Activity of the oral MEK inhibitor trametinib in patients with advanced melanoma: a phase 1 dose-escalation trial

Abstract

Background: MEK is a member of the MAPK signalling cascade that is commonly activated in melanoma. Direct inhibition of MEK blocks cell proliferation and induces apoptosis. We aimed to analyse safety, efficacy, and genotyping data for the oral, small-molecule MEK inhibitor trametinib in patients with melanoma.

Methods: We undertook a multicentre, phase 1 three-part study (dose escalation, cohort expansion, and pharmacodynamic assessment). The main results of this study are reported elsewhere; here we present data relating to patients with melanoma. We obtained tumour samples to assess BRAF mutational status, and available tissues underwent exploratory genotyping analysis. Disease response was measured by Response Evaluation Criteria in Solid Tumors, and adverse events were defined by common toxicity criteria. This study is registered with ClinicalTrials.gov, number NCT00687622.

Findings: 97 patients with melanoma were enrolled, including 81 with cutaneous or unknown primary melanoma (36 BRAF mutant, 39 BRAF wild-type, six BRAF status unknown), and 16 with uveal melanoma. The most common treatment-related adverse events were rash or dermatitis acneiform (n=80; 82%) and diarrhoea (44; 45%), most of which were grade 2 or lower. No cutaneous squamous-cell carcinomas were recorded. Of 36 patients with BRAF mutations, 30 had not received a BRAF inhibitor before; two complete responses (both confirmed) and ten partial responses (eight confirmed) were noted in this subgroup (confirmed response rate, 33%). Median progression-free survival of this subgroup was 5·7 months (95% CI 4·0-7·4). Of the six patients who had received previous BRAF inhibition, one unconfirmed partial response was recorded. Of 39 patients with BRAF wild-type melanoma, four partial responses were confirmed (confirmed response rate, 10%).

Interpretation: Our data show substantial clinical activity of trametinib in melanoma and suggest that MEK is a valid therapeutic target. Differences in response rates according to mutations indicate the importance of mutational analyses in the future.

Funding: GlaxoSmithKline.

Figure 1. Efficacy data for BRAF-mutant, BRAF wild-type, and uveal melanoma patients

Best unconfirmed response for patients with BRAF-mutant, BRAF inhibitor-naïve melanoma (Panel A), BRAF wild-type melanoma (Panel C) and uveal melanoma (Panel E). Positive values indicate tumour growth, negative values indicate tumour reduction, and the dashed line represents the threshold for partial response by RECIST (version 1.0). Loading doses of 6 mg on Day 1, and 6 or 10 mg on Days 1 and 2 followed by 2 mg or 3 mg QD are designated 6/2, 6/6/2, and 10/10/3, respectively. Run-in doses of 0·5 and 1 mg followed by 2 mg or 2·5 mg QD are designated 0·5/2, 1/2, and 1/2·5. All other doses are 2, 2·5 or 3 mg QD. In some instances, patients had less than 20% growth in target lesions but had progressive disease for other reasons (e.g. appearance of new lesions or progression of non-target lesions). Three patients are not included in Panel A due to clinical PD (n=1), SD with no target lesions (n=1), and withdrawal prior to first disease assessment (n=1). Four patients are not included in Panel C due to clinical progression (n=3) and non-radiological assessment (n=1). In Panel E, G indicates either a GNAQ or GNA11 mutation. Panels B (BRAF-mutant melanoma), D (BRAF wild-type melanoma), and F (uveal melanoma) show duration of treatment as measured by time from first dose to time of last dose. Best unconfirmed response and clinical mutational status are indicated. The arrowheads indicate the patient was still receiving treatment at the time of data cut-off. Circles indicate time of disease progression or death, whichever came first. Black triangles indicate patients who continued on treatment for longer than 8 weeks after progression. In panel B, E indicates patients with V600E mutation and K indicates patients with a V600K mutation. Absence of an E or K indicates that there was a BRAF mutation, but the mutation was not specified.

