Practical Radiation Oncology (2012) 2, e95–e100
www.practicalradonc.org
Original Report
The efficacy of external beam radiotherapy and stereotactic body radiotherapy for painful spinal metastases from renal cell carcinoma Grant K. Hunter MDa , Ehsan H. Balagamwala BAa , Shlomo A. Koyfman MDa , Trevor Bledsoe BSa , Lawrence J. Sheplan MDa , Chandana A. Reddy MSa , Samuel T. Chao MDa , Toufik Djemil PhDa , Lilyana Angelov MDb , Gregory M.M. Videtic MDa,⁎ a
Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio Department of Neurosurgery, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio
b
Received 18 November 2011; revised 9 January 2012; accepted 25 January 2012
Abstract Purpose: Palliative radiotherapy is routinely used to treat painful spinal metastases from renal cell carcinoma (RCC). Conventionally planned external beam radiotherapy (CRT) has been standard, with high-dose stereotactic body radiotherapy (SBRT) becoming increasingly common given the radioresistant nature of RCC. We compared the efficacy and durability of pain relief produced by these 2 modalities. Methods and Materials: Patients with painful spinal metastases from RCC treated from 2002-2010 were included. Response was defined similar to the Radiation Therapy Oncology Group 0631 protocol: complete response (CR) being resolution of pain without increased narcotics; partial response (PR) included patients with an incomplete reduction in pain without increased narcotics. Patients who experienced a CR or PR were coded as having pain relief, while those with persistent pain or additional narcotics requirements were coded as failures. Achievement of pain relief was analyzed using competing risk analysis with death as the competing event. Time to pain relief was plotted using cumulative incidence analysis. Results: A total of 110 patients (34 CRT; 76 SBRT) were included. Median follow-up was 4.3 months (range, 0.2-38). Median Karnofsky performance score was higher for patients treated with SBRT compared with CRT (80 vs 70; P = .0004). Overall pain response rates were 68% for CRT and 62% for SBRT, with respective CR and PR rates of 12% and 56% for CRT, and 33% and 29% for SBRT (P = .01). Median time to pain relief was 0.6 weeks for CRT versus 1.2 weeks for SBRT (P = .29). For patients who achieved pain relief (n = 79), median duration was 1.7 months for CRT versus 4.8 months for SBRT (P = .095). On univariate analysis no factors were significantly related to pain relief.
Conflicts of interest: None. ⁎ Corresponding author. Department of Radiation Oncology, 9500 Euclid Ave, Desk T-28, Cleveland, OH 44195. E-mail address: [email protected] (G.M.M. Videtic). 1879-8500/$ – see front matter © 2012 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.prro.2012.01.005
e96
G.K. Hunter et al
Practical Radiation Oncology: October-December 2012
Conclusions: CRT was not statistically different than SBRT for pain relief in symptomatic spine metastases from RCC and should be used as first line treatment. The appropriate use of SBRT in this population merits prospective study. © 2012 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction Around 61,000 cases of kidney cancer are expected to be diagnosed in 2011 with renal cell carcinoma (RCC) constituting 92% of the total number of malignancies. 1 Patients with RCC frequently present with metastatic disease. 2 The second most common site of metastatic disease (following pulmonary metastases) is bone, occurring in 5% of all patients with RCC, and the spine is the most common site of bone involvement. 3-5 Patients who present with spinal metastatic disease commonly have severe back pain as a presenting symptom and are routinely treated for palliation. 6,7 The general treatment options employed to treat malignant spine metastases include the following: surgery, including kyphoplasty and vertebroplasty; radiation therapy, which includes conventionally planned and delivered external beam radiation or stereotactic body radiotherapy (SBRT); and medical management that includes narcotics, nonsteroidal antiinflammatory drugs, or both; or some combination of therapies. Radical surgery, either in the form of laminectomy or corpectomy, is often employed when patients present with malignant spinal cord compression or when the patient has a pathologic vertebral body fracture for stabilization purposes. Radiation therapy is used for patients postoperatively after the metastatic osseous lesion is resected to eliminate microscopic disease, or for those patients who are medically unfit for surgery or do not warrant surgical intervention as deemed by the surgical oncologist. 8 Radiation therapy for spinal metastases has traditionally been delivered with conventional radiation therapy (CRT), which entails single or multiple fractions of daily radiation delivered to the involved region of the spine with a margin of 1 vertebral body above and below the affected area. Two-dimensional field design with 1 or 2 beams is commonly used, and the dose is prescribed to a point. Recent technologic advances in both immobilization and planning techniques have created the ability to accurately and safely deliver high doses of radiation to the spine while respecting spinal cord tolerance via SBRT. 9,10 While RCC has been shown in preclinical models and in older clinical series to be relatively radioresistant, SBRT offers an attractive strategy to overcome this radioresistance with the use of high doses of single fraction radiotherapy. 11,12 Though this hypothesis is likely contributing to the increasing use of SBRT in this
particular patient population, there is a paucity of comparative efficacy data between CRT and the more costly and labor intensive SBRT in this setting for any histology of metastasis. This is currently being investigated in the on-going Radiation Therapy Oncology Group (RTOG) 0631 phase II/III randomized trial, with patients with RCC included in the phase III component. 13 The primary objective of this study is to retrospectively compare pain relief outcomes for patients with spinal metastases from RCC treated (in a nonrandomized fashion) with these 2 modalities.
