None of three combination therapies tested among patients with advanced renal cell carcinoma in the BeST trial came close to achieving the primary objective of a 67% improvement in median progression-free survival compared to single-agent bevacizumab (Avastin), Keith T. Flaherty, MD, reported at the 11th International Kidney Cancer Symposium in Chicago.
“We were looking to see a 67% improvement, a rather bold attempt admittedly. But we figured that based on the phase I data, to absorb the cost of toxicity and then potentially the cost of doublet therapy that might become chronic treatment in the future, one would really need to see a significant improvement in this screening study to warrant further investigation,” said Dr. Flaherty. He is lead investigator of the BeST trial (the Eastern Cooperative Oncology Group’s E2804), a randomized phase II study of VEGF, RAF kinase, and mTOR combination targeted therapy with bevacizumab, sorafenib (Nexavar), and temsirolimus (Torisel) in advanced renal cell carcinoma. He is also Director of The Henri and Belinda Termeer Center for Targeted Therapy, Cancer Center, Massachusetts General Hospital, and Associate Professor, Department of Medicine, Harvard Medical School, Boston.
Reporting data that matured sufficiently for presentation just 2 weeks before the symposium, Dr. Flaherty added that all three combination arms met one of the secondary objectives of response rates greater than 20%. For the key secondary endpoint of safety, “the severe toxicity rate was substantially higher with each of the two-drug combinations compared to single-agent bevacizumab,” Dr. Flaherty said. “You see a near doubling in terms of the rate of grade 3 and a clear doubling if you combine grade 3 and 4 toxicities for each of the combination arms compared to single-agent bevacizumab.”
The symptomatic toxicities were hypertension, fatigue, hand-foot syndrome, and diarrhea. The laboratory toxicities were hypophosphatemia, proteinuria, and hyperglycemia. “For both symptomatic toxicities and laboratory toxicities, we are seeing an accentuation of expected toxicities in terms of greater severity, but not new unexpected toxicities,” Dr. Flaherty said.
Targeting Multiple Pathways
“The design of the trial was not to be a definitive study, but rather a randomized phase II trial specifically to try to identify the most active two-drug combination,” Dr. Flaherty explained. That combination could then be tested against the best available single-agent therapy in a phase III trial. “At the time, sunitinib [Sutent] looked to be that agent,” Dr. Flaherty said, but “concurrent with the launch of this trial, the results were emerging from two sunitinib phase I trials suggesting that it could not be safely combined with either temsirolimus or bevacizumab.”
In the BeST trial, bevacizumab, with known single-agent efficacy, served as the control (arm A) and was compared to three combination regimens identified in previous phase I trials. Arm B consisted of bevacizumab/temsirolimus, arm C was bevacizumab/sorafenib, and arm D was sorafenib/temsirolimus. Each arm included between 81 and 87 patients, mostly male.
At the time the BeST trial was conceived, the hypothesis was that targeting multiple growth and survival pathways in vascular endothelial cells would result in suppression of escape mechanisms arising with antiangiogenic therapy in renal cell carcinoma. “Bevacizumab, sorafenib, and temsirolimus had all shown single-agent efficacy to a greater or lesser degree,” Dr. Flaherty stated. Combining bevacizumab with sorafenib or temsirolimus was envisioned as a way to block escape mechanisms upstream and downstream, and combining sorafenib and temsirolimus was considered a parallel pathway blockade.
Study Design
Key eligibility criteria were > 75% clear cell histology, with prior nephrectomy required unless there was a high burden of disease elsewhere, and no prior therapy with VEGF, VEGFR, or mTOR inhibitors. Single-agent bevacizumab was given at the previously evaluated dose for renal cell carcinoma—10 mg/kg administered intravenously every 2 weeks (on days 1 and 15). This dose was also used in arm B, combined with temsirolimus at 25 mg IV weekly (days 1, 8, 15, and 22). Bevacizumab doses were attenuated in arm C to 5 mg/kg IV every 2 weeks (days 1 and 15), combined with sorafenib at 200 mg administered orally twice daily on days 1 to 5, 8 to 12, 15 to 19, and 22 to 26. In arm D, sorafenib at 200 mg orally twice daily on days 1 to 28 was combined with temsirolimus at 25 mg IV weekly (days 1, 8, 15, and 22).
Among patients who were able to stay free of disease progression and on therapy, “bevacizumab as a single agent as well as in combination held its dose intensity reasonably well from cycle to cycle,” Dr. Flaherty reported. Doses of temsirolimus and sorafenib were attenuated, and sorafenib, which was started at a substantial dose reduction compared to its single-agent phase II dose, was further attenuated due to toxicity.
Response and Survival Rates
Disease progression, occurring in 60% of patients “was the main reason that patients discontinued therapy, although adverse events were notable at 21%,” Dr. Flaherty said. The median progression-free survival was 8.7 months in the bevacizumab-alone arm, 7.3 months for the bevacizumab/temsirolimus arm, and 7.7 months for the sorafenib/temsirolimus arm.
“Only the bevacizumab/sorafenib arm had the suggestion of a slightly improved outcome”—11.3 months—Dr. Flaherty continued. But after doing the preplanned statistical analyses, the hazard ratio (0.84) only slightly favored this arm, a 16% improvement in progression-free survival, he added. “We were targeting a 67% improvement, meaning the hazard ratio needed to be a good deal lower to approximate statistical significance. The P values did not come close to significance [see Table 1].”
In the absence of improved progression-free survival, response rates were “deemed to be the potential tie-breaker,” Dr. Flaherty noted, with regimens producing response rates greater than 20% considered to be worthy of further study. “Although the progression-free survival data did not suggest superiority, … response rate was superior with each of the doublets, compared to single-agent bevacizumab,” Dr. Flaherty said. “Just focusing on objective responses, it does appear that the combinations have an impact on tumor regression,” he added, noting that the pairwise comparisons between arms were all statistically significant (see Table 2).
For overall survival—another secondary endpoint—there was “no hint of difference in outcome, not even a suggestion of improvement,” Dr. Flaherty said. “It is fairly clear that no combination arm was superior to single-agent bevacizumab for the progression-free survival primary endpoint.”
In response to a question from a symposium participant about whether dialing up and down the doses of the drugs could have changed the outcome, Dr. Flaherty noted that this was not allowed in the E2804 trial. He acknowledge the possibility that “full-dose bevacizumab with just a small fraction of a dose of another agent, or the other way around—say, a whisper of bevacizumab with full-dose sorafenib or temsirolimus—could have been potentially more advantageous, but we did not have the freedom to explore that,” Dr. Flaherty said.
He also noted that survival data continue to be gathered and will reflect not only the impact of the study therapies but of subsequent treatment received by trial participants. Those data are currently immature, he said, but they could eventually prove useful in understanding the course of the disease after treatment with these trial regimens. ■
Disclosure: Dr. Flaherty reported no potential conflicts of interest.
