Brain Metastasis From Renal Cell Carcinoma. BACKGROUND. Patients with renal cell carcinoma brain metastases (RCCBM) are

Transcription

1 1641 Brain Metastasis From Renal Cell Carcinoma Presentation, Recurrence, and Survival Brian Shuch, MD 1 Jeff C. La Rochelle, MD 1 Tobias Klatte, MD 1 Stephen B. Riggs, MD 1 Weiqing Liu, MS 2 Fairooz F. Kabbinavar, MD 1,3 Allan J. Pantuck, MD, MSN 1 Arie S. Belldegrun, MD 1 1 Department of Urology, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, California. 2 Department of Biostatistics, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, California. 3 Department of Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, California. BACKGROUND. Patients with renal cell carcinoma brain metastases (RCCBM) are frequently excluded from trials and to the authors knowledge no guidelines currently exist regarding central nervous system (CNS) surveillance or treatment. The objective of the current study was to assist in the creation of treatment guidelines. METHODS. Patients undergoing evaluation for RCCBM from 1989 to 2006 were identified. Their characteristics, symptoms, pathologic variables, number and size of RCCBM, CNS treatment, CNS recurrence, overall survival, and use of systemic therapy were reviewed. RESULTS. A total of 138 patients were identified with RCCBM, of whom 92% had clear cell RCC and 95% had synchronous extracranial metastases. CNS symptoms were noted in 67% of patients. Symptomatic CNS tumors were larger (2.1 cm vs 1.3 cm; P <.001) and more frequently required a craniotomy (P <.001). The median overall survival after a diagnosis of RCCBM was 10.7 months; the 1-year, 2- year, and 5-year survival rates were 48%, 30%, and 12%, respectively. Median CNS recurrence was 9 months after RCCBM treatment. The initial number of tumors (>1 tumor) was found to be an independent predictor of CNS recurrence (hazards ratio of 3.72; P <.001). Those patients with 1 and >1 lesion had a median CNS recurrence-free survival of 13 months and 4 months, respectively (P <.001). Patients receiving interleukin-2 after CNS treatment had a response rate of 17%. CONCLUSIONS. Patients with metastatic RCC should undergo CNS screening to allow the identification of smaller lesions that are more amenable to treatment. Those patients with solitary RCCBM are less likely to develop CNS recurrence after local therapy. Selected patients with good performance status may exhibit prolonged survival and should be offered aggressive therapy. Cancer 2008;113: Ó 2008 American Cancer Society. Address for reprints: Allan J. Pantuck, MD, MSN, Department of Urology, David Geffen School of Medicine, University of California at Los Angeles, Le Conte Ave., B7-298A CHS, Box , Los Angeles, CA ; Fax: (310) ; apantuck@mednet. ucla.edu Received February 6, 2008; revision received April 7, 2008; accepted April 21, KEYWORDS: brain metastasis, radiosurgery, renal cell cancer, kidney cancer. Brain metastases (BM) are commonly observed in patients with metastatic renal cell carcinoma (RCC), with a reported incidence of 2% to 17%. 1-4 RCC brain metastases (RCCBM) are associated with large amounts of peritumoral edema and, because of their vascularity, are susceptible to spontaneous intracranial hemorrhage. 5,6 As such, localizing and nonlocalizing central nervous system (CNS) symptoms are frequently observed. Symptoms such as headaches, confusion, altered behavior, and seizures have been reported in 80% to 98% of patients. 7-9 CNS screening of asymptomatic patients has led to the detection of small, asymptomatic lesions in other malignancies. 10 However, these asymptomatic CNS metastases do not appear to represent indolent disease and, left untreated, all become symptomatic with time. 11 Although RCCBM are typically highly symptomatic, small lesions may be found by screening before becoming symptomatic. ª 2008 American Cancer Society DOI /cncr Published online 31 July 2008 in Wiley InterScience (

2 1642 CANCER October 1, 2008 / Volume 113 / Number 7 To our knowledge, there currently are no clear guidelines for CNS imaging in patients with metastatic RCC. Several centers screen before initiating immunotherapy because of concerns for limited efficacy and the risk of CNS toxicity, whereas others perform imaging only when CNS symptoms are present. For patients with localized disease, CNS surveillance after nephrectomy is generally not recommended unless symptoms develop. Up to 90% of patients who develop RCCBM will have concurrent metastases at other sites and therefore most surveillance recommendations advise against routine CNS screening in asymptomatic patients without evidence of disease in the more common sites for recurrence. 