Methods and materials Patients with radiographically evident, painful spine metastases from RCC (C1-sacrum), who were treated with either CRT or SBRT between 2002 and 2010, were included in this Institutional Review Board approved retrospective, nonrandomized review. Of the SBRT patients, only those treated with single fraction SBRT were included. There were no dose or fractionation criteria for CRT patients. Patients with prior surgery or external beam radiotherapy to the treated area were also included. CRT was delivered using a single posterior-anterior beam or anterior-posterior/posterior anterior technique prescribed to a depth, or the isocenter, respectively. Simulations were done with either fluoroscopic or computed tomographic (CT)-based imaging. For CRT planning, the target was the involved vertebral body (bodies) with an additional margin of 1 vertebral body superior and inferior to the lesion(s). The choice of dose and fractionation was at the discretion of the treating radiation oncologist of which there were 12 over the study time period. Patients treated with SBRT were immobilized using an Elekta BodyFix stereotactic body frame system (Elekta AB, Stockholm, Sweden) and underwent CT simulation in the supine position. The CT simulation entailed obtaining axial CT images with 1.5-mm slice thickness. A high definition magnetic resonance imaging (MRI) scan (1.5-mm slices) of the spine at the involved vertebral level(s) was also obtained. Axial MRI images were then fused to the planning CT to aid with spinal cord and tumor delineation. Treatment planning was performed using iPlan (BrainLab, Munich, Germany) and image guidance was accomplished using the ExacTrac system (Novalis, Palo Alto, CA) during treatment delivery. The clinical target volume (CTV) was defined as the entire vertebral body or entire posterior vertebral elements
Practical Radiation Oncology: October-December 2012
depending on tumor involvement. Paraspinal and epidural disease were routinely included in the CTV, with 62% of the patients in the SBRT cohort having some epidural involvement. In the SBRT cohort included in our review, 17 patients or 22% had compression fractures evident. No patients were treated postoperatively (including postkyphoplasty) in the SBRT cohort. Our institutional practice is to not use SBRT for patients with frank spinal cord compression. Noncontinuous vertebral bodies were allowed to be treated, with 9 patients (15% of SBRT patients) having noncontiguous bodies treated on the same day and 8 patients (14% of SBRT patients) having noncontiguous bodies treated on different days. No additional margins were added to the CTV to account for setup error. The spinal cord was delineated on axial high definition MRI images and was contoured at the level of the CTV with 3-mm cranial and caudal extensions. The following dose constraints were used as our institution standards: 14 Gy maximum point dose and V10 Gy b10% for the spinal cord, and 16 Gy maximum point dose and V12 Gy b10% for the cauda equina. Figure 1A and B depicts 2 representative treatment plans; one for CRT (Fig 1A) and one for SBRT (Fig 1B) with typical isodose coverage included. We characterized pain relief as it was defined by RTOG 0631, which includes both an assessment of pain response and an assessment of narcotics usage. The RTOG protocol specifies its definitions of complete response (CR) and partial response (PR) with the incorporation of the Numerical Rating Pain Scale (NRPS) in the definition of each response. 14 CR was graded if there was a pain score of 0 at the index site or if there was complete resolution of pain without an increase in narcotics usage. PR included patients who had a reduction in pain without an increase in narcotics usage. The PR definition in the RTOG protocol states that the pain score must be reduced by 3 points or more without increased narcotics. This rigorous definition of pain response (utilizing NRPS scores and narcotics usage) is clearly able to be done when data are collected in a prospective fashion, but is difficult if not impossible when attempting to use the same scale in a retrospective manner when the NRPS score was not assessed at the time of follow-up. For our study we used the above scale to assess the pain response of each patient, incorporating the NRPS as the RTOG does if such information was available for the patient. We adapted the RTOG scale (based on the NRPS) for those of our patients who did not have numerical scores coded. All patients did have qualitative assessments of their pain and narcotic usage documented at baseline and follow-up that was used for pain relief assessment in the absence of NRPS quantitative assessments. All patients were asked at follow-up a qualitative assessment of their pain (ie, mild, moderate, or severe) and if it had improved compared with before
External beam/SBRT for RCC of the spine
e97
Figure 1 Pictures of a typical plan utilizing a (A) single posterior-anterior beam for treatment of T11-L1 to 3000 cGy/10 fractions above a comparative stereotactic body radiotherapy (SBRT) plan (B) for treatment of L2 to 1600 cGy/1 fraction with representative isodose lines (100%, 90%, and 70% for the conventional treatment and 100%, 90% and 65% for SBRT).
treatment. Using this qualitative scoring along with the patient's narcotics usage data we were able to assign a corresponding pain score for comparison. Patients who underwent SBRT all had NRPS numbers obtained at the time of follow-up to help score their pain response. Patients who experienced a CR or PR were coded as having pain relief while patients with persistent pain or who required an increase in narcotics were coded as failures of treatment. Achievement of pain relief was analyzed using competing risk analysis with death as the competing
e98
G.K. Hunter et al
Table 1
Practical Radiation Oncology: October-December 2012
Patient and treatment characteristics
Variables
CRT
SBRT
P value
Median age, y (range)
62 (43-81) 1.75 (0.2-28) 12:22 70 (20-100) 8 (23%) 6 (18%)
57 (41-80) 4.95 (0.3-38) 13:63 80 (50-90) 0 12 (16%)
.16
N/A .58
8 (24%) 20 (8-30)
45 (59%) 15 (8-16)
.98 b.0001
5 (1-10)
1
b.0001
1 (0-6)
6 (1-19)
.0001
Median follow-up, mo (range) Male:Female Median KPS (range) Palliative surgery Prior radiation treated to spine Multiple sites treated Median total dose, Gy (range) Median no. of fractions (range) Time from sim to tx, d (range)
.07 .04 .0004
CRT, conventional radiation therapy; KPS, Karnofsky performance score; N/A, not applicable; SBRT, stereotactic body radiotherapy; sim, simulation; tx, treatment.
event. Time to pain relief was plotted using cumulative incidence analysis. Fine and Gray regression was used to determine if any of the following variables were significantly associated with pain relief: Karnofsky performance score, age, sex, prior palliative spinal surgery, prior CRT to the same site, number of fractions (for CRT), and total dose. A P value of b .5 was used to determine if a variable was significantly associated with pain relief.
