Cabozantinib versus Everolimus in Advanced Renal-Cell Carcinoma

Patients

Eligible patients were 18 years of age or older with advanced or metastatic renal-cell carcinoma with a clear-cell component and measurable disease. Patients must have received prior treatment with at least one VEGFR-targeting tyrosine kinase inhibitor and must have had radiographic progression during treatment or within 6 months after the most recent dose of the VEGFR inhibitor. Patients with known brain metastases that were adequately treated and stable were eligible. There was no limit to the number of previous anticancer therapies, which could include cytokines, chemotherapy, and monoclonal antibodies, including those targeting VEGF, the programmed death 1 (PD-1) receptor, or its ligand PD-L1. Eligible patients also had a Karnofsky performance-status score of at least 70% (on a scale from 0 to 100%, with higher scores indicating better performance status) and adequate organ and marrow function. Key exclusion criteria were previous therapy with an mTOR inhibitor or cabozantinib or a history of uncontrolled, clinically significant illness.

Study Design and Treatment

Patients were randomly assigned in a 1:1 ratio to receive either cabozantinib or everolimus. Randomization was stratified according to the number of previous VEGFR-targeting tyrosine kinase inhibitors (1 or ≥2) and prognostic risk category (favorable, intermediate, or poor) according to the Memorial Sloan Kettering Cancer Center (MSKCC) criteria²⁴ (for details on the MSKCC criteria, see Table S1 in the Supplementary Appendix, available with the full text of this article at NEJM.org).

Cabozantinib and everolimus were provided by the sponsor (Exelixis). Cabozantinib was administered orally at a dose of 60 mg once daily, and everolimus was administered orally at a dose of 10 mg once daily. Dose reductions for cabozantinib (40 mg, then 20 mg) and everolimus (5 mg, then 2.5 mg) and interruptions of study treatment were specified for management of adverse events. Treatment was continued as long as clinical benefit was observed by the investigator or until the development of unacceptable toxic effects. Crossover between treatment groups was not allowed.

End Points and Assessments

The primary end point was duration of progression-free survival, defined as the interval between the dates of randomization and first documentation of disease progression (assessed by an independent radiology review committee) or death from any cause. Secondary efficacy end points were duration of overall survival and objective response rate. Tumor response and progression were assessed according to Response Evaluation Criteria in Solid Tumors, version 1.1,²⁵ in all patients at screening, every 8 weeks after randomization during the first 12 months, and every 12 weeks thereafter. Routine safety evaluations were performed and adverse-event severity was assessed by the investigator with the use of the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0.²⁶

Study Oversight

The protocol was approved by the institutional review board or ethics committee at each center, and the study was conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki. Safety was monitored by an independent data monitoring committee. Data were collected by the sponsor and were analyzed in collaboration with the authors. The authors vouch for the accuracy and completeness of the data and for the fidelity of the study to the protocol. The first draft of the manuscript was written by the first and last authors, with all the authors contributing to subsequent drafts. Medical-writing support, funded by the sponsor, was provided by Bellbird Medical Communications. All the authors made the decision to submit the manuscript for publication. The study protocol and statistical analysis plan are available at NEJM.org.

Statistical Analysis

The trial was designed to provide adequate power for assessment of both the primary end point of progression-free survival and the secondary end point of overall survival. For the primary end point, we estimated that 259 events (disease progression or death) would be required to provide 90% power to detect a hazard ratio of 0.667 (7.5 months with cabozantinib vs. 5 months with everolimus), using the log-rank test and a two-sided significance level of 0.05. For the overall-survival end point, assuming a single interim analysis at the time of the primary end-point analysis and a subsequent final analysis, we estimated that 408 deaths would be required to provide 80% power to detect a hazard ratio of 0.75 (20 months with cabozantinib vs. 15 months with everolimus), using the log-rank test and a two-sided significance level of 0.04.

Figure 1. Figure 1. Study Design.

VEGFR denotes vascular endothelial growth factor receptor.