3,12,13 Immunotherapy has played a limited role in the management of RCCBM because the brain is considered an immune-privileged site. In addition, interleukin-2 (IL-2) is believed to decrease the seizure threshold and is contraindicated in patients with active CNS disease. 14 As part of the treatment of RCCBM, patients may require systemic steroids that can decrease the efficacy of immune-based treatments. Chemotherapy also has limited CNS efficacy because of the blood brain barrier. Large molecules cannot freely pass into normal brain parenchyma decreasing therapeutic efficacy. Previous phase 3 trials with the recently approved agents sorafenib, sunitinib, and temsirolimus all excluded patients with CNS disease, and therefore their efficacy in the CNS is uncertain at this time To assist with the formulation of management guidelines, the outcome of patients with RCCBM was reviewed at a single RCC referral center. We hypothesized that a subset of patients could experience extended survival. Recent National Comprehensive Cancer Network (NCCN) guidelines suggest stereotactic radiosurgery (STRS) could be used for the management of small (<2 cm), asymptomatic CNS metastases. 18 We hypothesized that CNS screening could identify smaller, asymptomatic tumors, allowing for a less invasive treatment strategy. In addition, we sought to identify those patients less likely to experience CNS recurrence after treatment of their original RCCBM, a finding that may influence the choice of subsequent systemic therapy. MATERIALS AND METHODS Patients at the University of California at Los Angeles (UCLA) who were undergoing evaluation for RCC from 1989 through 2006 were reviewed. Those with radiographic or pathologic documentation of RCCBM were identified. Through a retrospective review of an Institutional Review Board (IRB)-approved database, demographic, clinical, and pathologic data were gathered. At our institution, patients presenting with metastatic disease or demonstrating disease recurrence after nephrectomy underwent CNS screening before consideration for systemic therapy. The timing of BM occurrence, concurrent sites of metastases, CNS symptoms, size and number of lesions, type of CNS treatment (if performed), and the use and response to systemic therapy were recorded. The Eastern Cooperative Group (ECOG) performance status was evaluated at the time of BM diagnosis. BM size was recorded for the largest individual lesion based on imaging. CNS tumor characteristics were compared using a Student t test and graphically represented with standard boxplots. The study endpoints were overall survival calculated from the date of RCCBM diagnosis to date of death or last follow-up, and CNS recurrence after the date of treatment of the original RCCBM. Survival was calculated using the Kaplan-Meier method and groups were compared with the log-rank test. For timing of BM occurrence, patients were placed into 3 groups: those presenting with RCCBM within 1 month of RCC diagnosis, those with a disease-free interval after nephrectomy for localized disease who experienced a disease recurrence that included RCCBM, and those with non-cns metastatic disease that later progressed to brain involvement. Because these groups differ with regard to the timing of disease and systemic therapy usage, a detailed survival analysis was performed only for the largest group, those with non-cns metastatic disease that progressed to brain involvement. A multivariate Cox regression analysis was applied to identify independent prognostic factors of overall survival in this group of patients. Variables incorporated into the model were ECOG performance status, Fuhrman grade, metastatic disease at the time of presentation, systemic therapy before and after RCCBM, and months between RCC diagnosis and BM occurrence. Follow-up CNS imaging after intervention was reviewed to evaluate the development of new CNS disease for all patients. CNS recurrence was defined as the presence of new CNS lesions observed during follow-up. Local recurrence was defined as regrowth of a treated lesion and was considered primary treatment failure. These are considered rare and were excluded from our definition of CNS recurrence. CNS recurrence-free survival was calculated using the Kaplan-Meier method and groups were compared with the log-rank test. Multivariate Cox regression analysis was applied to identify independent prognostic factors of CNS recurrence after CNS treatment. Variables incorporated into the model were ECOG