Results A total of 110 patients (34 CRT; 76 SBRT) were included in this study. Patient and treatment characteristics are detailed in Table 1. The overall median follow-up
Figure 2 Incidence of pain relief after conventional radiation therapy (CRT) or stereotactic body radiotherapy (SBRT) for painful spinal renal cell carcinoma metastases.
time was 4.3 months (range, 0.2-38). Patients treated with SBRT had longer median follow-up than those treated with CRT (4.95 vs 1.75 months; P = .07). Median Karnofsky performance score was significantly higher for the SBRT group (80 vs 70; P = .0004). The number of fractions used to deliver CRT varied between 1 and 10, with a median of 5 fractions. The total dose delivered by CRT varied from 8 Gy to 30 Gy with a median of 20 Gy. The total dose was significantly higher in patients treated with CRT compared with SBRT (20 Gy vs 15 Gy; P b .0001) and the total number of fractions was significantly greater in patients treated with CRT (5 vs 1; P b .0001). Treatment doses and fractionations used are detailed in Table 2. While a small number of both groups of patients had received prior radiation therapy to the treated area, only patients in the CRT group had prior decompressive surgery (23%). The time from simulation to treatment was significantly longer for patients treated with SBRT (median of 6 days vs 1 day, P b .0001).
Table 2 Dose and fractionation schemes used for palliative treatments CRT
SBRT
Dose/fractions
No. pts (%)
Dose/fractions
No. pts (%)
20 Gy/5 fxs 30 Gy/10 fxs 8 Gy/1 fx 15 Gy/5 fxs 27.3 Gy/7 fxs 21 Gy/7 fxs 16 Gy/4 fxs 12 Gy3 fxs
13 12 3 2 1 1 1 1
16 Gy/1 fx 15 Gy/1 fx 14 Gy/1 fx 13 Gy/1 fx 12 Gy/1 fx 10 Gy/1 fx 8 Gy/1 fx
32 (42) 12 (17) 14 (18) 4 (5) 9 (12) 4 (5) 1 (1)
(38) (35) (9) (6) (3) (3) (3) (3)
CRT, conventional radiation therapy; fxs, fraction(s); pts, patients; SBRT, stereotactic body radiotherapy.
Figure 3 Duration of pain relief for patients with pain relief after radiation therapy (RT). CRT, conventional radiation therapy; SBRT, stereotactic body radiotherapy; SD, standard deviation.
Practical Radiation Oncology: October-December 2012 Table 3 Fine and Gray competing risk regression for pain relief: univariate analysis Variable
P value
HR
95% CI
Univariate analysis CRT vs SBRT Age (cont) KPS (cont) Dose (cont) Gender (female vs male) Prior EBRT # Fx (CRT only) Palliative surgery (CRT only)
.33 .38 .4 .11 .73 .22 .051 .46
1.28 0.99 0.99 1.04 0.92 1.37 1.14 0.70
0.78-2.08 0.96-1.02 0.97-1.01 0.99-1.09 0.57-1.48 0.83-2.29 1-1.31 0.27-1.82
CI, confidence interval; CRT, conventional radiation therapy; EBRT, external beam radiotherapy; # Fx, number of fractions; HR, hazard ratio; KPS, Karnofsky performance score; SBRT, stereotactic body radiotherapy.
The overall pain response rates were similar for the CRT and SBRT groups (68% vs 62%, respectively), with CR rates being lower for the CRT group (12% vs 33%) and PR rates being higher for the CRT group (56% vs 29%, P = .01). The median time to achieve pain relief was shorter for the CRT group (0.6 vs 1.2 weeks; P = .29). The cumulative incidence for pain control by treatment is depicted in Fig 2, which shows a nonsignificant difference between the 2 groups. For those patients who achieved pain relief (n = 79), the median duration of pain relief was nonsignificantly shorter for the CRT group (1.7 vs 4.8 months; P = .095) (Fig 3). For the few patients whose pain recurred after an initial pain response, the time to recurrence was similar for the CRT and SBRT groups (4.7 vs 5.0 months; P = .58). On competing risk regression univariate analysis there were no factors that were significantly associated with pain relief (Table 3).
Discussion To our knowledge, this study is the first to compare pain relief outcomes for patients with spinal metastases from RCC who were treated with CRT or SBRT. Overall, our data suggest that these 2 modalities provide equivalent rates of pain relief, which argues against the belief that conventional radiotherapy is ineffective for this histology. These data are in line with previously published studies that report pain relief rates of 77%-83% in patients with RCC metastases. 15,16 Another interesting finding from this dataset was that CRT produced a pain response at least as quickly as SBRT. Time to pain relief is an extremely significant endpoint in this patient population with incurable disease, especially when pain is the most common reason they temporarily discontinue their systemic treatments in favor of focal palliative radiotherapy. Our CRT data are consistent with older series that report that by 2 weeks
External beam/SBRT for RCC of the spine
e99
post-conventional radiotherapy, nearly 60% of patients achieved some pain relief. 17 On the other hand, while SBRT may be slower to produce a symptomatic response, the pain relief appears more durable than CRT. Although in our series the duration of pain relief was not statistically significantly different comparing CRT to SBRT, the SBRT patients had a higher rate of complete response. While the 12% CR rate in the CRT arm is identical to the CR rate at 8 weeks post treatment on the RTOG 97-14 phase III randomized study, SBRT led to a nearly threefold increase in complete pain relief. Both of these observations can be explained by the hypothesis that high-dose single fraction SBRT delivers a higher radiobiologic effective dose, thereby enhancing tumor cell kill and leading to enhanced responses that last longer. Taken together, these findings highlight an important area of uncertainty in the management of these patients. SBRT is an application of advanced technology that may have some advantages over CRT; however, the right setting in which SBRT is appropriate remains an active area of investigation as it is more costly and labor intensive treatment modality compared with CRT. While some argue that its radioresistant nature warrants SBRT to be used as first line in spinal metastases from RCC, our data suggest that a more nuanced treatment algorithm