Abstract
BACKGROUND
This study was designed to evaluate the impact of tyrosine kinase inhibitors (TKIs) on incidence of brain metastasis (brain metastasis) and overall survival (OS) in patients with metastatic renal cell cancer (mRCC).
METHODS
All patients who presented with mRCC but no brain metastasis in the intervals 2002 to 2003 and 2006 to 2007 were identified using the institutional tumor registry. The following data were collected: age, sex, Fuhrman grade, disease sites, nephrectomy, systemic therapy including TKIs (sorafenib or sunitinib), Memorial Sloan-Kettering Cancer Center risk category, brain metastasis treatment, and vital status. Statistical analysis was performed using the Cox proportional hazards model and the Kaplan-Meier method.
RESULTS
Of the 338 patients who were identified; 154 (46%) were treated with a TKI before brain metastasis, and 184 (54%) were not. There were no significant differences in age, histology, nephrectomy, involved sites of disease other than lung, or Memorial Sloan-Kettering Cancer Center risk category between the groups. Median OS was longer in the TKI-treated group (25 months vs 12.1 months, P < .0001). In multivariate analysis, TKI treatment (hazard ratio [HR], 0.53; 95% confidence interval [CI], 0.38–0.74; P < .001) was associated with improved OS. Forty-four (13%) patients developed a brain metastasis, including 29 (15.8%) of the non-TKI group and 15 (9.7%) of the TKI group. The 5-year actuarial rate of brain metastasis was 40% versus 17%, respectively (P < .001). TKI treatment was associated with lower incidence of brain metastasis in Cox multivariate analysis (HR, 0.39; 95% CI, 0.21–0.73; P = .003). Lung metastasis increased the risk of brain metastasis (HR, 9.61; 95% CI, 2.97–31.1; P < .001).
CONCLUSIONS
Treatment with TKI agents reduces the incidence of brain metastasis in mRCC. Lung metastasis is a risk factor for brain metastasis development.
Keywords: renal cell carcinoma, tyrosine kinase inhibitors, sunitinib, sorafenib, brain metastasis
Cancers of the kidney and renal pelvis account for approximately 3.9% of all cancers in the United States, with an estimated 57,760 new cases and 12,980 deaths in 2009.1 Most of these cancers arise in the kidney, and the most common histological type, renal cell carcinoma (RCC), presents with metastatic disease in up to 30% of cases.2 The most common sites of metastasis include lung, lymph nodes, and bone,2 with the brain involved in 10% to 20% of cases.3,4 Because metastatic RCC (mRCC) is generally resistant to chemotherapy, immunologic therapy with interferon or interleukin-2 (IL-2) has been the most commonly used treatment, despite substantial toxicity5 and disappointingly low response rates (5%–20%).2
The advent of tyrosine kinase inhibitors (TKIs) has drastically altered the management of mRCC. These agents act by inhibiting intracellular kinases in the vascular endothelial growth factor pathway, the up-regulation of which is central to the pathogenesis of RCC.6–8 During the past 5 years, 3 TKIs, sunitinib, sorafenib, and pazopanib, were approved by the US Food and Drug Administration for use in advanced RCC. Sunitinib was approved in 2006 for RCC after phase 2 and 3 trials9 demonstrated that sunitinib, given as first-line therapy, doubled the progression-free survival (PFS) time compared with that obtained with interferon. Sorafenib was approved for RCC in 2005 after phase 3 trials similarly showed that it improved PFS compared with placebo for patients with cytokine-refractory disease,10 and pazopanib was approved in 2009 after another phase 3 trial showed that it led to improved PFS compared with placebo for both previously untreated and cytokine-refractory disease.11
The outcome for patients with RCC who develop brain metastasis is typically poor, with median survival times of only 4 to 11 months after diagnosis3,12 even after surgical resection, whole-brain radiotherapy, or stereotactic radiosurgery. The influence of TKIs on the development or progression of brain metastasis has not yet been clearly established. Several case reports have been published describing brain metastasis responding to treatment with sunitinib or sorafenib,13,14 raising the possibility that use of TKIs may improve the overall outcome for patients with RCC and brain metastasis. In addition, a recent review of patients enrolled in the TARGET trial found a lower incidence of brain metastasis in the group that received sorafenib,15 suggesting that these agents may also help to prevent or delay the development of brain metastasis.
