Georgina V. Long, M.D., Ph.D., Daniil Stroyakovskiy, M.D., Helen Gogas, M.D., Evgeny Levchenko, M.D., Filippo de Braud, M.D., James Larkin, M.D., Claus Garbe, M.D., Ph.D., Thomas Jouary, M.D., Axel Hauschild, M.D., Ph.D., Jean Jacques Grob, M.D., Ph.D., Vanna Chiarion Sileni, M.D., Celeste Lebbe, M.D., Mario Mandalà, M.D., Michael Millward, M.D., Ana Arance, M.D., Ph.D., Igor Bondarenko, M.D., Ph.D., John B.A.G. Haanen, M.D., Ph.D., Johan Hansson, M.D., Ph.D., Jochen Utikal, M.D., Ph.D., Virginia Ferraresi, M.D., Nadezhda Kovalenko, M.D., Ph.D., Peter Mohr, M.D., Volodymyr Probachai, M.D., Dirk Schadendorf, M.D., Ph.D., Paul Nathan, M.D., Caroline Robert, M.D., Antoni Ribas, M.D., Ph.D., Douglas J. DeMarini, Ph.D., Jhangir G. Irani, M.A., Michelle Casey, Ph.D., Daniele Ouellet, Ph.D., Anne-Marie Martin, Ph.D., Ngocdiep Le, M.D., Ph.D., Kiran Patel, M.D., and Keith Flaherty, M.D.
Targeted inhibition of the RAF–MEK–ERK (MAPK) pathway with BRAF inhibitors dabrafenib or vemurafenib, as compared with chemotherapy, improves the progression-free and overall survival of patients who have metastatic melanoma with BRAF V600 mutations.1,2 However, resistance develops in a majority of patients, resulting in a median progression-free survival of 6 to 7 months.3,4 Most reported resistance mechanisms reactivate the MAPK pathway.5-7 In addition, BRAF-inhibitor–induced paradoxical activation of the MAPK pathway8-10 can result in secondary cancers, including cutaneous squamous-cell carcinoma, and may reactivate RAS-mutant tumors.11-13 Independently, single-agent trametinib, a MEK inhibitor, improves the overall survival of patients with BRAF V600 mutation–positive metastatic melanoma, as compared with chemotherapy, and is not associated with paradoxical activation.14
The combination of BRAF and MEK inhibition, as compared with single-agent BRAF inhibition, delayed the emergence of resistance and decreased the incidence of cutaneous hyperproliferative lesions in preclinical models.12,15,16 This finding was confirmed in an unblinded, randomized, phase 2 study of a combination of dabrafenib and trametinib, as compared with dabrafenib alone, with crossover permitted at the time of disease progression for those receiving dabrafenib alone.17The median progression-free survival was 9.4 months with combination therapy versus 5.8 months with dabrafenib alone (hazard ratio for progression or death in the dabrafenib–trametinib group, 0.39; 95% confidence interval [CI], 0.25 to 0.62; P<0.001) and the corresponding response rate was 76% versus 54% (P=0.03).17 The rate of cutaneous squamous-cell carcinoma was lower in the dabrafenib–trametinib group than in the dabrafenib-only group (7% vs. 19%).
We conducted a double-blind, randomized, phase 3 study without crossover to compare the combination of dabrafenib and trametinib with dabrafenib alone as first-line therapy in patients who had metastatic melanoma with BRAF V600E or V600K mutations.
Eligible patients had histologically confirmed, unresectable stage IIIC or stage IV metastatic melanoma with BRAF V600E or V600K mutations, as determined by means of an investigational-use-only polymerase-chain-reaction assay (ThxID BRAF Assay, bioMérieux) performed at a central reference laboratory. Patients were not eligible if they had previous systemic anticancer therapy (including BRAF or MEK inhibitors). Patients with brain metastases that had been definitively treated and stable for at least 12 weeks were eligible to participate. Additional inclusion and exclusion criteria are provided in the Supplementary Appendix, available with the full text of this article at NEJM.org. Written informed consent was obtained from all study patients.