3 Brain Metastases From RCC/Shuch et al 1643 TABLE 1 Baseline Patient Characteristics Category No. % Man Mean age at diagnosis of RCC, y 57 Mean age at diagnosis of RCCBM, y 60 Nephrectomy Yes T classification N classification N N Grade Other metastasis No Yes Histology Clear cell Papillary Chromophobe Unclassified ECOG PS Timing of disease Presentation Recurrence Progression Systemic therapy after treatment Yes RCC indicates renal cell carcinoma; RCCBM, renal cell carcinoma brain metastases; ECOG PS, Eastern Cooperative Oncology Group performance status. performance status, CNS symptoms, number of lesions, size of BM, interval between RCC diagnosis and RCCBM, and interval between metastatic disease and RCCBM. Because to our knowledge no other US Food and Drug Administration (FDA)-approved therapy was available for the majority of the study period, patients with good ECOG performance status were generally offered various intravenous, high-dose IL-2 regimens after treatment of the CNS lesion. Response to IL-2-based systemic therapy was prospectively evaluated based on the Response Evaluation Criteria in Solid Tumors (RECIST) after the first cycle of therapy. A cutoff of P.05 was used to determine statistical significance. SAS statistical software (Cary, NC) was used for all statistical analyses. RESULTS Of the 1855 patients who were evaluated for RCC during the study period, a total of 138 patients were identified with BM. Patient characteristics are displayed in Table 1. Patients most frequently had CNS TABLE 2 CNS Tumor and Treatment Characteristics Category No. % CNS symptoms Yes No No. of CNS tumors > Size of tumors, cm < > CNS therapy None Craniotomy STRS WBRT Craniotomy plus STRS Other CNSindicates central nervous system; STRS, stereotactic radiosurgery; WBRT, whole-brain radiotherapy. involvement after progression of prior non-cns metastatic disease, as demonstrated in 79 patients (57%) with a median and mean occurrence of CNS involvement at 10.6 months and 16.2 months, respectively (standard deviation of 16.9 months). CNS involvement at the time of diagnosis of RCC also was observed in 37 patients (27%). Overall, isolated CNS involvement was rare and was found in 7 (5.1%) of patients. Clear cell histology was the most frequent histologic subtype in 102 patients (92.7%). Patients with RCCBM frequently had good ECOG performance status (0 of 1) at the time of BM diagnosis (87 cases [79.1%]). A solitary site of BM was observed in 92 patients (68.1%) (Table 2). The number of BM varied from 1 to 12 (mean, 1.73; median, 1). Lesion size varied from 0.2 cm to 4.0 cm (mean, 1.83 cm; median, 1.7 cm). A total of 42 patients (32.6%) with BM were asymptomatic at the time of presentation. A total of 132 patients (95.4%) received treatment for their CNS disease that included STRS (35.1%), craniotomy (16%), whole-brain radiotherapy (12.2%), craniotomy and STRS (20.6%), or other combination therapies (11.5%). Lesion size, but not the number of BM, was associated with both the presence of CNS symptoms and the need for craniotomy. The mean size of asymptomatic lesions was 1.3 cm versus 2.1 cm for symptomatic lesions (P <.001) (Fig. 1a). Craniotomy was performed more frequently for larger size tumors (mean size of 2.2 cm versus 1.6 cm for those not undergoing craniotomy) (P <.001) (Fig. 1b). Overall, 115 patients (83.3%) died during followup. The median follow-up of surviving patients was 8.7 months. The median overall survival after the diagnosis of RCCBM was 10.7 months (Fig. 2). For all

4 1644 CANCER October 1, 2008 / Volume 113 / Number 7 FIGURE 2. Overall survival for all 138 patients from the time of diagnosis of brain metastasis. FIGURE 1. (a) Symptoms and association with the number and size of the lesions, shown as the mean (standard deviation [SD]). NS indicates not significant. (b) Craniotomy and association with the number and size of the lesions. (c) Tumor size by symptoms. CNS indicates central nervous system. (d) Tumor size by use of craniotomy. patients, the 1-year, 2-year, and 5-year overall survival rates were 48%, 30%, and 12%, respectively. Patients with prior non-cns metastatic disease progressing to brain involvement demonstrated a median survival of 10.4 months from RCCBM diagnosis with a 1-year, 2-year, and 5-year overall survival rates of 49%, 29%, and 11%, respectively (Fig. 3a). Analysis of the use of systemic therapy before CNS disease, systemic metastases at the time of presentation, Fuhrman grade, and time from RCC diagnosis to BM diagnosis were not found to be associated with overall survival. Only ECOG performance status (P <.001) and use of systemic therapy after RCCBM diagnosis (P 5.026) were associated with improved survival (Figs. 3b and 3c). In our multivariate model, only ECOG