In this retrospective analysis, we sought to evaluate the effect of treatment with 1 of 2 TKIs, sunitinib and sorafenib, on brain metastasis development in patients with mRCC. Given the timing of the approval and increasing use of these agents in early 2006, we selected 2 groups of patients, 1 that presented to The University of Texas MD Anderson Cancer Center in 2002 to 2003 and the other in 2006 to 2007 with the expectation that these periods would segregate patients by TKI exposure status. We extracted a wide array of clinicopathologic characteristics and used univariate and multivariate analyses to evaluate potential treatment effects on the incidence of brain metastasis, time to the diagnosis of brain metastasis, and outcomes, including survival and local control, after systemic therapy or treatment of brain metastasis.
MATERIALS AND METHODS
Patient Selection
The institutional review board of The University of Texas MD Anderson Cancer Center approved this retrospective analysis. Patients were identified by searching an institutional tumor registry database for those with stage IV clear cell RCC who presented for evaluation at The University of Texas MD Anderson Cancer Center in 2002 to 2003 or in 2006 to 2007. We excluded patients with sarcomatoid, papillary, or chromophobe features on histological examination, as well as those who had brain metastasis at their initial staging evaluation.
Data Review
The following data were extracted from medical records for review: age; sex; ethnic group; Fuhrman nuclear grade; involved sites of metastatic disease at diagnosis (including liver, adrenal glands, lung, bone, retroperitoneal nodes, and mediastinal nodes); whether the patient had undergone nephrectomy; Memorial Sloan Kettering Cancer Center risk category,16 a widely accepted risk stratification method calculated using serum lactate dehydrogenase, hemoglobin, corrected serum calcium, performance status, and time from diagnosis to systemic therapy; treatment with the TKIs sunitinib or sorafenib or other systemic agents; development of brain metastasis, treatment modality for brain metastasis, and location and treatment of brain recurrence; vital status; and duration of follow-up. To evaluate the potential effect of TKIs on brain metastasis development, patients were included in the TKI group only if they had received the agent before a brain metastasis was diagnosed. Survival calculations included all patients treated with TKIs regardless of when they received the agent relative to the development of brain metastasis.
Statistical Analysis
The frequencies of all clinical and pathological features were compared by using the chi-square statistic. Continuous variables were compared using Student t test. Overall survival, brain metastasis-free survival, and central nervous system (CNS) recurrence-free survival were estimated by using the Kaplan-Meier method. Patients who had not developed brain metastasis at last follow-up were censored at that time. The Cox proportional hazards method17 was used to evaluate the effect of patient, tumor, and other predictor factors of significance on survival and development of brain metastasis for different groups. The estimated hazard is reported, and the Wald test was used to assess the role of covariates in the model. All statistical analysis was performed using Stata/MP 11.1 for Windows.18
RESULTS
Patient Population
We identified 338 patients with stage IV clear cell mRCC but without brain metastasis at diagnosis (Table 1). The median age at diagnosis for all patients was 59 years (range, 18–83 years); 138 (41%) patients had bone metastasis, and 220 (65%) had lung metastasis at diagnosis. The median follow-up time for all patients was 16.2 months (range, 0–107 months).
Table 1.
