Case study: ethical issues in the BRIM-3 trial

BRAF targets in melanoma. Biological mechanisms, resistance, and drug discovery. Cancer drug discovery and development. Volume 82. Ed. Ryan J. Sullivan. Springer (2015)

Advanced melanoma is a uniformly lethal disease, with a 5 year survival of less than 10 % [66]. Prior to 2011, dacarbazine, which has a response rate of 5–15 % [67], and high-dose interleukin 2 (response rate 16 %) were the only agents FDAapproved for the systemic therapy for advanced melanoma. While numerous other agents or combination strategies showed promising phase 2 data, all failed to show an overall survival benefit when tested in phase 3 studies against dacarbazine or other routinely used, and similarly effective/ineffective, chemotherapies [68].

In 2002, mutations in the gene encoding BRAF, a serine-threonine kinase of the mitogen-activated protein kinase (MAPK) pathway, were first reported to be present in over half of malignant melanomas, marking the discovery of a critical potential therapeutic target [38]. In June 2009, the results of a phase I clinical trial testing vemurafenib (PLX4032, RG7204, RO5185426) in patients with metastatic melanoma and other solid tumors were presented at the annual American Society of Clinical Oncology (ASCO) meeting. In the seven patients harboring BRAF mutations, five had tumor regression, with disease control lasting up to 14 months [69].

In August 2010, updated results from the phase I trial of vemurafenib were reported in the New England Journal of Medicine; of the 49 patients with melanoma enrolled, 16 harbored the BRAF V600E mutation, and an additional 32 metastatic melanoma patients found to have the V600E mutation were later enrolled in the expansion phase. Of the initial 16 patients who enrolled on the dose-escalation phase, 10 had a partial response and 1 had a complete response, while 24 of the 32 patients enrolled in the expansion had a partial response and 2 had a complete response [70]. Symptom improvement was observed within 1–2 weeks of starting therapy. Of equal importance, no responses were seen in any of the patients treated whose tumors did not harbor a BRAF mutation.

Soon after, the New York Times published a series of articles notably detailing the story of two cousins with metastatic, BRAF-mutant melanoma [71]. One cousin enrolled on the phase II trial of vemurafenib (BRIM2) and responded well to treatment for 9 months at the time of the publication. In contrast, the other cousin was enrolled onto the phase III trial of vemurafenib (BRIM3) and was randomized to the control arm (dacarbazine). Unfortunately, the cousin who was randomized to dacarbazine had rapidly progressing disease which led to clinical decline, and was denied access to vemurafenib according to the protocol, which did not allow for crossover to the treatment arm. He died of metastatic melanoma just over 6 months after diagnosis. While the story of the two cousins did not account for confounding factors that may have explained why one did more poorly than the other, it made a powerful point and sparked a debate as to the ethical nature of the BRIM3 study.

The differing views of the debate were described as part of these series of articles and presented the views of prominent oncologists and researchers in the melanoma community regarding the BRIM3 study. While some argued against extrapolating a survival benefit from a phase I trial showing response, and one in which OS was not a primary endpoint, others argued that the phase I results were convincing proof that vemurafenib should not be withheld from a subset of patients who stand to benefit the most from it. The controversy over the ethics of withholding a non-approved drug from patients with BRAF-mutant melanoma continued for months. Following the first interim analysis, it was clear that the patients randomized to vemurafenib were doing much better than those treated with dacarbazine; results which led to a meeting with the FDA, study sponsor, and study investigators. As a result of this meeting, the end-point of the study was changed from OS to a composite end-point of OS and PFS. Additionally, cross-over to vemurafenib was then allowed.

The results of the BRIM3 study were published in the New England Journal of Medicine on June 30, 2011 [72]. At the time of the interim analysis conducted after 98 deaths, a relative reduction of 63 % in the risk of death was seen with vemurafenib therapy among the 675 previously untreated BRAF-mutant melanoma patients randomly assigned to receive either vemurafenib or dacarbazine. In the analysis at 6 months, overall survival with vemurafenib was significantly improved compared to dacarbazine (84 % vs. 64 %, p < 0.001) [72]. In June 2012, an updated analysis of the BRIM3 study was presented. Importantly, while treatment with vemurafenib was still associated with an improvement in overall survival, the risk reduction of death was 30 %, as opposed to the initial presented 63 %, and in subgroup analysis, the bulk of this could be accounted for in patients with the most aggressive disease (AJCC M1c) [73].

Ethical considerations

  1. Was there equipoise between the two treatments tested in the BRIM-3 randomized clinical trial? What element(s) of trial design could have been modified to address clinicians’ concerns that dying patients were denied a potentially life-prolonging therapy?

As Freedman defines equipoise as the state in which “there is no consensus within the expert clinical community about the comparative merits of the alternatives to be tested” [30], one might question whether equipoise existed for this study. Specifically, nearly every investigator and researcher in the medical community was certain that the BRIM3 study was going to find that vemurafenib was a superior agent to dacarbazine and would improve survival of patients. Still, the melanoma community (and the FDA) had seen a number of promising agents fail in phase 3 studies following compelling phase 2 data; thus the burden of proof was on the sponsor and the investigators to show that vemurafenib was indeed better. The implementation of an interim analysis was used appropriately and action was taken to functionally end the study as originally composed. With the benefit of hindsight it is clear that patients with rapidly growing disease had worse outcomes with dacarbazine than with vemurafenib, and patients with less aggressive disease (unresectable Stage III, Stage IV M1a and M1b) did not. While the first finding would have been predicted from the initial clinical studies with BRAF inhibitors in general and vemurafenib specifically, the latter conclusion would not have been. In our opinion, this information would still likely have been discovered if crossover from dacarbazine to vemurafenib was allowed from the beginning of the study. The FDA typically considers overall survival as the benchmark for oncology drug approvals, although accelerated approval may be granted based on surrogate endpoints such as PFS or objective response rate (ORR) [74]. Still, in the case of bevacizumab in metastatic breast cancer, the FDA revoked its accelerated approval in 2011 [75], stating that the time-to-progression benefit upon which the drug was initially approved failed to translate into a significant improvement in overall survival [76, 77, 78]. OS is an ideal end-point, but one that is difficult to achieve in crossover studies. In the end, the sponsor and investigators carried out the BRIM3 study in good faith to define the effectiveness of a promising agent on the hardest end-point, built in early analysis, and ultimately changed the protocol to allow crossover based on the results of this interim analysis.

  1. Does dacarbazine represent a reference standard against which novel therapies should be measured? Would the use of a placebo as the standard arm be more or less ethical?

When the BRIM3 study was being designed, the frontline setting was chosen for evaluation given the lack of a truly accepted standard of care for this disease. Based on the lethality of this disease and the well defined, though quite modest, activity of chemotherapy, an active control was chosen; as it has been in nearly every randomized, phase III study in the field. Dacarbazine has been used most often in these studies, though other agents/regimens such as single-agent temozolomide, single-agent paclitaxel, and the combination of carboplatin and paclitaxel have also been implemented. The use of placebo would not have been ethical based on the possibility of rapidly progressing, fatal disease in the absence of effective therapy [68].

[contact-form-7 id=»5168″ title=»Контактная форма 1″]


Добавить комментарий

Войти с помощью: 

Ваш e-mail не будет опубликован. Обязательные поля помечены *