Study Design and Treatments
From May 2012 through January 2013, we screened 947 patients at 113 centers worldwide. We randomly assigned 423 of these patients, in a 1:1 ratio, to receive either a combination of oral dabrafenib (150 mg twice daily) and oral trametinib (2 mg once daily) or oral dabrafenib (150 mg twice daily) and placebo. Patients were stratified according to the baseline lactate dehydrogenase level and BRAF genotype (Figure S1 in the Supplementary Appendix).
Study End Points
The primary end point was investigator-assessed progression-free survival, defined as the time from randomization until radiologic disease progression or death from any cause. The secondary end points were overall survival, response rate, response duration, safety, and pharmacokinetics, as defined in the Supplementary Appendix. An independent central review committee whose members were unaware of study-group assignments reviewed radiologic findings on which a sensitivity analysis of progression-free survival was based. No interim analyses were performed for efficacy or futility with respect to the primary end point.
Tumor assessments were conducted according to Response Evaluation Criteria in Solid Tumors (RECIST), version 1.1,18 at baseline, at week 8, every 8 weeks until week 56, and then every 12 weeks until disease progression, death, or withdrawal from the study. Patients who were eligible for continued treatment beyond progression underwent tumor assessments according to the protocol. (Eligibility criteria for this assessment are provided in the Supplementary Appendix.) All responses were confirmed on either computed tomography or magnetic resonance imaging, with scanning obtained at least 4 weeks after the first RECIST response. Adverse events were graded by the site investigator according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0,19 throughout the study until 30 days after the discontinuation of study treatment.
This study was funded by the sponsor, GlaxoSmithKline. The protocol was approved by the institutional review board or human research ethics committee at each participating center and complied with country-specific regulatory requirements. The study was conducted in accordance with the provisions of the Declaration of Helsinki and Good Clinical Practice guidelines. The first and last authors and representatives of the sponsor designed the study. Data collection was performed by staff members employed at each study site and was monitored by the sponsor. The first author wrote the initial drafts of the manuscript, with support from the last author and representatives of the sponsor. All authors and sponsor representatives had full access to the study data and were involved in the data analysis. All the authors made the decision to submit the manuscript for publication and vouch for the accuracy of the data and the fidelity of the study to the protocol, which is available at NEJM.org. No one who is not listed as an author contributed to the writing of the manuscript.
This report is based on data as of August 2013, when the prespecified number of disease progressions or deaths (whichever came first) had occurred. The study was initially designed with a power of more than 90% to detect a 41% reduction in the risk of disease progression or death (hazard ratio, 0.59) in the dabrafenib–trametinib group, as compared with the dabrafenib-only group, at a one-sided type I error rate of 0.025. The study was overenrolled by approximately 24% (423 actual enrollees vs. the target of 340). To increase the precision of the median progression-free survival estimate in the dabrafenib–trametinib group, the final analysis was planned after 193 events had occurred, which maintained the same ratio of events to patients as originally planned. As a result, the power increased from 90% to 95%. A prespecified interim analysis of overall survival was planned at the time of the analysis of progression-free survival. The stopping boundary for the interim analysis of overall survival was a two-sided alpha level of less than 0.00028. A final overall survival analysis will be conducted when 70% of the patients who underwent randomization have died or been lost to follow-up.
We used a Kaplan–Meier analysis to estimate progression-free and overall survival, and treatment comparisons were made with the use of a stratified log-rank test. Post hoc subgroup analyses were performed with the use of an unstratified log-rank test. Efficacy was determined in all patients in the intention-to-treat population; safety analyses were performed in all patients who received at least one dose of a study drug. Treatment beyond progression was defined as the receipt of a study drug more than 15 days after radiologic progression, as defined by RECIST.
Of 947 patients who were screened, 423 patients were randomly assigned to receive either dabrafenib plus trametinib (211 patients) or dabrafenib plus placebo (212 patients) (Table 1TABLE 1Characteristics of the Patients at Baseline (Intention-to-Treat Population)., and Figure S1 in the Supplementary Appendix). The lack of a BRAF V600E or V600K mutation was the most common reason for exclusion (in 245 patients). Baseline characteristics were similar in the two study groups (Table 1). The median follow-up time was 9 months (range, 0 to 16). Details of patient status and follow-up are provided in Figure S1 in the Supplementary Appendix.