Patient Characteristics
| Characteristic | All Patients, n=338 | No TKI Therapy, n=184 | TKI Therapy, n=154 | P |
|---|---|---|---|---|
| Age, median (range), y | 59 (18–83) | 58 (34–82) | 60 (18–83) | .45 |
| Sex | ||||
| Male | 227 (67%) | 123 (67%) | 104 (67%) | .89 |
| Female | 111 (33%) | 61 (33%) | 54 (33%) | |
| Fuhrman nuclear grade | .21 | |||
| 2 | 37 (11%) | 22 (12%) | 15 (10%) | |
| 3 | 112 (33%) | 54 (29%) | 58 (38%) | |
| 4 | 75 (22%) | 47 (26%) | 28 (18%) | |
| Not available | 115 (34%) | 61 (35%) | 54 (35%) | |
| Sites of distant metastases at diagnosis | ||||
| Liver | 62 (18%) | 36 (20%) | 26 (17%) | .53 |
| Adrenal glands | 62 (18%) | 35 (19%) | 27 (18%) | .53 |
| Retroperitoneal nodes | 113 (33%) | 66 (36%) | 47 (31%) | .30 |
| Mediastinal nodes | 97 (29%) | 53 (29%) | 44 (29%) | .96 |
| Bone | 138 (41%) | 73 (40%) | 65 (42%) | .64 |
| Lung | 220 (65%) | 129 (70%) | 91 (59%) | .034 |
| Nephrectomy | .45 | |||
| Yes | 193 (57%) | 102 (55%) | 91 (59%) | |
| No | 145 (43%) | 82 (45%) | 63 (41%) | |
| Systemic agents | ||||
| Capecitabine and fluorouracil | 121 (36%) | 84 (46%) | 37 (24%) | <.001 |
| Gemcitabine | 119 (35%) | 82 (45%) | 37 (24%) | <.001 |
| Interferon | 83 (25%) | 67 (36%) | 16 (10%) | <.001 |
| Interleukin-2 | 38 (11%) | 30 (16%) | 8 (5%) | .018 |
| Thalidomide | 54 (16%) | 46 (25%) | 8 (5%) | <.001 |
| Bevacizumab | 65 (19%) | 17 (9%) | 48 (31%) | <.001 |
| Erlotinib | 17 (5%) | 5 (2.7%) | 12 (8%) | .034 |
| Temsirolimus | 34 (10%) | 2 (1%) | 32 (21%) | <.001 |
| MSKCC risk score | ||||
| 0 | 23 (7%) | 13 (8%) | 10 (6%) | .99 |
| 1–2 | 190 (56%) | 103 (56%) | 87 (56%) | |
| 3–5 | 112 (33%) | 61 (33%) | 51 (33%) | |
| Not available | 13 (4%) | 7 (4%) | 6 (4%) | |
Abbreviations: MSKCC, Memorial Sloan-Kettering Cancer Center; TKI, tyrosine kinase inhibitor.
Of the 338 patients, 154 (46%) had received a TKI before the development of a brain metastasis, and 184 (54%) had not (Table 1). Six patients were first treated with a TKI subsequent to the diagnosis of brain metastasis, and were considered in the non-TKI group for the purpose of evaluating brain metastasis development. No significant differences were found between the TKI and non-TKI groups with respect to age, sex, ethnic group, Fuhrman nuclear grade, having undergone nephrectomy, or Memorial Sloan-Kettering Cancer Center risk score. The non-TKI group had a higher incidence of lung metastasis at diagnosis than the TKI group (70% vs 59%, P = .034); no other differences were found in sites of metastatic disease between the groups. As expected for patients treated over 2 different periods, the systemic therapies used were different for the TKI and non-TKI groups; specifically, a higher percentage of patients in the non-TKI group received capecitabine, gemcitabine, interferon, IL-2, or thalidomide, whereas more patients in the TKI group had received bevacizumab, erlotinib, or temsirolimus.
Overall Survival
The median overall survival time for all 338 patients was 19.1 months. Median survival time was longer in the TKI group than in the non-TKI group (25 months vs 12.1 months, P < .0001) (Table 2). When stratified according to the Memorial Sloan-Kettering Cancer Center risk score, patients with low-risk disease (score of 0) had a median survival time of 79.6 months, those with intermediate-risk disease (score of 1–2) 23.5 months, and those with high-risk disease (score of 3–5) 11.6 months (log-rank P < .0001) (Fig. 1).
Table 2.
Survival Outcomes
| Cohort | Median Survival Time, mo | ||
|---|---|---|---|
| TKI Group | Non-TKI Group | P (Log-Rank) | |
| All patients | 25 | 12.1 | <.0001 |
| Without brain metastasis | 26.2 | 11.7 | <.0001 |
| After diagnosis of brain metastasis | 6.15 | 5.23 | .33 |
| MSKCC low-risk disease (0) | Not reached | 27.1 | .027 |
| MSKCC intermediate-risk disease (1–2) | 32.6 | 17.2 | <.0001 |
| MSKCC high-risk disease (3–5) | 15.9 | 7.20 | .004 |
Abbreviations: MSKCC, Memorial Sloan-Kettering Cancer Center; TKI, tyrosine kinase inhibitor.