should be employed. CRT appears to be an appropriate first line therapy for many patients with RCC spinal metastases, while SBRT may be reserved for those with a history of prior CRT, or perhaps those with an extended life expectancy where durability of response is a critical measure of successful treatment. To this point, Chao et al 18 recently published a recursive partitioning analysis for patients who underwent spine SBRT, including patients with RCC, which identified 3 distinct subgroups of patients with significantly disparate overall survival. This provides a helpful first step in tailoring the use of SBRT to patients who may benefit from it most. There are several limitations to this study. Its retrospective nature, coupled with the varying quality of follow-up information, especially in the older CRT cohort, introduces recall and follow-up biases that may impact some of the study results, especially vis-à-vis durability of response data. That 23% of patients in the CRT group had prior surgery can confound the baseline pain and treatment response data in those patients as well. Another limitation to our study is the retrospective pain assessment. As previously described, in order to best compare the 2 modalities the assignments of pain response were either using the NRPS scores that were collected at the time of follow-up or based on the patient's qualitative description of pain relief with an assessment of narcotics usage. The conclusions drawn would be stronger if all patients had NRPS scores taken at the time of follow-up to more rigorously assess pain response. Finally, the small numbers of patients in this study prohibit any firm
e100
G.K. Hunter et al
conclusions to be drawn. Despite these shortcomings, the study achieves its goal of providing the first comparative analysis of these 2 modalities in this patient population and provides data that at least reaffirm the knowledge gap. This study also argues that the question of CRT versus SBRT is as relevant for patients with RCC as it is for other histologies currently included in the ongoing RTOG 0631 study. We would therefore advocate that these patients be included in that study, and future investigations should continue to focus on identifying which patients with RCC are most likely to benefit from SBRT.
Conclusions Both CRT and SBRT provide effective palliation of symptomatic spine metastases from RCC. CRT as a treatment modality requires less intensive treatment planning and quality assurance, and is less costly and was not statistically different compared with SBRT. The use of SBRT in this population merits prospective study. Conventional treatment remains appropriate as the first line treatment for palliation with SBRT being appropriate for selected patients.
References 1. American Cancer Society. Cancer Facts & Figures 2011. Available at: http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/ documents/document/acspc-029771.pdf. Accessed February 7, 2012. 2. Levy DA, Slaton JW, Swanson DA, Dinney CP. Stage specific guidelines for surveillance after radical nephrectomy for local renal cell carcinoma. J Urol. 1998;159:1163-1167. 3. Kume H, Kakutani S, Yamada Y, et al. Prognostic factors for renal cell carcinoma with bone metastasis: who are the long-term survivors? J Urol. 2011;185:1611-1614. 4. Swanson DA, Orovan WL, Johnson DE, Giacco G. Osseous metastases secondary to renal cell carcinoma. Urology. 1981;18: 556-561.
Practical Radiation Oncology: October-December 2012 5. Shvarts O, Lam JS, Kim HL, Han KR, Figlin R, Belldegrun A. Eastern Cooperative Oncology Group performance status predicts bone metastasis in patients presenting with renal cell carcinoma: implication for preoperative bone scans. J Urol. 2004;172:867-870. 6. Hartsell WF, Scott CB, Bruner DW, et al. Randomized trial of shortversus long-course radiotherapy for palliation of painful bone metastases. J Natl Cancer Inst. 2005;97:798-804. 7. Langdon J, Way A, Heaton S, Bernard J, Molloy S. The management of spinal metastases from renal cell carcinoma. Ann R Coll Surg Engl. 2009;91:649-652. 8. Faul CM, Flickinger JC. The use of radiation in the management of spinal metastases. J Neurooncol. 1995;23:149-161. 9. Hamilton AJ, Lulu BA, Fosmire H, Stea B, Cassady JR. Preliminary clinical experience with linear accelerator-based spinal stereotactic radiosurgery. Neurosurgery. 1995;36:311-319. 10. Ryu SI, Chang SD, Kim DH, et al. Image-guided hypo-fractionated stereotactic radiosurgery to spinal lesions. Neurosurgery. 2001;49: 838-846. 11. Wronski M, Maor MH, Davis BJ, Sawaya R, Levin VA. External radiation of brain metastases from renal carcinoma: a retrospective study of 119 patients from the M. D. Anderson Cancer Center. Int J Radiat Oncol Biol Phys. 1997;37:753-759. 12. Onufrey V, Mohiuddin M. Radiation therapy in the treatment of metastatic renal cell carcinoma. Int J Radiat Oncol Biol Phys. 1985;11:2007-2009. 13. RTOG 0631 Protocol Information. Available at: http://www.rtog. org/ClinicalTrials/ProtocolTable/StudyDetails.aspx?study=0631. Accessed April 29, 2011. 14. Jensen MP, Turner JA, Romano JM, Fisher LD. Comparative reliability and validity of chronic pain intensity measures. Pain. 1999;83:157-162. 15. Halperin EC, Harisiadis L. The role of radiation therapy in the management of metastatic renal cell carcinoma. Cancer. 1983;51: 614-617. 16. Sundaresan N, Scher H, DiGiacinto GV, Yagoda A, Whitmore W, Choi IS. Surgical treatment of spinal cord compression in kidney cancer. J Clin Oncol. 1986;4:1851-1856. 17. Reddy S, Hendrickson FR, Hoeksema J, Gelber R. The role of radiation therapy in the palliation of metastatic genitourinary tract carcinomas. A study of the Radiation Therapy Oncology Group. Cancer. 1983;52:25-29. 18. Chao ST, Koyfman SA, Woody N, et al. Recursive partitioning analysis index is predictive for overall survival in patients undergoing spine stereotactic body radiation therapy for spinal metastases [e-pub ahead of print April 12, 2011]. Int J Radiat Oncol Biol Phys.