In the intention-to-treat population, the estimated median progression-free survival (the primary end point) was longer in the dabrafenib–trametinib group than in the dabrafenib-only group (9.3 months vs. 8.8 months); the hazard ratio for disease progression or death was 0.75 (95% CI, 0.57 to 0.99; P=0.03) (Figure 1AFIGURE 1Progression-free Survival and Subgroup Analyses.). The progression-free survival benefit favored ºthe dabrafenib–trametinib group in most of the subgroups that were analyzed (Figure 1B).
In patients with an elevated lactate dehydrogenase level, the median progression-free survival was 7.1 months in the dabrafenib–trametinib group as compared with 3.8 months in the dabrafenib-only group (hazard ratio for disease progression or death, 0.64; 95% CI, 0.42 to 0.95), with 49 events in each group (64% vs. 69%) (Figure 1C). Among patients in the two study groups, there were fewer events in the subgroup with a normal lactate dehydrogenase level: 53 events (40%) in the dabrafenib–trametinib group versus 60 events (43%) in the dabrafenib-only group (hazard ratio, 0.83; 95% CI, 0.57 to 1.20) (Figure S2A in the Supplementary Appendix).
Within the first 2 months after randomization, data for 6 patients in the dabrafenib–trametinib group and 18 patients in the dabrafenib-only group were censored in the progression-free survival analysis (for details, see the Results section in the Supplementary Appendix). Of the 18 patients in the dabrafenib-only group for whom data were censored, 13 had disease progression on the basis of clinical indications (without radiologic confirmation), as determined by the investigator, or had started a new anticancer therapy. In preplanned sensitivity analyses, when clinical progression or initiation of a new anticancer therapy was considered as an event, the hazard ratio for progression and the median progression-free survival for the dabrafenib–trametinib group remained stable (i.e., the median remained the same when clinical progression was considered or decreased by 0.1 month when the initiation of a new anticancer therapy was considered). In contrast, the median progression-free survival in the dabrafenib-only group decreased by 1.2 months when clinical progression was considered and by 1.6 months when the initiation of a new anticancer therapy was considered (Table S1 in the Supplementary Appendix).
At the time of the interim survival analysis, 40 patients (19%) in the dabrafenib–trametinib group and 55 patients (26%) in the dabrafenib-only group had died (hazard ratio for death, 0.63; 95% CI, 0.42 to 0.94; P=0.02). At 6 months, the overall survival rate was 93% in the dabrafenib–trametinib group and 85% in the dabrafenib-only group. However, the between-group difference did not cross the prespecified stopping boundary (two-sided P=0.00028). At the time of this report, neither group had reached a median overall survival (Figure 2AFIGURE 2Overall Survival.). Patients with an elevated lactate dehydrogenase level appeared to derive a greater survival benefit from combination therapy over monotherapy, with deaths of 24 of 77 patients (31%) and 36 of 71 patients (51%), respectively. The median survival among patients with an elevated lactate dehydrogenase level receiving combination therapy was 13.7 months versus 8.9 months among those receiving dabrafenib alone (hazard ratio for death, 0.48; 95% CI, 0.29 to 0.80) (Figure 2B). In contrast, among patients with a normal lactate dehydrogenase level, deaths were reported in only 16 of 133 patients (12%) receiving combination therapy and 19 of 140 patients (14%) receiving dabrafenib alone (Figure S2B in the Supplementary Appendix).
The overall response rate as assessed by investigators was 67% (95% CI, 60 to 73) in the dabrafenib–trametinib group versus 51% (95% CI, 45 to 58) in the dabrafenib-only group (P=0.002) (Table 2TABLE 2Disease Response, According to the Type of BRAF Mutation.). In the dabrafenib–trametinib group, 22 patients (10%) had a complete response, and 118 (56%) had a partial response. In the dabrafenib-only group, 18 (9%) had a complete response, and 90 (43%) had a partial response. The median duration of response was 9.2 months in the dabrafenib–trametinib group and 10.2 months in the dabrafenib-only group on the basis of data that were highly censored because the majority of investigator-assessed responses (60%) were still ongoing. Within each of the two study groups, the response rates were slightly lower in the BRAFV600K subgroup than in the predominant V600E subgroup (Table 2).