Figure 1.
Overall survival is shown according to (A) receipt of a tyrosine kinase inhibitor (TKI) or (B) Memorial Sloan-Kettering Cancer Center (MSKCC) risk classification.
Overall survival of all patients by TKI status. A, The difference in survival with and without TKI. B, The difference between the patient who never received TKI in comparison to those who received TKI before (naive) or after (progression) brain metastasis. TKI indicates tyrosine kinase inhibitor.
In Cox univariate analysis, receipt of TKI was associated with longer survival (hazard ratio [HR], 0.49; 95% confidence interval [CI], 0.38–0.64; P < .001). No other systemic agent was associated with an increase in survival. Other predictors of improved survival in the univariate analysis were receipt of nephrectomy (HR, 0.28; 95% CI, 0.21–0.37; P < .001) and belonging to the 2006 to 2007 cohort (HR, 0.71; 95% CI, 0.55–0.92; P = .009). Predictors of poor outcome included intermediate-risk or high risk Memorial Sloan-Kettering Cancer Center score; performance status (>2 on the Eastern Cooperative Oncology Group [ECOG] scale); presence of lung, liver, retroperitoneal, or mediastinal lymph node metastasis at diagnosis; and high Fuhrman nuclear grade (Table 3). Most of these variables also retained their significance in multivariate analysis; receipt of TKIs was a significant independent predictor of better survival (HR, 0.53; 95% CI, 0.38–0.74; P < .001), and high Fuhrman grade, the presence of lung or mediastinal nodal metastasis at diagnosis, and ECOG performance status >2 were significant predictors of worse survival (Table 3).
Table 3.
Univariate and Multivariate Predictors of Survival
| Characteristic | HR | 95% CI | P |
|---|---|---|---|
| Univariate | |||
| Fuhrman grade 4 | 1.96 | 1.40–2.74 | <.001 |
| Lung metastasis at diagnosis | 1.79 | 1.36–2.37 | <.001 |
| Liver metastasis at diagnosis | 2.57 | 1.88–3.53 | <.001 |
| Retroperitoneal node involvement at diagnosis | 1.67 | 1.28–2.18 | <.001 |
| Mediastinal node involvement at diagnosis | 1.48 | 1.12–1.94 | .006 |
| Nephrectomy | 0.28 | 0.21–0.37 | <.001 |
| Receipt of TKI | 0.49 | 0.38–0.64 | <.001 |
| MSKCC intermediate-risk disease (1–2) | 2.13 | 1.15–3.94 | .02 |
| MSKCC high-risk disease (31) | 4.03 | 2.14–7.54 | <.001 |
| Multivariate | |||
| Fuhrman grade 4 | 1.78 | 1.27–2.49 | .001 |
| Lung metastasis at diagnosis | 1.91 | 1.30–2.80 | .001 |
| Mediastinal node involvement at diagnosis | 1.60 | 1.12–2.29 | .01 |
| Receipt of TKI | 0.53 | 0.38–0.74 | <.001 |
| ECOG performance score >2 | 1.87 | 1.28–2.71 | .001 |
Abbreviations: CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; MSKCC, Memorial Sloan-Kettering Cancer Center; TKI, tyrosine kinase inhibitor.
Screening for Brain Metastasis
A total of 297 (88%) patients were documented to be free of brain metastases on at least 1 computed tomography or magnetic resonance imaging (MRI) scan of the brain, and 41 (12%) were not. Our institutional practice is to screen asymptomatic patients for brain metastasis with annual MRI. The median time between brain imaging, estimated by dividing the time from diagnosis to last brain imaging by the number of scans in the interim, was 3.6 months. The median time from last brain imaging to the last clinical follow-up date was 2.4 months.
Brain Metastasis to Last Clinical Follow-Up
Of the 338 patients, 44 (13%) eventually developed brain metastasis, including 29 (15.8%) patients in the non-TKI group and 15 (9.7%) in the TKI group (P = .10). Of the 44 patients, 21 (48%) had neurologic symptoms at the time of brain metastasis diagnosis. The remaining 23 (52%) were asymptomatic, with brain metastasis discovered on routine screening. The median time from diagnosis of mRCC to brain metastasis, in the 44 patients who developed brain metastasis, was 10 months. The median time to brain metastasis was 7.5 months in the non-TKI group and 12.2 months in the TKI group (P = .30).