Figure 1. Efficacy data for BRAF-mutant, BRAF wild-type, and uveal melanoma patients

Best unconfirmed response for patients with BRAF-mutant, BRAF inhibitor-naïve melanoma (Panel A), BRAF wild-type melanoma (Panel C) and uveal melanoma (Panel E). Positive values indicate tumour growth, negative values indicate tumour reduction, and the dashed line represents the threshold for partial response by RECIST (version 1.0). Loading doses of 6 mg on Day 1, and 6 or 10 mg on Days 1 and 2 followed by 2 mg or 3 mg QD are designated 6/2, 6/6/2, and 10/10/3, respectively. Run-in doses of 0·5 and 1 mg followed by 2 mg or 2·5 mg QD are designated 0·5/2, 1/2, and 1/2·5. All other doses are 2, 2·5 or 3 mg QD. In some instances, patients had less than 20% growth in target lesions but had progressive disease for other reasons (e.g. appearance of new lesions or progression of non-target lesions). Three patients are not included in Panel A due to clinical PD (n=1), SD with no target lesions (n=1), and withdrawal prior to first disease assessment (n=1). Four patients are not included in Panel C due to clinical progression (n=3) and non-radiological assessment (n=1). In Panel E, G indicates either a GNAQ or GNA11 mutation. Panels B (BRAF-mutant melanoma), D (BRAF wild-type melanoma), and F (uveal melanoma) show duration of treatment as measured by time from first dose to time of last dose. Best unconfirmed response and clinical mutational status are indicated. The arrowheads indicate the patient was still receiving treatment at the time of data cut-off. Circles indicate time of disease progression or death, whichever came first. Black triangles indicate patients who continued on treatment for longer than 8 weeks after progression. In panel B, E indicates patients with V600E mutation and K indicates patients with a V600K mutation. Absence of an E or K indicates that there was a BRAF mutation, but the mutation was not specified.

Figure 1. Efficacy data for BRAF-mutant, BRAF wild-type, and uveal melanoma patients

Best unconfirmed response for patients with BRAF-mutant, BRAF inhibitor-naïve melanoma (Panel A), BRAF wild-type melanoma (Panel C) and uveal melanoma (Panel E). Positive values indicate tumour growth, negative values indicate tumour reduction, and the dashed line represents the threshold for partial response by RECIST (version 1.0). Loading doses of 6 mg on Day 1, and 6 or 10 mg on Days 1 and 2 followed by 2 mg or 3 mg QD are designated 6/2, 6/6/2, and 10/10/3, respectively. Run-in doses of 0·5 and 1 mg followed by 2 mg or 2·5 mg QD are designated 0·5/2, 1/2, and 1/2·5. All other doses are 2, 2·5 or 3 mg QD. In some instances, patients had less than 20% growth in target lesions but had progressive disease for other reasons (e.g. appearance of new lesions or progression of non-target lesions). Three patients are not included in Panel A due to clinical PD (n=1), SD with no target lesions (n=1), and withdrawal prior to first disease assessment (n=1). Four patients are not included in Panel C due to clinical progression (n=3) and non-radiological assessment (n=1). In Panel E, G indicates either a GNAQ or GNA11 mutation. Panels B (BRAF-mutant melanoma), D (BRAF wild-type melanoma), and F (uveal melanoma) show duration of treatment as measured by time from first dose to time of last dose. Best unconfirmed response and clinical mutational status are indicated. The arrowheads indicate the patient was still receiving treatment at the time of data cut-off. Circles indicate time of disease progression or death, whichever came first. Black triangles indicate patients who continued on treatment for longer than 8 weeks after progression. In panel B, E indicates patients with V600E mutation and K indicates patients with a V600K mutation. Absence of an E or K indicates that there was a BRAF mutation, but the mutation was not specified.

Figure 1. Efficacy data for BRAF-mutant, BRAF wild-type, and uveal melanoma patients