www.practicalradonc.org
Original Report
The efficacy of external beam radiotherapy and stereotactic body radiotherapy for painful spinal metastases from renal cell carcinoma Grant K. Hunter MDa , Ehsan H. Balagamwala BAa , Shlomo A. Koyfman MDa , Trevor Bledsoe BSa , Lawrence J. Sheplan MDa , Chandana A. Reddy MSa , Samuel T. Chao MDa , Toufik Djemil PhDa , Lilyana Angelov MDb , Gregory M.M. Videtic MDa,⁎ a
Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio Department of Neurosurgery, Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio
b
Received 18 November 2011; revised 9 January 2012; accepted 25 January 2012
Abstract Purpose: Palliative radiotherapy is routinely used to treat painful spinal metastases from renal cell carcinoma (RCC). Conventionally planned external beam radiotherapy (CRT) has been standard, with high-dose stereotactic body radiotherapy (SBRT) becoming increasingly common given the radioresistant nature of RCC. We compared the efficacy and durability of pain relief produced by these 2 modalities. Methods and Materials: Patients with painful spinal metastases from RCC treated from 2002-2010 were included. Response was defined similar to the Radiation Therapy Oncology Group 0631 protocol: complete response (CR) being resolution of pain without increased narcotics; partial response (PR) included patients with an incomplete reduction in pain without increased narcotics. Patients who experienced a CR or PR were coded as having pain relief, while those with persistent pain or additional narcotics requirements were coded as failures. Achievement of pain relief was analyzed using competing risk analysis with death as the competing event. Time to pain relief was plotted using cumulative incidence analysis. Results: A total of 110 patients (34 CRT; 76 SBRT) were included. Median follow-up was 4.3 months (range, 0.2-38). Median Karnofsky performance score was higher for patients treated with SBRT compared with CRT (80 vs 70; P = .0004). Overall pain response rates were 68% for CRT and 62% for SBRT, with respective CR and PR rates of 12% and 56% for CRT, and 33% and 29% for SBRT (P = .01). Median time to pain relief was 0.6 weeks for CRT versus 1.2 weeks for SBRT (P = .29). For patients who achieved pain relief (n = 79), median duration was 1.7 months for CRT versus 4.8 months for SBRT (P = .095). On univariate analysis no factors were significantly related to pain relief.
Conflicts of interest: None. ⁎ Corresponding author. Department of Radiation Oncology, 9500 Euclid Ave, Desk T-28, Cleveland, OH 44195. E-mail address: [email protected] (G.M.M. Videtic). 1879-8500/$ – see front matter © 2012 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.prro.2012.01.005
e96
G.K. Hunter et al
Practical Radiation Oncology: October-December 2012
Conclusions: CRT was not statistically different than SBRT for pain relief in symptomatic spine metastases from RCC and should be used as first line treatment. The appropriate use of SBRT in this population merits prospective study. © 2012 American Society for Radiation Oncology. Published by Elsevier Inc. All rights reserved.
Introduction Around 61,000 cases of kidney cancer are expected to be diagnosed in 2011 with renal cell carcinoma (RCC) constituting 92% of the total number of malignancies. 1 Patients with RCC frequently present with metastatic disease. 2 The second most common site of metastatic disease (following pulmonary metastases) is bone, occurring in 5% of all patients with RCC, and the spine is the most common site of bone involvement. 3-5 Patients who present with spinal metastatic disease commonly have severe back pain as a presenting symptom and are routinely treated for palliation. 6,7 The general treatment options employed to treat malignant spine metastases include the following: surgery, including kyphoplasty and vertebroplasty; radiation therapy, which includes conventionally planned and delivered external beam radiation or stereotactic body radiotherapy (SBRT); and medical management that includes narcotics, nonsteroidal antiinflammatory drugs, or both; or some combination of therapies. Radical surgery, either in the form of laminectomy or corpectomy, is often employed when patients present with malignant spinal cord compression or when the patient has a pathologic vertebral body fracture for stabilization purposes. Radiation therapy is used for patients postoperatively after the metastatic osseous lesion is resected to eliminate microscopic disease, or for those patients who are medically unfit for surgery or do not warrant surgical intervention as deemed by the surgical oncologist. 8 Radiation therapy for spinal metastases has traditionally been delivered with conventional radiation therapy (CRT), which entails single or multiple fractions of daily radiation delivered to the involved region of the spine with a margin of 1 vertebral body above and below the affected area. Two-dimensional field design with 1 or 2 beams is commonly used, and the dose is prescribed to a point. Recent technologic advances in both immobilization and planning techniques have created the ability to accurately and safely deliver high doses of radiation to the spine while respecting spinal cord tolerance via SBRT. 9,10 While RCC has been shown in preclinical models and in older clinical series to be relatively radioresistant, SBRT offers an attractive strategy to overcome this radioresistance with the use of high doses of single fraction radiotherapy. 11,12 Though this hypothesis is likely contributing to the increasing use of SBRT in this
particular patient population, there is a paucity of comparative efficacy data between CRT and the more costly and labor intensive SBRT in this setting for any histology of metastasis. This is currently being investigated in the on-going Radiation Therapy Oncology Group (RTOG) 0631 phase II/III randomized trial, with patients with RCC included in the phase III component. 13 The primary objective of this study is to retrospectively compare pain relief outcomes for patients with spinal metastases from RCC treated (in a nonrandomized fashion) with these 2 modalities.