Other Anticancer Therapies
Fewer patients in the dabrafenib–trametinib group than in the dabrafenib-only group received a second type of anticancer therapy (43 [20%] vs. 65 [31%]) (Table S2 in the Supplementary Appendix). In contrast, more patients in the dabrafenib–trametinib group than in the dabrafenib-only group continued treatment beyond progression (41 [19%]) vs. 34 [16%]), including at the time of data cutoff (19 [46%] vs. 7 [21%]). Ipilimumab was the most common subsequent anticancer therapy in the two groups; however, fewer patients in the dabrafenib–trametinib group than in the dabrafenib-only group received ipilimumab (19 [9%] vs. 32 [15%]). A similar number of patients in the two groups received subsequent therapy with vemurafenib: 12 (6%) in the dabrafenib–trametinib group and 13 (6%) in the dabrafenib-only group. No patients received monoclonal antibodies targeting programmed death 1 and programmed death ligand 1.
Among patients who received at least one dose of a study drug, permanent discontinuations were reported in 19 of 209 patients (9%) in the dabrafenib–trametinib group and 11 of 211 patients (5%) in the dabrafenib-only group. In addition, dose reductions were required in 52 patients (25%) and 28 patients (13%), respectively, and interruption of a study drug because of adverse events occurred in 103 patients (49%) and 70 patients (33%), respectively. In both the dabrafenib–trametinib group and the dabrafenib-only group, pyrexia was the most common reason for dose interruptions (in 67 patients [32%] and 28 patients [13%]) and dose reductions (in 27 patients [13%] and 6 patients [3%]). The most common reason for permanent discontinuation was pyrexia in the dabrafenib–trametinib group (in 5 patients [2%]) and a decreased ejection fraction in the dabrafenib-only group (in 3 patients [1%]).
The most common adverse events in the two study groups were pyrexia, fatigue, nausea, headache, chills, diarrhea, arthralgia, rash, and hypertension (Table 3TABLE 3Adverse Events.). Fewer patients in the dabrafenib–trametinib group than in the dabrafenib-only group had incident cutaneous squamous-cell carcinoma (2% vs. 9%) or cutaneous hyperkeratoses (3% vs. 32%). Papillomas, alopecia, and the hand–foot syndrome were also less common in the dabrafenib–trametinib group. However, pyrexia, hypertension, peripheral edema, and diarrhea were more common in the dabrafenib–trametinib group than in the dabrafenib-only group (proportion of patients with pyrexia, 51% vs. 28%) (Table 3, and Figure S3 in the Supplementary Appendix). There were no cases of retinal-vein occlusion. Of the 107 patients in the dabrafenib–trametinib group who had pyrexia, 67 (63%) had a second episode, and 50 (47%) had three or more episodes. The median time to the onset of the first episode of pyrexia was 4.3 weeks, and the median duration was 3 days. During an acute episode of pyrexia, 63% of patients had a dose interruption and 25% had a dose reduction, with resolution of pyrexia in 97% of the patients. Glucocorticoids were used for 3 weeks or more as prophylaxis for pyrexia in 30 patients in the dabrafenib–trametinib group after the first or second event. Pharmacokinetic analysis showed a possible association between pyrexia and exposure to the hydroxy-dabrafenib metabolite and, to a lesser extent, to dabrafenib in the two study groups. Similarly, trametinib exposure was associated with pyrexia in the combination group (see the Methods and Results sections and Figure S4 and S5 in the Supplementary Appendix).
Grade 3 or 4 adverse events occurred in 73 patients (35%) in the dabrafenib–trametinib group and 79 patients (37%) in the dabrafenib-only group. Noncutaneous cancers and new primary melanomas were uncommon (Table S3 in the Supplementary Appendix). Four deaths related to adverse events occurred in the dabrafenib–trametinib group: three from cerebral hemorrhage (two during receipt of study treatment and one 5 days after cessation) and one from pneumonia (22 days after cessation of treatment). All four events were considered by the investigator to be unrelated to study treatment. No deaths related to adverse events occurred in the dabrafenib-only group (Table S3 in the Supplementary Appendix).