performance status appeared to independently impact overall survival, with a hazards ratio (HR) of 6.14 (95% confidence interval [95% CI], ; P <.001). For all patients with follow-up imaging, CNS recurrence was present in 50 of 92 (54.3%) patients, with a median CNS recurrence of 9 months (Fig. 4a). Univariate analysis of risk factors for CNS recurrence after treatment demonstrated no association with systemic therapy after treatment, CNS symptoms, ECOG performance status, and both time from RCC diagnosis to metastatic disease and metastatic disease to BM. Those with 1 and >1 brain metastases had a median CNS recurrence-free survival of 13 months and 4 months, respectively (P <.001) (Fig. 4b). In the multivariate model, only the number of CNS lesions (>1) was found to be independently associated with CNS recurrence (HR of 3.7; 95% CI, [P <.001]). No significant differences were found with regard to the CNS treatment and the incidence of recurrence. There was no difference noted with regard to the use of whole-brain radiotherapy

5 Brain Metastases From RCC/Shuch et al 1645 FIGURE 4. (a) Central nervous system (CNS) recurrence-free survival overall. (b) CNS recurrence-free survival based on the number of initial tumors. between those with and without CNS recurrence (20% vs 19%; P 5.88). Response to IL-2-based systemic therapy was prospectively assessed in those receiving therapy after treatment of CNS disease. A total of 18 patients were treated with IL-2, and 3 (16.6%) acheived an objective response to therapy. No complete responders to therapy were observed. FIGURE 3. Survival for 79 patients with disease progression to the brain. (a) Overall survival. (b) Overall survival by Eastern Cooperative Oncology Group performance status (ECOG PS). (c) Overall survival by postcentral nervous system (CNS) treatment systemic therapy. DISCUSSION This retrospective series is unique in that to our knowledge, it is the largest series reported to date of patients with metastatic RCC to the brain and includes the modern era of CNS treatment when STRS became available. Important observations

6 1646 CANCER October 1, 2008 / Volume 113 / Number 7 regarding the patient characteristics of those with RCCBM were made. Clear cell carcinoma appeared to be the predominant histology metastasizing to the brain, found to be present in >90% of patients. Isolated metastatic disease to the brain was a rare finding, confirming that imaging for CNS disease should not be part of routine surveillance after curative nephrectomy unless neurologic symptoms are present. Asymptomatic disease was present in nearly one-third of patients, contradicting the belief that RCC brain metastases are uniformly symptomatic. 7-9 Lesion size and not total number of lesions appeared to be a major determinant of symptoms and influenced the choice of management. These small incidentally detected tumors will likely progress to symptomatic, hemorrhagic lesions with time. Because these smaller, asymptomatic lesions appear to require craniotomy less frequently, perhaps earlier detection by screening can increase the use of less invasive treatments such as STRS. This outpatient treatment modality has a high local control rate of >90% and has minimal morbidity. 1 Based on the findings of the current study, we believe that all patients with metastatic RCC should receive initial CNS screening. In addition, future studies should assess whether patients with metastatic RCC demonstrating prolonged survival benefit from yearly CNS surveillance. This strategy will potentially allow early intervention before catastrophic complications arise. The survival analysis demonstrated that a proportion of patients with RCCBM display extended survival after the diagnosis of CNS involvement. The 1-year, 2-year, and 5-year survival rates of 48%, 30%, and 12%, respectively, was surprising and greater than previously reported. 7,9 This patient population has generally been considered to have poor survival and is frequently excluded from aggressive therapy. With these findings, select patients with good performance status should be considered for aggressive treatment including cytoreductive nephrectomy, CNS treatment, and systemic therapy. Active CNS disease is common, despite CNS treatment, usually resulting from the development of new CNS lesions rather than local recurrence. 