Methods and materials Patients with radiographically evident, painful spine metastases from RCC (C1-sacrum), who were treated with either CRT or SBRT between 2002 and 2010, were included in this Institutional Review Board approved retrospective, nonrandomized review. Of the SBRT patients, only those treated with single fraction SBRT were included. There were no dose or fractionation criteria for CRT patients. Patients with prior surgery or external beam radiotherapy to the treated area were also included. CRT was delivered using a single posterior-anterior beam or anterior-posterior/posterior anterior technique prescribed to a depth, or the isocenter, respectively. Simulations were done with either fluoroscopic or computed tomographic (CT)-based imaging. For CRT planning, the target was the involved vertebral body (bodies) with an additional margin of 1 vertebral body superior and inferior to the lesion(s). The choice of dose and fractionation was at the discretion of the treating radiation oncologist of which there were 12 over the study time period. Patients treated with SBRT were immobilized using an Elekta BodyFix stereotactic body frame system (Elekta AB, Stockholm, Sweden) and underwent CT simulation in the supine position. The CT simulation entailed obtaining axial CT images with 1.5-mm slice thickness. A high definition magnetic resonance imaging (MRI) scan (1.5-mm slices) of the spine at the involved vertebral level(s) was also obtained. Axial MRI images were then fused to the planning CT to aid with spinal cord and tumor delineation. Treatment planning was performed using iPlan (BrainLab, Munich, Germany) and image guidance was accomplished using the ExacTrac system (Novalis, Palo Alto, CA) during treatment delivery. The clinical target volume (CTV) was defined as the entire vertebral body or entire posterior vertebral elements
Practical Radiation Oncology: October-December 2012
depending on tumor involvement. Paraspinal and epidural disease were routinely included in the CTV, with 62% of the patients in the SBRT cohort having some epidural involvement. In the SBRT cohort included in our review, 17 patients or 22% had compression fractures evident. No patients were treated postoperatively (including postkyphoplasty) in the SBRT cohort. Our institutional practice is to not use SBRT for patients with frank spinal cord compression. Noncontinuous vertebral bodies were allowed to be treated, with 9 patients (15% of SBRT patients) having noncontiguous bodies treated on the same day and 8 patients (14% of SBRT patients) having noncontiguous bodies treated on different days. No additional margins were added to the CTV to account for setup error. The spinal cord was delineated on axial high definition MRI images and was contoured at the level of the CTV with 3-mm cranial and caudal extensions. The following dose constraints were used as our institution standards: 14 Gy maximum point dose and V10 Gy b10% for the spinal cord, and 16 Gy maximum point dose and V12 Gy b10% for the cauda equina. Figure 1A and B depicts 2 representative treatment plans; one for CRT (Fig 1A) and one for SBRT (Fig 1B) with typical isodose coverage included. We characterized pain relief as it was defined by RTOG 0631, which includes both an assessment of pain response and an assessment of narcotics usage. The RTOG protocol specifies its definitions of complete response (CR) and partial response (PR) with the incorporation of the Numerical Rating Pain Scale (NRPS) in the definition of each response. 14 CR was graded if there was a pain score of 0 at the index site or if there was complete resolution of pain without an increase in narcotics usage. PR included patients who had a reduction in pain without an increase in narcotics usage. The PR definition in the RTOG protocol states that the pain score must be reduced by 3 points or more without increased narcotics. This rigorous definition of pain response (utilizing NRPS scores and narcotics usage) is clearly able to be done when data are collected in a prospective fashion, but is difficult if not impossible when attempting to use the same scale in a retrospective manner when the NRPS score was not assessed at the time of follow-up. For our study we used the above scale to assess the pain response of each patient, incorporating the NRPS as the RTOG does if such information was available for the patient. We adapted the RTOG scale (based on the NRPS) for those of our patients who did not have numerical scores coded. All patients did have qualitative assessments of their pain and narcotic usage documented at baseline and follow-up that was used for pain relief assessment in the absence of NRPS quantitative assessments. All patients were asked at follow-up a qualitative assessment of their pain (ie, mild, moderate, or severe) and if it had improved compared with before
External beam/SBRT for RCC of the spine
e97
Figure 1 Pictures of a typical plan utilizing a (A) single posterior-anterior beam for treatment of T11-L1 to 3000 cGy/10 fractions above a comparative stereotactic body radiotherapy (SBRT) plan (B) for treatment of L2 to 1600 cGy/1 fraction with representative isodose lines (100%, 90%, and 70% for the conventional treatment and 100%, 90% and 65% for SBRT).
treatment. Using this qualitative scoring along with the patient's narcotics usage data we were able to assign a corresponding pain score for comparison. Patients who underwent SBRT all had NRPS numbers obtained at the time of follow-up to help score their pain response. Patients who experienced a CR or PR were coded as having pain relief while patients with persistent pain or who required an increase in narcotics were coded as failures of treatment. Achievement of pain relief was analyzed using competing risk analysis with death as the competing
e98
G.K. Hunter et al
Table 1
Practical Radiation Oncology: October-December 2012
Patient and treatment characteristics
Variables
CRT
SBRT
P value
Median age, y (range)
62 (43-81) 1.75 (0.2-28) 12:22 70 (20-100) 8 (23%) 6 (18%)
57 (41-80) 4.95 (0.3-38) 13:63 80 (50-90) 0 12 (16%)
.16
N/A .58
8 (24%) 20 (8-30)
45 (59%) 15 (8-16)
.98 b.0001
5 (1-10)
1
b.0001
1 (0-6)
6 (1-19)
.0001
Median follow-up, mo (range) Male:Female Median KPS (range) Palliative surgery Prior radiation treated to spine Multiple sites treated Median total dose, Gy (range) Median no. of fractions (range) Time from sim to tx, d (range)
.07 .04 .0004
CRT, conventional radiation therapy; KPS, Karnofsky performance score; N/A, not applicable; SBRT, stereotactic body radiotherapy; sim, simulation; tx, treatment.
event. Time to pain relief was plotted using cumulative incidence analysis. Fine and Gray regression was used to determine if any of the following variables were significantly associated with pain relief: Karnofsky performance score, age, sex, prior palliative spinal surgery, prior CRT to the same site, number of fractions (for CRT), and total dose. A P value of b .5 was used to determine if a variable was significantly associated with pain relief.