Best unconfirmed response for patients with BRAF-mutant, BRAF inhibitor-naïve melanoma (Panel A), BRAF wild-type melanoma (Panel C) and uveal melanoma (Panel E). Positive values indicate tumour growth, negative values indicate tumour reduction, and the dashed line represents the threshold for partial response by RECIST (version 1.0). Loading doses of 6 mg on Day 1, and 6 or 10 mg on Days 1 and 2 followed by 2 mg or 3 mg QD are designated 6/2, 6/6/2, and 10/10/3, respectively. Run-in doses of 0·5 and 1 mg followed by 2 mg or 2·5 mg QD are designated 0·5/2, 1/2, and 1/2·5. All other doses are 2, 2·5 or 3 mg QD. In some instances, patients had less than 20% growth in target lesions but had progressive disease for other reasons (e.g. appearance of new lesions or progression of non-target lesions). Three patients are not included in Panel A due to clinical PD (n=1), SD with no target lesions (n=1), and withdrawal prior to first disease assessment (n=1). Four patients are not included in Panel C due to clinical progression (n=3) and non-radiological assessment (n=1). In Panel E, G indicates either a GNAQ or GNA11 mutation. Panels B (BRAF-mutant melanoma), D (BRAF wild-type melanoma), and F (uveal melanoma) show duration of treatment as measured by time from first dose to time of last dose. Best unconfirmed response and clinical mutational status are indicated. The arrowheads indicate the patient was still receiving treatment at the time of data cut-off. Circles indicate time of disease progression or death, whichever came first. Black triangles indicate patients who continued on treatment for longer than 8 weeks after progression. In panel B, E indicates patients with V600E mutation and K indicates patients with a V600K mutation. Absence of an E or K indicates that there was a BRAF mutation, but the mutation was not specified.

Figure 1. Efficacy data for BRAF-mutant, BRAF wild-type, and uveal melanoma patients

Best unconfirmed response for patients with BRAF-mutant, BRAF inhibitor-naïve melanoma (Panel A), BRAF wild-type melanoma (Panel C) and uveal melanoma (Panel E). Positive values indicate tumour growth, negative values indicate tumour reduction, and the dashed line represents the threshold for partial response by RECIST (version 1.0). Loading doses of 6 mg on Day 1, and 6 or 10 mg on Days 1 and 2 followed by 2 mg or 3 mg QD are designated 6/2, 6/6/2, and 10/10/3, respectively. Run-in doses of 0·5 and 1 mg followed by 2 mg or 2·5 mg QD are designated 0·5/2, 1/2, and 1/2·5. All other doses are 2, 2·5 or 3 mg QD. In some instances, patients had less than 20% growth in target lesions but had progressive disease for other reasons (e.g. appearance of new lesions or progression of non-target lesions). Three patients are not included in Panel A due to clinical PD (n=1), SD with no target lesions (n=1), and withdrawal prior to first disease assessment (n=1). Four patients are not included in Panel C due to clinical progression (n=3) and non-radiological assessment (n=1). In Panel E, G indicates either a GNAQ or GNA11 mutation. Panels B (BRAF-mutant melanoma), D (BRAF wild-type melanoma), and F (uveal melanoma) show duration of treatment as measured by time from first dose to time of last dose. Best unconfirmed response and clinical mutational status are indicated. The arrowheads indicate the patient was still receiving treatment at the time of data cut-off. Circles indicate time of disease progression or death, whichever came first. Black triangles indicate patients who continued on treatment for longer than 8 weeks after progression. In panel B, E indicates patients with V600E mutation and K indicates patients with a V600K mutation. Absence of an E or K indicates that there was a BRAF mutation, but the mutation was not specified.

Figure 1. Efficacy data for BRAF-mutant, BRAF wild-type, and uveal melanoma patients

Best unconfirmed response for patients with BRAF-mutant, BRAF inhibitor-naïve melanoma (Panel A), BRAF wild-type melanoma (Panel C) and uveal melanoma (Panel E). Positive values indicate tumour growth, negative values indicate tumour reduction, and the dashed line represents the threshold for partial response by RECIST (version 1.0). Loading doses of 6 mg on Day 1, and 6 or 10 mg on Days 1 and 2 followed by 2 mg or 3 mg QD are designated 6/2, 6/6/2, and 10/10/3, respectively. Run-in doses of 0·5 and 1 mg followed by 2 mg or 2·5 mg QD are designated 0·5/2, 1/2, and 1/2·5. All other doses are 2, 2·5 or 3 mg QD. In some instances, patients had less than 20% growth in target lesions but had progressive disease for other reasons (e.g. appearance of new lesions or progression of non-target lesions). Three patients are not included in Panel A due to clinical PD (n=1), SD with no target lesions (n=1), and withdrawal prior to first disease assessment (n=1). Four patients are not included in Panel C due to clinical progression (n=3) and non-radiological assessment (n=1). In Panel E, G indicates either a GNAQ or GNA11 mutation. Panels B (BRAF-mutant melanoma), D (BRAF wild-type melanoma), and F (uveal melanoma) show duration of treatment as measured by time from first dose to time of last dose. Best unconfirmed response and clinical mutational status are indicated. The arrowheads indicate the patient was still receiving treatment at the time of data cut-off. Circles indicate time of disease progression or death, whichever came first. Black triangles indicate patients who continued on treatment for longer than 8 weeks after progression. In panel B, E indicates patients with V600E mutation and K indicates patients with a V600K mutation. Absence of an E or K indicates that there was a BRAF mutation, but the mutation was not specified.