In this phase 3 study, we found a significant 25% relative reduction in the risk of disease progression among patients with metastatic melanoma with BRAF V600E or V600K mutations who received first-line treatment with a combination of dabrafenib and trametinib, as compared with dabrafenib alone. The combination therapy, as compared with monotherapy, was associated with a 37% relative reduction in the risk of death at the interim survival analysis, but this result did not cross the prespecified stopping boundary. Notably, this difference in survival was observed despite the increased use of new therapies after the discontinuation of monotherapy, including therapies with a proven survival benefit (e.g., ipilimumab).20,21 The overall response rate was also significantly higher in the group receiving both dabrafenib and trametinib. Resistance to single-agent BRAF inhibition is acquired predominantly through reactivation of the MAPK pathway.5-7,22-26 Our study shows that the inhibition of the MAPK pathway at two nodes rather than one decreases the risk of progression (and therefore delays resistance) by 25%.
In a post hoc subgroup analysis, we found that patients with poorer prognostic features appeared to benefit more from a combination of dabrafenib and trametinib than did those in the overall study population. At the time of the interim survival analysis, patients with an elevated lactate dehydrogenase level who received the combination therapy, as compared with monotherapy, had a 52% relative reduction in the risk of death (hazard ratio, 0.48). Since events tend to occur earlier in patients with poorer prognostic features, longer follow-up will be required to assess the effects of combination therapy as compared with monotherapy in patients with better prognostic features, such as a normal lactate dehydrogenase level or stage M1a or M1b melanoma (i.e., only distant skin, subcutaneous, or nodal metastases or lung metastases).
Our study used a control regimen that had a proven survival benefit.2 It had a double-blind design and did not allow crossover at the time of progression. In addition, the median progression-free survival of 8.8 months in the dabrafenib-only group was longer than that in previous trials of dabrafenib and vemurafenib in which the median progression-free survival was 5.5 to 6.9 months.1-4,17,27-29 The long plateau at the median progression-free survival point in the dabrafenib-only group may account in part for its increased value in our study (Figure 1A). In addition, the preplanned sensitivity analysis showed that the median progression-free survival for dabrafenib was unstable. Data for patients who had clinical progression or received a new anticancer therapy without radiographic evidence of progression were censored (which occurred more frequently in the dabrafenib-only group than in the combination-therapy group in the first 2 months of the study). Thus, the median progression-free survival for the dabrafenib group decreased from 8.8 months to 7.6 months when clinical progression was included as an event and decreased from 8.8 months to 7.2 months when receipt of a new anticancer therapy was included.
Patients receiving the combination therapy required more dose modifications than did those receiving monotherapy. There was no significant between-group difference in the frequency of adverse events, including grade 3 and 4 toxic effects. However, the types of toxic effects differed. Hyperproliferative cutaneous events, including cutaneous squamous-cell carcinoma and other events related to paradoxical activation of the MAPK pathway, which show the oncogenic potential of BRAF inhibitors,11-13 were significantly abrogated by the addition of the MEK inhibitor trametinib.12 There were no deaths in either study group that were considered by the investigators to be related to treatment.
Pyrexia occurred more frequently in the combination group than in the monotherapy group and was the most common reason for treatment discontinuation. Although there was an apparent relationship between pyrexia and exposure to the hydroxy-dabrafenib metabolite and trametinib in this study, the cause of pyrexia remains unknown. Pyrexia has been shown to respond well to immediate temporary cessation of dabrafenib or both dabrafenib and trametinib, which were restarted once the patient was well and afebrile for at least 24 hours; prophylactic glucocorticoids may prevent recurrent episodes.30 The education of patients regarding pyrexia and the need to interrupt treatment at the first sign of this side effect is critical to ensure that the episode is not prolonged or complicated by dehydration or hypotension.
In conclusion, in our study involving previously untreated patients who had metastatic melanoma with BRAF V600E or V600EK mutations, the combination of dabrafenib and trametinib, as compared with dabrafenib alone, reduced the risk of progression by 25% and improved the response rate. At the time of the interim analysis, overall survival was longer in the combination group, but the prespecified stopping boundary was not crossed.