Kaplan-Meier analysis showed that the rates of brain metastasis at 1, 2, and 5 years were 11%, 18%, and 26%, respectively. The actuarial rates of brain metastasis in the non-TKI group were 16%, 25%, and 40% at 1, 2, and 5 years, respectively. The rates of brain metastasis in the TKI group at the same time points were significantly lower at 5%, 11%, and 17%, respectively (log-rank P = .001) (Fig. 2).
Figure 2.
Brain metastasis rates are shown according to (A) receipt of a tyrosine kinase inhibitor (TKI) or (B) receipt of a TKI and the presence or absence of lung metastasis at diagnosis.
Brain metastasis local control by the different local treatment modalities. SRS indicates stereotactic radiosurgery; WBRT, whole brain radiotherapy.
In the Cox univariate analysis, receipt of TKIs was protective with regard to the development of brain metastasis (HR, 0.34; 95% CI, 0.18–0.63; P = .001). None of the other systemic agents we assessed was associated with a decrease in the incidence of brain metastasis. The presence of lung metastasis at diagnosis was a risk factor for the development of brain metastasis (HR, 10.58; 95% CI, 3.27–34.21; P < .001) (Table 4). These variables also retained significance in Cox multivariate analysis, with receipt of TKI associated with a lower incidence of brain metastasis (HR, 0.39; 95% CI, 0.21–0.73; P = .003) and having lung metastasis at diagnosis associated with a higher incidence of brain metastasis (HR, 9.61; 95% CI, 2.97–31.14; P < .001). (Table 4).
Table 4.
Univariate and Multivariate Predictors of Brain Metastasis
| Characteristic | HR | 95% CI | P |
|---|---|---|---|
| Univariate | |||
| Receipt of bevacizumab or Temsirolimus | 0.54 | 0.27–1.1 | .09 |
| Receipt of erlotinib | 0.26 | 0.04–1.88 | .18 |
| Receipt of TKI before brain metastasis | 0.34 | 0.18–0.63 | .001 |
| Lung metastasis at diagnosis | 10.58 | 3.27–34.21 | <.001 |
| Multivariate | |||
| Lung metastasis at diagnosis | 9.61 | 2.97–31.14 | <.001 |
| Receipt of TKI before brain metastasis | 0.39 | 0.21–0.73 | .003 |
Abbreviations: CI, confidence interval; HR, hazard ratio; TKI, tyrosine kinase inhibitor.
Brain Metastasis to Last Brain Imaging
We also performed Kaplan-Meier analysis using the date of last brain imaging, instead of the last clinical follow-up, as the endpoint. In this analysis, the actuarial rate of brain metastasis in the non-TKI group at 1 and 2 years was 28% and 37%, respectively. The rate of brain metastasis in the TKI group was again significantly lower, at 15% and 15%, respectively (log-rank P = .0017).
In Cox univariate analysis using last brain imaging as the endpoint, receipt of TKI was again found to be associated with a significant reduction in brain metastasis (HR, 0.31; 95% CI, 0.17–0.59; P < .001). Treatment with bevacizumab or temsirolimus was also associated with a reduced rate of brain metastasis (HR, 0.45; 95% CI, 0.22–0.91; P = .03). Mediastinal nodal metastasis (HR, 1.93; 95% CI, 1.01–3.70; P = .05) and lung metastasis (HR, 9.46; 95% CI, 2.92–30.6; P = <.001) were associated with increased incidence of brain metastasis. In Cox multivariate analysis, TKI treatment was associated with lower incidence of brain metastasis (HR, 0.39; 95% CI, 0.21–0.74; P = .004), and lung metastasis was associated with higher risk of brain metastasis (HR, 8.11; 95% CI, 2.49–26.4; P = .001).