Efficacy was evaluated in two populations according to the intention-to-treat principle. To evaluate the secondary end point of overall survival, 650 patients were planned (the overall-survival population). However, only 375 patients were required to achieve appropriate statistical power for the primary end point of progression-free survival. Thus, the study was designed to evaluate the primary end point in the first 375 patients who underwent randomization (the progression-free–survival population) to allow longer follow-up of progression-free survival (Figure 1).

Hypothesis testing for progression-free and overall survival was performed with the use of the stratified log-rank test according to the stratification factors used at randomization. Median duration of progression-free survival and overall survival and associated 95% confidence intervals for each treatment group were estimated with the Kaplan–Meier method. Hazard ratios were estimated with a Cox regression model. A prespecified interim analysis for overall survival was conducted at the time of the primary end-point analysis. The type I error for the interim analysis was controlled by a Lan–DeMets alpha spending function, with O’Brien–Fleming boundaries, to account for the fraction of planned events at the time of the analysis.

Patients

From August 2013 through November 2014, a total of 658 patients from 173 centers in 26 countries were randomly assigned to receive cabozantinib (330 patients) or everolimus (328 patients); these patients together compose the overall-survival population (Fig. S1 in the Supplementary Appendix). The first 375 patients who underwent randomization (187 assigned to cabozantinib and 188 assigned to everolimus) compose the progression-free–survival population for the primary end-point analysis (Fig. S2 in the Supplementary Appendix). The safety population comprises all patients who received study treatment (331 received cabozantinib and 322 received everolimus) (Fig. S1 in the Supplementary Appendix).

As of the data-cutoff date of May 22, 2015, a total of 133 patients assigned to cabozantinib and 67 patients assigned to everolimus were continuing to receive study treatment. Minimum follow-up time was 11 months in the progression-free–survival population and 6 months in the overall-survival population. The most common reason for discontinuing treatment was progression of disease on radiography.

Table 1. Table 1. Baseline Demographic and Clinical Characteristics.

The treatment groups were balanced with respect to baseline demographic and disease characteristics (Table 1). The most common previous therapy was sunitinib, and the majority of patients had received only one prior VEGFR inhibitor.

Efficacy

Figure 2. Figure 2. Kaplan–Meier Estimates of Progression-free Survival.

Disease progression was assessed by an independent radiology review committee.

The duration of progression-free survival was determined by an independent radiology review committee in the first 375 patients who underwent randomization. The estimated median progression-free survival was 7.4 months (95% confidence interval [CI], 5.6 to 9.1) with cabozantinib and 3.8 months (95% CI, 3.7 to 5.4) with everolimus. The rate of disease progression or death was 42% lower with cabozantinib than with everolimus (hazard ratio for progression or death, 0.58; 95% CI, 0.45 to 0.75; P<0.001) (Figure 2). The results were similar in a supportive analysis involving investigator assessment of progression-free survival (median, 7.4 months [95% CI, 6.3 to 7.6] with cabozantinib vs. 5.3 months [95% CI, 3.8 to 5.6] with everolimus; hazard ratio for progression or death, 0.60; 95% CI, 0.47 to 0.76; P<0.001) (Fig. S3 in the Supplementary Appendix).

A progression-free survival benefit associated with cabozantinib was consistently observed in prespecified subgroups defined according to the number of prior VEGFR inhibitors and MSKCC prognostic risk category (Fig. S4 in the Supplementary Appendix). In a post hoc analysis of the subgroup of 153 patients who received sunitinib as their only prior VEGFR inhibitor, the estimated median progression-free survival was 9.1 months (95% CI, 5.6 to 11.2) with cabozantinib and 3.7 months (95% CI, 1.9 to 4.2) with everolimus (hazard ratio for progression or death, 0.41).

