OS Prolonged With I/O Before Targeted Therapy in Advanced BRAF-Mutant Melanoma

During the November 2021 Session of the American Society of Clinical Oncology Plenary Series, Michael B. Atkins, MD, presented data from the DREAMseq study, which compared 2 treatment sequences for advanced BRAF-mutant melanoma.

The combination of nivolumab (Opdivo) and ipilimumab (Yervoy) followed by the combination of dabrafenib (Tafinlar) and trametinib (Mekinist) appears to be a superior sequence for patients with advanced BRAF V600-mutant metastatic melanoma, compared with the converse sequence of combinations, according to findings from the phase 3 DREAMseq trial.

During the November 2021 Session of the American Society of Clinical Oncology (ASCO) Plenary Series, the data were presented by Michael B. Atkins, MD, the deputy director, and professor, as well as the acting chief in the Division of Hematology/Oncology at MedStar Georgetown University Hospital.

“As background, BRAF V600 driver mutations are present in approximately 50% of patients with metastatic melanoma, BRAF/MEK inhibitor combinations in the 3 [drugs] that are FDA approved, produce significant progression-free survival [PFS] and overall survival [OS] benefits relative to BRAF inhibitor monotherapy,” said Atkins.

The 3 FDA-approved targeted therapy combinations for the treatment of advanced BRAF V600-mutant metastatic melanoma include dabrafenib plus trametinib, vemurafenib (Zelboraf) with cobimetinib (Cotellic), encorafenib (Braftovi) plus binimetinib (Mektovi). The immunotherapy combination of nivolumab plus ipilimumab is also an FDA-approved option for the patient population.

According to Atkins, roughly half of the patients with advanced BRAF V600-mutant metastatic melanoma treated in the United States receive the combination of a BRAF inhibitor and a MEK inhibitor, and about 25% receive nivolumab plus ipilimumab. Despite these combinations being the most popular treatment strategies, no phase 3 prospective research investigated sequencing of these therapies until DREAMseq (NCT02224781).

In the study, 300 patients were randomized into 4 treatment arms. In arm A, patients received nivolumab/ipilimumab induction for 12 weeks followed by nivolumab maintenance for 72 weeks. Patients in arm B were treated with continuous dabrafenib in combination with trametinib. At disease progression, patients in arm C received continuous dabrafenib plus trametinib while those arm D receive nivolumab plus ipilimumab induction followed by nivolumab maintenance.

To evaluate the study end points, patients were stratified by their ECOG performance status as well as their baseline lactate dehydrogenase (LDH) levels.

The primary end point of the study was the 2-year landmark OS in the 270 evaluable patients. The study was 90% powered to detect a superior 2-year OS rate for 70% with the sequence arm A to arm C compared with 50% with the sequence of arm B to arm D. The secondary end points of the study included 3-year landmark OS, FR for death at key time points, objective response rate, PFS, and safety among those with a BRAF mutation treated with nivolumab/ipilimumab, activity with each sequence, the feasibility of crossover, and platform for correlative studies.

As of the data cutoff date of July 16, 2021, 133 patients were treated with nivolumab plus ipilimumab followed by dabrafenib plus trametinib and 132 patients received the converse sequence. At baseline, the median age in arm A was 61 years (range, 35-85) compared with 61 years (range, 30-84) in arm B. The ECOG performance status was 0 in 68% of arm A versus 67% of arm B. The majority of patients has stag M1C disease including 62% of the arm A arm and 58% of arm B. LDH level at baseline were 40% in both sequence groups. Finally, 12% of arm A versus 17% of arm B received prior treatment in the adjuvant setting.

At a median follow-up of 27.7 months, the 2-year OS rate was 72% (95% CI, 62%-79%) when immunotherapy was administered before BRAF/MEK inhibitor therapy versus 52% (95% CI, 42%-60%) with the converse sequence for 20% difference (95% RCI, 3%-38%; log-rank P =.0095). Time to early death was less than 10 months in arm A. In 24 patients, the median OS was 3 months (95% CI, .09-8.4).

“The data safety monitoring committee felt that there was a clinically meaningful difference in overall survival and recommended that the study be close to approval in patients I'm first line [arm B] be given the option to switch arm D without the need for disease progression,” noted Atkins.

Fifty-two percent of patients in the study progressed and crossed over to arm D. Of those patients, 30% from arm were still alive at data cutoff with 10% of arm B. One patient each in arms A and B died after 6 months of progression.

PFS outcomes were assessed in 214 patients from arm A versus arm B. The median PFS was 11.8 months (95%CI, 5.9-33.5) in arm A compared with 8.5 months (6.5-11.3) in arm B for a difference of P =.054. The 1-year PFS rate in arm A was 49% (95% CI, 38%-58%) versus 36% (95% CI, 28%-46%) in arm B. At the 2-year mark, the PFS rate in arm A was 42% (95% CI, 31%-52%) compared with 19% (12%-27%) in arm B.

ORR was evaluated in each treatment arm. Patients in arm A had an ORR of 46%, and the ORR was 43% in am B, 48% in am C, and 30% in arm D. Duration of response (DOR) was compared in arm A versus arm. The median DOR was not reached (NR) in arm A (95% CI, 29.3 to NR) compared with 12.7 months (95% CI, 8.2-NR) in arm B (P <.001).

Grade 3 treatment-related adverse events were observed in all treatment arms. Grade 5 TRAEs were more abundant in arms A and B in 11 and 10 patients, respectively compared with 3 patients each in arms C and D.

In a subgroup analysis of 2-years OS based on the stratification factors, it was shown that having a performance status of 0 and LDH of below the upper limit of normal correlated with a favorable OS outcome.

“Efforts are underway to further identify and best manage this small subset of patients,” Atkins shared in closing.


Atkins MB, Lee S, Chmielowski B, et al. DREAMseq (Doublet, Randomized Evaluation in Advanced Melanoma Sequencing): A phase III trial—ECOG-ACRIN EA6134. Presented at: ASCO Plenary Series: November 2021 Session; November 16, 2021.