Landmark Analysis
Landmark analysis was performed beginning 3 and 6 months from diagnosis of RCC. In landmark analysis beginning 3 months after diagnosis, the 1-year and 2-year rates of brain metastasis in the non-TKI group were 17% and 24.5%, respectively. These rates were significantly lower than the rates at the same time points in the TKI group (9% and 11.4%, respectively, P = .002). In landmark analysis using the 6-month cutoff time, the 1-year and 2-year rates of brain metastasis in the non-TKI group were 5% and 14%, respectively, versus 4% and 6% in the TKI group (P = .14).
Outcomes After Brain Metastasis
Brain metastases were treated with a variety of methods, alone or in combination, including observation, stereotactic radiosurgery, surgical resection, or whole brain radiotherapy. The median overall survival time after the diagnosis of brain metastasis was 6.15 months and was no different according to receipt of TKI (5.23 months for the TKI group and 6.15 months for the non-TKI group, logrank P = .33). Rates of local recurrence and development of new CNS metastases were also no different for the TKI group and the non-TKI group (log-rank P = .32). Outcomes are presented in Table 2.
DISCUSSION
In our analysis, we examined the development of brain metastasis in a large, diverse cohort of patients with advanced RCC and found that treatment with TKIs was associated with both improved survival and a reduced incidence of brain metastasis in patients with stage IV RCC. Our results suggest that the use of TKIs is altering the natural history of RCC by preventing the development of brain metastasis.
Although the impact of TKIs on mRCC is of great interest given the lack of effective treatments for this disease, exactly how they influence the development or progression of brain metastasis is not yet known. In their review of patients treated in the TARGET trial of sorafenib, Massard et al15 provided the first evidence suggesting that TKIs may protect the brain from mRCC when they reported that those who received sorafenib had a lower incidence of brain metastasis versus placebo (12% vs 3%). However, this study was limited by the relatively modest number of patients (n = 139), the inclusion of only patients for whom first-line therapy had failed, and the lack of multivariate analysis. In this context, our review of 338 patients extends the results reported by Massard et al by including analysis of a wide range of clinicopathologic criteria in a diverse group of patients, including those who had not undergone any treatment; in addition, the large number of patients in our study allowed us to perform univariate and multivariate analysis to evaluate the contributions of specific variables.
Our Kaplan-Meier analysis showed the rates of brain metastasis in the TKI group to be considerably lower than those in the group that did not receive TKIs. These 2 groups were composed of patients with similar clinical features, including Memorial Sloan-Kettering Cancer Center score, with the major difference being the type of treatment received. As we expected, patients who were not treated with TKIs tended to be treated with older agents such as IL-2, capecitabine, gemcitabine, and thalidomide, whereas more of the patients in the TKI group were treated with newer agents: erlotinib, temsirolimus, and bevacizumab. In multivariate analysis, receipt of TKI was associated with significant protection from brain metastasis (HR, 0.39; P = .003), and no other agent was associated with a reduction in brain metastasis in univariate or multivariate analysis. We believe that the lower incidence of brain metastasis observed can be attributed to receipt of TKIs, further supporting their use for patients with mRCC.
Although temsirolimus, erlotinib, and bevacizumab did not account for the observed protection from brain metastasis in our study, it is possible that these agents have a protective role with regard to brain metastasis that was unable to be detected because of the small numbers of patients in our sample who received those agents. Further study is needed to properly evaluate their influence on brain metastasis in mRCC. Another interesting finding was that in our study, lung metastasis was associated with a strong risk of developing a brain metastasis, with an HR >9 in both univariate and multivariate analysis. Of the 44 patients who developed brain metastasis, 90% had had lung metastasis at diagnosis, and only 2.5% of patients without lung metastasis at diagnosis ultimately developed a brain metastasis. A possible explanation is that the lung disease is a nidus for metastasis to the brain, in which case we would expect to observe lung metastasis to occur before brain metastasis. A second explanation is that there are underlying characteristics of these tumor cells that predispose them to metastasize to both lung and brain. In either case, this finding suggests that the presence of lung metastasis at diagnosis could be used as a tool to identify patients with RCC at high risk for brain metastasis and increase the surveillance of these patients accordingly. We further found that the incidence of brain metastasis reached a plateau at 4 years after the initial diagnosis, raising the possibility that long-term survivors of RCC are at lower risk for brain metastasis and perhaps might require less frequent screening. Finally, the finding that approximately half of the patients with brain metastasis in our cohort had asymptomatic brain metastasis discovered on routine screening has implications for the appropriate screening frequency for these patients. Our institutional practice is to screen stable, asymptomatic patients with mRCC for brain metastasis with annual MRI. Although there is no consensus screening guideline for brain metastasis, our data suggest that routine screening is perhaps indicated to detect a potentially high rate of asymptomatic brain metastasis.