8,9 Analysis of risk factors for CNS recurrence demonstrated that the initial number of CNS lesions (>1) was an independent predictor of CNS recurrence. Those patients presenting with a solitary BM exhibited CNS failure at a median of 13 months, whereas those with >1 BM had a median recurrence of 4 months. Those patients with a solitary BM appear to have a prolonged CNS disease-free interval. Patients with multiple CNS lesions appear to have CNS-homing tumors that preferentially metastasize to the brain. Chemokines have recently been implicated in determining organ-specific metastasis in RCC and may be responsible for the preferential spread of these tumors. 19 Our finding has implications for the selection of appropriate systemic therapy in addition to CNS surveillance after treatment. To our knowledge IL-2 is the only approved agent known to produce long-term cure. However, in this patient population, many centers avoid the use of IL-2 despite successful CNS treatment. IL-2 use in this patient population poses theoretic risks including peritumoral cerebral edema from capillary leak syndrome, cerebral hemorrhage from treatmentinduced thrombocytopenia, altered mental status, seizures, and treatment-related confusion. A retrospective review of 64 patients with CNS disease (61 with melanoma and 3 with RCC) who received IL-2 was performed by the National Cancer Institute. 20 Patients with treated and untreated BM did not demonstrate any significant neurotoxicity compared with those without CNS disease. However, patients with CNS disease were withdrawn from therapy more frequently for disorientation. Guirgus et al 20 analyzed the response to therapy for patients with treated BM that appeared to be similar to those patients without CNS disease (18.5% vs 19.7%). In patients without CNS treatment, the response to therapy was found to be significantly worse, with a response rate of 5.6%. Similarly, our data in patients with RCC demonstrated that patients with treated BM had objective response rates (16.7%) that were similar to those reported in other published series. 21,22 Both these findings demonstrate that select patients with treated CNS disease can be safely and effectively treated with IL-2. On the basis of the current study data regarding CNS recurrence, patients with a solitary treated CNS lesion can be considered for therapy that may not have CNS efficacy, including IL-2. A risk of CNS recurrence exists, but patients can demonstrate a long CNS disease-free interval, allowing them to receive systemic therapy to treat their extracranial disease. Those patients with multiple lesions appear to have CNS-homing tumors that demonstrate rapid CNS progression. For these patients, an agent such as IL-2 may not be the best option because of concerns regarding CNS efficacy and potential toxicity. The development of targeted agents with small molecular size has brought the hope of CNS penetration and efficacy. Recently, a different targeted agent, gefitinib, demonstrated efficacy against brain metastases from nonsmall cell lung cancer. 23,24 In RCC, the new agents sunitinib and sorafenib have demonstrated early promise in treating CNS disease. A

7 Brain Metastases From RCC/Shuch et al 1647 recent case report indicated that sunitinib was safe and led to a partial response in a patients with RCCBM. 25 In addition, in a phase 3 trial of sorafenib, patients had a 75% reduction in the incidence of brain metastasis compared with placebo. 26 Currently, a phase 2 clinical trial is enrolling patients to specifically evaluate the efficacy of sunitinib in treating patients with asymptomatic RCCBM (NCT ). There are many limitations of this retrospective, observational study. This patient population may not be representative of all patients with brain metastases due the nature of our tertiary referral pattern and the finding that those with poor performance status and advanced disease may have been unable to present to our institution. Although CNS screening was performed initially for the majority of patients presenting with metastatic disease or with disease recurrence, those with disease progression who were referred to our center may not have undergone initial screening. In addition, those patients who developed progressive disease did not undergo a standard surveillance protocol for the detection of asymptomatic lesions. As such, the results of the current study may not be applicable for CNS surveillance recommendations. In addition, patients received different CNS treatment modalities and multiple factors other than tumor size and number went into those complex neurosurgical treatment decisions. Conclusions Patients with metastatic RCC should be evaluated with CNS screening, thereby allowing for the identification of smaller, asymptomatic lesions that may be amenable to less invasive treatment. Some patients exhibit prolonged survival and those with good performance status should be treated aggressively. Those patients with a solitary BM are less likely to develop CNS recurrence. Those with multiple CNS lesions appear to have CNS-homing tumors that demonstrate rapid CNS progression. REFERENCES 1. Sheehan JP, Sun MH, Kondziolka D, et al. Radiosurgery in patients with renal cell carcinoma metastasis to the brain: long-term outcomes and prognostic factors influencing survival and local tumor control. J Neurosurg. 