Among the first 375 patients who underwent randomization, the objective response rate, as assessed by an independent radiology review committee, was significantly higher with cabozantinib than with everolimus (partial responses in 40 of the 187 patients [21%] assigned to cabozantinib vs. 9 of the 188 patients [5%] assigned to everolimus; P<0.001) (Table S2 in the Supplementary Appendix). A best response of stable disease occurred in 116 patients (62%) in each group, and progressive disease occurred in 26 patients (14%) assigned to cabozantinib versus 51 patients (27%) assigned to everolimus. In the subgroup of 153 patients who received sunitinib as their only prior VEGFR inhibitor, objective responses occurred in 17 of the 76 patients assigned to cabozantinib (22%; 95% CI, 14 to 33) and in 2 of the 77 patients assigned to everolimus (3%; 95% CI, 0 to 9).

Figure 3. Figure 3. Kaplan–Meier Estimates of Overall Survival.

At the prespecified interim analysis of overall survival, 202 deaths had occurred in the overall-survival population. A trend toward longer overall survival with cabozantinib than with everolimus was observed (hazard ratio for death, 0.67; unadjusted 95% CI, 0.51 to 0.89; P=0.005) (Figure 3). The P value of ≤0.0019 required to achieve statistical significance at the time of the interim analysis was not reached, and survival follow-up is continuing to the planned final analysis after 408 deaths occur. The trend toward longer survival with cabozantinib occurred despite more frequent use of subsequent anticancer therapies in the everolimus group (155 of 328 patients [47%]) than in the cabozantinib group (126 of 330 patients [38%]) (Table S3 in the Supplementary Appendix). The most common subsequent anticancer therapies were axitinib in the everolimus group (74 patients [23%]) and everolimus in the cabozantinib group (75 patients [23%]).

Safety

The median duration of treatment was 7.6 months among patients who received cabozantinib and 4.4 months among patients who received everolimus. Dose reductions occurred among 197 of the 331 patients (60%) treated with cabozantinib and 79 of the 322 patients (25%) treated with everolimus (Table S4 in the Supplementary Appendix). The median average daily dose was 44 mg of cabozantinib and 9 mg of everolimus. The rate of treatment discontinuation due to adverse events not related to renal-cell carcinoma was 9% (31 patients) in the cabozantinib group and 10% (31 patients) in the everolimus group.

Table 2. Table 2. Adverse Events.

The incidence of adverse events (any grade), regardless of whether the event was considered by the investigator to be related to the study treatment, was 100% with cabozantinib and more than 99% with everolimus, and the incidence of adverse events of grade 3 or 4 was 68% with cabozantinib and 58% with everolimus (Table 2). The most common grade 3 or 4 adverse events with cabozantinib were hypertension (15%), diarrhea (11%), and fatigue (9%) and with everolimus were anemia (16%), fatigue (7%), and hyperglycemia (5%). The most common adverse events (any grade) leading to dose reductions with cabozantinib were diarrhea (16%), the palmar–plantar erythrodysesthesia syndrome (11%), and fatigue (10%), and the most common with everolimus were pneumonitis (4%), fatigue (3%), and stomatitis (3%). Grade 5 adverse events occurred in 22 patients (7%) in the cabozantinib group and in 25 patients (8%) in the everolimus group and were primarily related to disease progression. Grade 5 events that were considered to be treatment-related occurred in 1 patient in the cabozantinib group (death not otherwise specified) and in 2 patients in the everolimus group (aspergillus infection and aspiration pneumonia).

Progression-free survival was longer with cabozantinib than with everolimus in this randomized, phase 3 trial involving patients with renal-cell carcinoma that had previously progressed during at least one VEGFR-targeted therapy. The efficacy with cabozantinib was robust, with an estimated median progression-free survival of 7.4 months, as compared with 3.8 months with everolimus, and a hazard ratio of 0.58, corresponding to a 42% reduction in the rate of disease progression or death. Objective tumor responses were observed in 21% of the patients assigned to cabozantinib, as compared with 5% of the patients assigned to everolimus.

Data pertaining to overall survival, a secondary end point in this trial, were immature at the prespecified interim analysis. Nonetheless, the rate of death was 33% lower with cabozantinib than with everolimus, a finding that indicates a strong trend toward longer survival. The interim boundary for significance was not reached, and follow-up for survival is continuing to the planned final analysis.