Figure 2. Efficacy and exploratory genetics in melanoma subgroups

Exploratory genetics and waterfall plots depicting best unconfirmed response for patients with BRAF-mutant melanoma (Panel A), BRAF wild-type melanoma (Panel B), and BRAF wild-type melanoma with NRAS status (Panel C). Patients in each panel are grouped based on results from local laboratories and/or RGI. Genes commonly implicated in tumourogenesis that were evaluated on the Illumina platform are listed beneath the waterfall plots (Panels A and B); dark squares indicate a mutation determined by Illumina genotyping. Patients who received prior BRAF-inhibitor therapy are highlighted (Panel A) and those who had non-BRAFV600 mutations identified by Illumina genotyping are marked with an asterisk (Panel C). BRAFV600^WT/NRAS^mut patients are represented by hatched bars, and BRAFV600^WT/NRAS^WT patients are represented by solid bars (Panel C). Illumina genotyping identified six NRAS mutations among 23 patients evaluated; two of these six mutations were known from clinical data and four were not. Among the 31 patients who had either clinical or Illumina data for NRAS, 11 (35%) were NRAS-mutant. Scan data were unavailable for three patients (NRAS wild-type, n=1; and NRAS-mutant, n=2) due to progressive disease prior to the first disease assessment.

Figure 2. Efficacy and exploratory genetics in melanoma subgroups

Exploratory genetics and waterfall plots depicting best unconfirmed response for patients with BRAF-mutant melanoma (Panel A), BRAF wild-type melanoma (Panel B), and BRAF wild-type melanoma with NRAS status (Panel C). Patients in each panel are grouped based on results from local laboratories and/or RGI. Genes commonly implicated in tumourogenesis that were evaluated on the Illumina platform are listed beneath the waterfall plots (Panels A and B); dark squares indicate a mutation determined by Illumina genotyping. Patients who received prior BRAF-inhibitor therapy are highlighted (Panel A) and those who had non-BRAFV600 mutations identified by Illumina genotyping are marked with an asterisk (Panel C). BRAFV600^WT/NRAS^mut patients are represented by hatched bars, and BRAFV600^WT/NRAS^WT patients are represented by solid bars (Panel C). Illumina genotyping identified six NRAS mutations among 23 patients evaluated; two of these six mutations were known from clinical data and four were not. Among the 31 patients who had either clinical or Illumina data for NRAS, 11 (35%) were NRAS-mutant. Scan data were unavailable for three patients (NRAS wild-type, n=1; and NRAS-mutant, n=2) due to progressive disease prior to the first disease assessment.

Figure 2. Efficacy and exploratory genetics in melanoma subgroups

Exploratory genetics and waterfall plots depicting best unconfirmed response for patients with BRAF-mutant melanoma (Panel A), BRAF wild-type melanoma (Panel B), and BRAF wild-type melanoma with NRAS status (Panel C). Patients in each panel are grouped based on results from local laboratories and/or RGI. Genes commonly implicated in tumourogenesis that were evaluated on the Illumina platform are listed beneath the waterfall plots (Panels A and B); dark squares indicate a mutation determined by Illumina genotyping. Patients who received prior BRAF-inhibitor therapy are highlighted (Panel A) and those who had non-BRAFV600 mutations identified by Illumina genotyping are marked with an asterisk (Panel C). BRAFV600^WT/NRAS^mut patients are represented by hatched bars, and BRAFV600^WT/NRAS^WT patients are represented by solid bars (Panel C). Illumina genotyping identified six NRAS mutations among 23 patients evaluated; two of these six mutations were known from clinical data and four were not. Among the 31 patients who had either clinical or Illumina data for NRAS, 11 (35%) were NRAS-mutant. Scan data were unavailable for three patients (NRAS wild-type, n=1; and NRAS-mutant, n=2) due to progressive disease prior to the first disease assessment.