, Hodgson D, Chow E, Bezjak A, Catton P, Tsuji D, et al: A phase II trial of palliative radiotherapy for metastatic renal cell carcinoma. Cancer 104: , Lipworth L, Tarone RE, McLaughlin JK: The epidemiology of renal cell carcinoma. J Urol 176: , Nguyen QN, Shiu AS, Rhines LD, Wang H, Allen PK, Wang XS, et al: Management of spinal metastases from renal cell carcinoma using stereotactic body radiotherapy. Int J Radiat Oncol Biol Phys 76: , Onufrey V, Mohiuddin M: Radiation therapy in the treatment of metastatic renal cell carcinoma. Int J Radiat Oncol Biol Phys 11: , Patil S, Figlin RA, Hutson TE, Michaelson MD, Négrier S, Kim ST, et al: Prognostic factors for progression-free and overall survival with sunitinib targeted therapy and with cytokine as first-line therapy in patients with metastatic renal cell carcinoma. Ann Oncol 22: , Ryu S, Rock J, Jain R, Lu M, Anderson J, Jin JY, et al: Radiosurgical decompression of metastatic epidural compression. Cancer 116: , Steinmetz MP, Mekhail A, Benzel EC: Management of metastatic tumors of the spine: strategies and operative indications. Neurosurg Focus 11(6):e2, Verma J, Jonasch E, Allen P, Tannir N, Mahajan A: Impact of tyrosine kinase inhibitors on the incidence of brain metastasis in metastatic renal cell carcinoma. Cancer 117: , Weber K, Doucet M, Kominsky S: Renal cell carcinoma bone metastasis elucidating the molecular targets. Cancer Metastasis Rev 26: , Weber KL, Doucet M, Price JE: Renal cell carcinoma bone metastasis: epidermal growth factor receptor targeting. Clin Orthop Relat Res (415 Suppl):S86 S94, Yuasa T, Urakami S, Yamamoto S, Yonese J, Saito K, Takahashi S, et al: Treatment outcome and prognostic factors in renal cell cancer patients with bone metastasis. Clin Exp Metastasis 28: , 2011 Manuscript submitted March 23, Accepted August 28, Please include this information when citing this paper: published online September 28, 2012; DOI: / SPINE Address correspondence to: Samuel T. Chao, M.D., Department of Radiation Oncology, Desk T28, Taussig Cancer Institute, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, Ohio chaos@ccf.org. 564