2003;98: Sandock DS, Seftel AD, Resnick MI. A new protocol for the followup of renal cell carcinoma based on pathological stage. J Urol. 1995;154: Ljungberg B, Alamdari FI, Rasmuson T, et al. Follow-up guidelines for nonmetastatic renal cell carcinoma based on the occurrence of metastases after radical nephrectomy. BJU Int. 1999;84: Levy DA, Slaton JW, Swanson DA, et al. Stage specific guidelines for surveillance after radical nephrectomy for local renal cell carcinoma. J Urol. 1998;159: Bucci L, Giannini A, Bellotti R, et al. [Role of hypernephroma in a case of intracerebral hemorrhage as a 1st sign of metastasis. Case report.] Riv Neurol. 1986;56: Bitoh S, Hasegawa H, Ohtsuki H, et al. Cerebral neoplasms initially presenting with massive intracerebral hemorrhage. Surg Neurol. 1984;22: Mori Y, Kondziolka D, Flickinger JC, et al. Stereotactic radiosurgery for brain metastasis from renal cell carcinoma. Cancer. 1998;83: Muacevic A, Kreth FW, Mack A, et al. Stereotactic radiosurgery without radiation therapy providing high local tumor control of multiple brain metastases from renal cell carcinoma. Minim Invasive Neurosurg. 2004;47: Hoshi S, Jokura H, Nakamura H, et al. Gamma-knife radiosurgery for brain metastasis of renal cell carcinoma: results in 42 patients. Int J Urol. 2002;9: ; discussion 626; author reply Seaman EK, Ross S, Sawczuk IS. High incidence of asymptomatic brain lesions in metastatic renal cell carcinoma. J Neurooncol. 1995;23: Seute T, Leffers P, Wilmink JT, et al. Response of asymptomatic brain metastases from small-cell lung cancer to systemic first-line chemotherapy. J Clin Oncol. 2006;24: Wronski M, Arbit E, Russo P, et al. Surgical resection of brain metastases from renal cell carcinoma in 50 patients. Urology. 1996;47: Lam JS, Shvarts O, Leppert JT, et al. Postoperative surveillance protocol for patients with localized and locally advanced renal cell carcinoma based on a validated prognostic nomogram and risk group stratification system. J Urol. 2005;174: ; discussion 472; quiz Bukowski RM, Young J, Goodman G, et al. Polyethylene glycol conjugated interleukin-2: clinical and immunologic effects in patients with advanced renal cell carcinoma. Invest New Drugs. 1993;11: Hudes G, Carducci M, Tomczak P, et al. Temsirolimus, interferon alfa, or both for advanced renal-cell carcinoma. N Engl J Med. 2007;356: Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma. N Engl J Med. 2007;356: Escudier B, Eisen T, Stadler WM, et al. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med. 2007;356: National Comprehensive Cancer Network. Central nervous system cancers. In: NCCN Clinical Practice Guidelines in Oncology. Version 1. Fort Washington, PA: National Comprehensive Cancer Network; Pan J, Burdick MD, Belperio JA, et al. CXCR3/CXCR3 ligand biological axis impairs RENCA tumor growth by a mechanism of immunoangiostasis. J Immunol. 2006;176: Guirguis LM, Yang JC, White DE, et al. Safety and efficacy of high-dose interleukin-2 therapy in patients with brain metastases. J Immunother. 2002;25: Figlin R, Gitlitz B, Franklin J, et al. Interleukin-2-based immunotherapy for the treatment of metastatic renal cell carcinoma: an analysis of 203 consecutively treated patients. Cancer J Sci Am. 1997;3(suppl 1):S92-S97.

8 1648 CANCER October 1, 2008 / Volume 113 / Number Walther MM, Yang JC, Pass HI, et al. Cytoreductive surgery before high dose interleukin-2 based therapy in patients with metastatic renal cell carcinoma. J Urol. 1997;158: Chiu CH, Tsai CM, Chen YM, et al. Gefitinib is active in patients with brain metastases from non-small cell lung cancer and response is related to skin toxicity. Lung Cancer. 2005;47: Ceresoli GL, Cappuzzo F, Gregorc V, et al. Gefitinib in patients with brain metastases from non-small-cell lung cancer: a prospective trial. Ann Oncol. 2004;15: Thibault F, Billemont B, Rixe O. Regression of brain metastases of renal cell carcinoma with antiangiogenic therapy. J Neurooncol. 2008;86: Massard C, Zonierek J, Laplanche A, Schwartz B, Szczylik C, Escudier B. Incidence of brain metastasis in advanced renal cell carcinoma among patients randomized in a phase III trial of sorafenib, an oral multi-kinase inhibitor. Ann Oncol. 2006;17(S):1X148-1X149. Abstract 454P.