The appearance of brain metastasis from RCC usually portends poor survival despite several options for local treatment. We found no significant difference in survival between patients with brain metastasis who received TKI drugs versus those who did not. Although it is difficult to draw conclusions from a small sample size of 44 patients, there are several possible explanations for this observation. First, we defined the TKI group as those who received TKIs before brain metastasis, meaning that in all cases brain metastasis represented progression after exposure to the agent. Those patients may also have had progression at other disease sites. Furthermore, progression on TKIs could indicate the emergence of a resistant clone, limiting the further utility of TKI treatment and thus removing 1 treatment option. All of these factors could negate the survival advantage of TKI drugs in this cohort.
In addition, we found no significant difference between the groups in the rate of subsequent brain metastasis after first brain metastasis. With the same caveat of a small sample size, and the additional consideration that many patients with brain metastasis have poor follow-up because of the nature of the disease, we offer a possible explanation for the reduced incidence of brain metastasis: that TKI agents stabilize systemic sites of disease, and thus reduce the number of circulating tumor cells with metastatic potential. This theory is somewhat supported by the finding that lung metastasis was a strong risk factor for future brain metastasis development. In this case, brain metastasis development could signal the emergence of a TKI-resistant clone, which the agent is no longer able to contain. Further study is certainly needed to better understand the role of TKIs in patients with brain metastasis.
It remains unclear whether TKI agents in combination with local treatments such as surgery or radiotherapy could improve local control or overall survival for patients with RCC and brain metastasis. Some case reports have described responses after treatment with sunitinib or sorafenib alone, suggesting that some benefit may be present. One recently published study reported that TKIs had limited benefit for patients with RCC and brain metastasis, 19 but that study included only 12 patients with brain metastasis, and neither PFS nor overall survival was statistically affected. To our knowledge, no study has evaluated outcomes of patients with RCC and brain metastasis according to TKI treatment status.
This study had some limitations, first among them being its retrospective nature. Retrospective reviews are inherently subject to bias, particularly selection bias. Although we compared groups in terms of a variety of clinicopathologic criteria, patients were not randomly assigned to treatment groups. Clinical factors at the time of treatment could lead to patients with a better prognosis being preferentially assigned to 1 treatment or another, confounding our results. Some patients in the earlier cohort (2002–2003) had survived long enough to receive a TKI; these survivors may have had underlying favorable factors aside from receipt of a TKI that led to improved outcomes. Finally, this retrospective review was limited to the data available, which did not always include brain imaging at preferred intervals.
In conclusion, in a diverse cohort of patients with mRCC without brain metastasis, TKIs both provide a significant survival benefit and may reduce the incidence of brain metastasis, further supporting their use in these patients. The presence of lung metastasis may have value in identifying patients at higher risk for brain metastasis and adjusting surveillance accordingly. Further investigation is needed to characterize the role of TKI treatment in the management and outcomes of patients after development of brain metastasis, and this remains a topic of great interest in our group.
Acknowledgments
FUNDING SOURCES
No specific funding was disclosed.