Results A total of 110 patients (34 CRT; 76 SBRT) were included in this study. Patient and treatment characteristics are detailed in Table 1. The overall median follow-up
Figure 2 Incidence of pain relief after conventional radiation therapy (CRT) or stereotactic body radiotherapy (SBRT) for painful spinal renal cell carcinoma metastases.
time was 4.3 months (range, 0.2-38). Patients treated with SBRT had longer median follow-up than those treated with CRT (4.95 vs 1.75 months; P = .07). Median Karnofsky performance score was significantly higher for the SBRT group (80 vs 70; P = .0004). The number of fractions used to deliver CRT varied between 1 and 10, with a median of 5 fractions. The total dose delivered by CRT varied from 8 Gy to 30 Gy with a median of 20 Gy. The total dose was significantly higher in patients treated with CRT compared with SBRT (20 Gy vs 15 Gy; P b .0001) and the total number of fractions was significantly greater in patients treated with CRT (5 vs 1; P b .0001). Treatment doses and fractionations used are detailed in Table 2. While a small number of both groups of patients had received prior radiation therapy to the treated area, only patients in the CRT group had prior decompressive surgery (23%). The time from simulation to treatment was significantly longer for patients treated with SBRT (median of 6 days vs 1 day, P b .0001).
Table 2 Dose and fractionation schemes used for palliative treatments CRT
SBRT
Dose/fractions
No. pts (%)
Dose/fractions
No. pts (%)
20 Gy/5 fxs 30 Gy/10 fxs 8 Gy/1 fx 15 Gy/5 fxs 27.3 Gy/7 fxs 21 Gy/7 fxs 16 Gy/4 fxs 12 Gy3 fxs
13 12 3 2 1 1 1 1
16 Gy/1 fx 15 Gy/1 fx 14 Gy/1 fx 13 Gy/1 fx 12 Gy/1 fx 10 Gy/1 fx 8 Gy/1 fx
32 (42) 12 (17) 14 (18) 4 (5) 9 (12) 4 (5) 1 (1)
(38) (35) (9) (6) (3) (3) (3) (3)
CRT, conventional radiation therapy; fxs, fraction(s); pts, patients; SBRT, stereotactic body radiotherapy.
Figure 3 Duration of pain relief for patients with pain relief after radiation therapy (RT). CRT, conventional radiation therapy; SBRT, stereotactic body radiotherapy; SD, standard deviation.
Practical Radiation Oncology: October-December 2012 Table 3 Fine and Gray competing risk regression for pain relief: univariate analysis Variable
P value
HR
95% CI
Univariate analysis CRT vs SBRT Age (cont) KPS (cont) Dose (cont) Gender (female vs male) Prior EBRT # Fx (CRT only) Palliative surgery (CRT only)
.33 .38 .4 .11 .73 .22 .051 .46
1.28 0.99 0.99 1.04 0.92 1.37 1.14 0.70
0.78-2.08 0.96-1.02 0.97-1.01 0.99-1.09 0.57-1.48 0.83-2.29 1-1.31 0.27-1.82
CI, confidence interval; CRT, conventional radiation therapy; EBRT, external beam radiotherapy; # Fx, number of fractions; HR, hazard ratio; KPS, Karnofsky performance score; SBRT, stereotactic body radiotherapy.
The overall pain response rates were similar for the CRT and SBRT groups (68% vs 62%, respectively), with CR rates being lower for the CRT group (12% vs 33%) and PR rates being higher for the CRT group (56% vs 29%, P = .01). The median time to achieve pain relief was shorter for the CRT group (0.6 vs 1.2 weeks; P = .29). The cumulative incidence for pain control by treatment is depicted in Fig 2, which shows a nonsignificant difference between the 2 groups. For those patients who achieved pain relief (n = 79), the median duration of pain relief was nonsignificantly shorter for the CRT group (1.7 vs 4.8 months; P = .095) (Fig 3). For the few patients whose pain recurred after an initial pain response, the time to recurrence was similar for the CRT and SBRT groups (4.7 vs 5.0 months; P = .58). On competing risk regression univariate analysis there were no factors that were significantly associated with pain relief (Table 3).
Discussion To our knowledge, this study is the first to compare pain relief outcomes for patients with spinal metastases from RCC who were treated with CRT or SBRT. Overall, our data suggest that these 2 modalities provide equivalent rates of pain relief, which argues against the belief that conventional radiotherapy is ineffective for this histology. These data are in line with previously published studies that report pain relief rates of 77%-83% in patients with RCC metastases. 15,16 Another interesting finding from this dataset was that CRT produced a pain response at least as quickly as SBRT. Time to pain relief is an extremely significant endpoint in this patient population with incurable disease, especially when pain is the most common reason they temporarily discontinue their systemic treatments in favor of focal palliative radiotherapy. Our CRT data are consistent with older series that report that by 2 weeks
External beam/SBRT for RCC of the spine
e99
post-conventional radiotherapy, nearly 60% of patients achieved some pain relief. 17 On the other hand, while SBRT may be slower to produce a symptomatic response, the pain relief appears more durable than CRT. Although in our series the duration of pain relief was not statistically significantly different comparing CRT to SBRT, the SBRT patients had a higher rate of complete response. While the 12% CR rate in the CRT arm is identical to the CR rate at 8 weeks post treatment on the RTOG 97-14 phase III randomized study, SBRT led to a nearly threefold increase in complete pain relief. Both of these observations can be explained by the hypothesis that high-dose single fraction SBRT delivers a higher radiobiologic effective dose, thereby enhancing tumor cell kill and leading to enhanced responses that last longer. Taken together, these findings highlight an important area of uncertainty in the management of these patients. SBRT is an application of advanced technology that may have some advantages over CRT; however, the right setting in which SBRT is appropriate remains an active area of investigation as it is more costly and labor intensive treatment modality compared with CRT. While some argue that its radioresistant nature warrants SBRT to be used as first line in spinal metastases from RCC, our data suggest that a more nuanced treatment algorithm