The safety profile of cabozantinib in this trial was similar to previous experience in this patient population.²³ Common adverse events with cabozantinib included diarrhea, fatigue, nausea, decreased appetite, the palmar–plantar erythrodysesthesia syndrome, and hypertension, which are also observed with other VEGFR tyrosine kinase inhibitors in patients with renal-cell carcinoma.^8,12 Adverse events that occurred at higher rates and with greater severity with everolimus than with cabozantinib included pneumonitis, peripheral edema, anemia, and hyperglycemia. Dose reductions for management of adverse events occurred more frequently with cabozantinib than with everolimus, underlining the need for careful adverse-event monitoring, as is the case with other VEGFR inhibitors. Discontinuation of study treatment owing to adverse events not related to renal-cell carcinoma occurred in 9% of the patients who received cabozantinib and in 10% of the patients who received everolimus.

This study used a “trial within a trial” design because the sample size required to properly evaluate the primary end point of progression-free survival is too small to adequately assess the important secondary end point of overall survival. An event-driven analysis of progression-free survival in the larger sample required for overall survival could have been overweighted with patients who have early progression, and patients with longer times to progression might not have been sufficiently represented. Therefore, to provide longer follow-up for an event-driven analysis of progression-free survival, the primary analysis of this end point was prespecified to occur in the first 375 patients who underwent randomization.

Everolimus was used as the comparator because it is a standard treatment for patients who previously had disease progression with a VEGFR-targeted therapy. A growing body of evidence suggests greater efficacy with VEGFR inhibitors than with mTOR inhibitors in patients with renal-cell carcinoma.^13,27,28 However, owing to the absence of comparative data from phase 3 trials, an area of controversy has been the relative benefit of a VEGFR inhibitor as compared with everolimus as a second-line treatment.²⁹ More than 70% of our study population were previously treated with only one VEGFR inhibitor, primarily sunitinib. A benefit with respect to progression-free survival was observed with cabozantinib in the subgroup of patients who received one prior VEGFR-targeted therapy, a finding that is consistent with the overall results.

Axitinib is also an option as a second-line treatment for patients with renal-cell carcinoma, given the results of the phase 3 AXIS trial, which showed a benefit in progression-free survival as compared with sorafenib as a second-line therapy.¹² The eligibility criteria of the AXIS trial allowed varied first-line therapies, and the two largest populations were patients who were previously treated with sunitinib (54%) or cytokines (35%). The estimated median progression-free survival in the overall population was 6.7 months with axitinib, as compared with 4.7 months with sorafenib, and the benefit was strongest among patients previously treated with cytokines. The subgroup of patients who received sunitinib as their first-line therapy had an estimated median progression-free survival of 4.8 months and an objective response rate of 11% with axitinib.^12,30 Thus, the estimated median progression-free survival of 9.1 months and the objective response rate of 22% with cabozantinib in a similar population of patients are noteworthy, potentially reflecting the unique mechanism of action of cabozantinib, beyond targeting VEGFR, with the addition of MET and AXL inhibition. Further investigation is required to clearly define the roles of these targets in the clinical activity of cabozantinib.

In a study now reported in the Journal, nivolumab, whose mechanism of action is immune checkpoint inhibition, was associated with an overall survival benefit as compared with everolimus among patients with previously treated advanced renal-cell carcinoma.³¹ Combinations of this agent with cabozantinib in patients with genitourinary cancers, including renal-cell carcinoma, are currently being investigated (ClinicalTrials.gov number, NCT02496208).

In conclusion, cabozantinib, a multitargeted MET, VEGFR, and AXL tyrosine kinase inhibitor, was associated with longer progression-free survival, as compared with everolimus, among patients with renal-cell carcinoma that had progressed after prior VEGFR inhibitor therapy. A strong trend toward longer overall survival with cabozantinib than with everolimus was shown in an interim analysis. A majority of patients who received cabozantinib required dose reductions to manage adverse events.