Footnotes
Presented at the Genitourinary Cancer Symposium of the American Society of Clinical Oncology, Orlando, Florida, February 17–19, 2011.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
References
- 1.Jemal A, Siegel R, Ward E, Yongping H, Xu J, Thun M. Global cancer statistics, 2009. CA Cancer J Clin. 2009;59:225–249. doi: 10.3322/caac.20006. [DOI] [PubMed] [Google Scholar]
- 2.Motzer RJ, Bander NH, Nanus DM. Renal-cell carcinoma. N Engl J Med. 1996;335:865–875. doi: 10.1056/NEJM199609193351207. [DOI] [PubMed] [Google Scholar]
- 3.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. doi: 10.1016/s0360-3016(97)00006-0. [DOI] [PubMed] [Google Scholar]
- 4.Saitoh H. Distant metastasis of renal adenocarcinoma. Cancer. 1981;48:1487–1491. doi: 10.1002/1097-0142(19810915)48:6<1487::aid-cncr2820480635>3.0.co;2-9. [DOI] [PubMed] [Google Scholar]
- 5.Fyfe G, Fisher RI, Rosenberg SA, et al. Results of treatment of 255 patients with metastatic renal cell carcinoma who received high-dose recombinant interleukin-2 therapy. J Clin Oncol. 1995;13:688–696. doi: 10.1200/JCO.1995.13.3.688. [DOI] [PubMed] [Google Scholar]
- 6.Kim WY, Kaelin WG. Role of VHL gene mutation in human cancer. J Clin Oncol. 2004;22:4991–5004. doi: 10.1200/JCO.2004.05.061. [DOI] [PubMed] [Google Scholar]
- 7.Rini BI, Small EJ. Biology and clinical development of vascular endothelial growth factor-targeted therapy in renal cell carcinoma. J Clin Oncol. 2005;23:1028–1043. doi: 10.1200/JCO.2005.01.186. [DOI] [PubMed] [Google Scholar]
- 8.Dvorak HF. Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy. J Clin Oncol. 2002;20:4368–4380. doi: 10.1200/JCO.2002.10.088. [DOI] [PubMed] [Google Scholar]
- 9.Motzer RJ, Hutson TE, Tomczak P, et al. Sunitinib versus interferon alfa in metastatic renal cell carcinoma. N Engl J Med. 2007;356:115–124. doi: 10.1056/NEJMoa065044. [DOI] [PubMed] [Google Scholar]
- 10.Escudier B, Eisen T, Stadler WM, et al. for the TARGET Study Group. Sorafenib in advanced clear-cell renal-cell carcinoma. N Engl J Med. 2007;356:125–134. doi: 10.1056/NEJMoa060655. [DOI] [PubMed] [Google Scholar]
- 11.Sternberg CN, Davis ID, Mardiak J, et al. Pazopanib in locally advanced or metastatic renal cell carcinoma: results of a randomized phase III trial. J Clin Oncol. 2010;28:1061–1068. doi: 10.1200/JCO.2009.23.9764. [DOI] [PubMed] [Google Scholar]
- 12.Shuch B, La Rochelle JC, Klatte T, et al. Brain metastasis from renal cell carcinoma: presentation, recurrence, and survival. Cancer. 2008;113:1641–1648. doi: 10.1002/cncr.23769. [DOI] [PubMed] [Google Scholar]
- 13.Walid MS, Johnston KW. Successful treatment of a brain metastasized renal cell carcinoma. Ger Med Sci. 2009;7:Doc28. doi: 10.3205/000087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Helgason HH, Mallo HA, Droogendijk H, et al. Brain metastases in patients with renal cell cancer receiving new targeted treatment. J Clin Oncol. 2008;26:152–154. doi: 10.1200/JCO.2007.13.5814. [DOI] [PubMed] [Google Scholar]
- 15.Massard C, Zonierek J, Gross-Goupil M, Fizazi K, Szczylik C, Escudier B. Incidence of brain metastasis in renal cell carcinoma treated with sorafenib. Ann Oncol. 2010;21:1027–1031. doi: 10.1093/annonc/mdp411. [DOI] [PubMed] [Google Scholar]
- 16.Motzer RJ, Bacik J, Murphy BA, Russo P, Mazumdar M. Interferon-alfa as a comparative treatment for clinical trials of new therapies against advanced renal cell carcinoma. J Clin Oncol. 2002;20:289–296. doi: 10.1200/JCO.2002.20.1.289. [DOI] [PubMed] [Google Scholar]
- 17.Cox DR. Regression models and life tables. J Royal Stat Soc Ser B. 1972;34:187–220. [Google Scholar]
- 18.StataCorp. Statistical software. College Station, TX: StataCorp; 2009. Stata: Release 11. [Google Scholar]
- 19.Vogl UM, Bojic M, Lamm W, et al. Extracerebral metastases determine the outcome of patients with brain metastases from renal cell carcinoma. BMC Cancer. 2010;10:480. doi: 10.1186/1471-2407-10-480. [DOI] [PMC free article] [PubMed] [Google Scholar]