should be employed. CRT appears to be an appropriate first line therapy for many patients with RCC spinal metastases, while SBRT may be reserved for those with a history of prior CRT, or perhaps those with an extended life expectancy where durability of response is a critical measure of successful treatment. To this point, Chao et al 18 recently published a recursive partitioning analysis for patients who underwent spine SBRT, including patients with RCC, which identified 3 distinct subgroups of patients with significantly disparate overall survival. This provides a helpful first step in tailoring the use of SBRT to patients who may benefit from it most. There are several limitations to this study. Its retrospective nature, coupled with the varying quality of follow-up information, especially in the older CRT cohort, introduces recall and follow-up biases that may impact some of the study results, especially vis-à-vis durability of response data. That 23% of patients in the CRT group had prior surgery can confound the baseline pain and treatment response data in those patients as well. Another limitation to our study is the retrospective pain assessment. As previously described, in order to best compare the 2 modalities the assignments of pain response were either using the NRPS scores that were collected at the time of follow-up or based on the patient's qualitative description of pain relief with an assessment of narcotics usage. The conclusions drawn would be stronger if all patients had NRPS scores taken at the time of follow-up to more rigorously assess pain response. Finally, the small numbers of patients in this study prohibit any firm
e100
G.K. Hunter et al
conclusions to be drawn. Despite these shortcomings, the study achieves its goal of providing the first comparative analysis of these 2 modalities in this patient population and provides data that at least reaffirm the knowledge gap. This study also argues that the question of CRT versus SBRT is as relevant for patients with RCC as it is for other histologies currently included in the ongoing RTOG 0631 study. We would therefore advocate that these patients be included in that study, and future investigations should continue to focus on identifying which patients with RCC are most likely to benefit from SBRT.
Conclusions Both CRT and SBRT provide effective palliation of symptomatic spine metastases from RCC. CRT as a treatment modality requires less intensive treatment planning and quality assurance, and is less costly and was not statistically different compared with SBRT. The use of SBRT in this population merits prospective study. Conventional treatment remains appropriate as the first line treatment for palliation with SBRT being appropriate for selected patients.
References 1. American Cancer Society. Cancer Facts & Figures 2011. Available at: http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/ documents/document/acspc-029771.pdf. Accessed February 7, 2012. 2. Levy DA, Slaton JW, Swanson DA, Dinney CP. Stage specific guidelines for surveillance after radical nephrectomy for local renal cell carcinoma. J Urol. 1998;159:1163-1167. 3. Kume H, Kakutani S, Yamada Y, et al. Prognostic factors for renal cell carcinoma with bone metastasis: who are the long-term survivors? J Urol. 2011;185:1611-1614. 4. Swanson DA, Orovan WL, Johnson DE, Giacco G. Osseous metastases secondary to renal cell carcinoma. Urology. 1981;18: 556-561.
Practical Radiation Oncology: October-December 2012 5. Shvarts O, Lam JS, Kim HL, Han KR, Figlin R, Belldegrun A. Eastern Cooperative Oncology Group performance status predicts bone metastasis in patients presenting with renal cell carcinoma: implication for preoperative bone scans. J Urol. 2004;172:867-870. 6. Hartsell WF, Scott CB, Bruner DW, et al. Randomized trial of shortversus long-course radiotherapy for palliation of painful bone metastases. J Natl Cancer Inst. 2005;97:798-804. 7. Langdon J, Way A, Heaton S, Bernard J, Molloy S. The management of spinal metastases from renal cell carcinoma. Ann R Coll Surg Engl. 2009;91:649-652. 8. Faul CM, Flickinger JC. The use of radiation in the management of spinal metastases. J Neurooncol. 1995;23:149-161. 9. Hamilton AJ, Lulu BA, Fosmire H, Stea B, Cassady JR. Preliminary clinical experience with linear accelerator-based spinal stereotactic radiosurgery. Neurosurgery. 1995;36:311-319. 10. Ryu SI, Chang SD, Kim DH, et al. Image-guided hypo-fractionated stereotactic radiosurgery to spinal lesions. Neurosurgery. 2001;49: 838-846. 11. Wronski M, Maor MH, Davis BJ, Sawaya R, Levin VA. External radiation of brain metastases from renal carcinoma: a retrospective study of 119 patients from the M. D. Anderson Cancer Center. Int J Radiat Oncol Biol Phys. 1997;37:753-759. 12. Onufrey V, Mohiuddin M. Radiation therapy in the treatment of metastatic renal cell carcinoma. Int J Radiat Oncol Biol Phys. 1985;11:2007-2009. 13. RTOG 0631 Protocol Information. Available at: http://www.rtog. org/ClinicalTrials/ProtocolTable/StudyDetails.aspx?study=0631. Accessed April 29, 2011. 14. Jensen MP, Turner JA, Romano JM, Fisher LD. Comparative reliability and validity of chronic pain intensity measures. Pain. 1999;83:157-162. 15. Halperin EC, Harisiadis L. The role of radiation therapy in the management of metastatic renal cell carcinoma. Cancer. 1983;51: 614-617. 16. Sundaresan N, Scher H, DiGiacinto GV, Yagoda A, Whitmore W, Choi IS. Surgical treatment of spinal cord compression in kidney cancer. J Clin Oncol. 1986;4:1851-1856. 17. Reddy S, Hendrickson FR, Hoeksema J, Gelber R. The role of radiation therapy in the palliation of metastatic genitourinary tract carcinomas. A study of the Radiation Therapy Oncology Group. Cancer. 1983;52:25-29. 18. Chao ST, Koyfman SA, Woody N, et al. Recursive partitioning analysis index is predictive for overall survival in patients undergoing spine stereotactic body radiation therapy for spinal metastases [e-pub ahead of print April 12, 2011]. Int J Radiat Oncol Biol Phys.
