5-Year Outcomes with Cobimetinib Plus Vemurafenib in BRAFV600 Mutation–

Home , Cobimetinib

Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

1 5-Year Outcomes with Cobimetinib Plus Vemurafenib in BRAFV600 Mutation–

2 Positive Advanced Melanoma: Extended Follow-Up of the coBRIM Study

5 Paolo A. Ascierto1, Brigitte Dréno2, James Larkin3, Antoni Ribas4, Gabriella Liszkay5,

6 Michele Maio6, Mario Mandalà7, Lev Demidov8, Daniil Stroyakovskiy9, Luc Thomas10,

7 Luis de la Cruz-Merino11, Victoria Atkinson12, Caroline Dutriaux13, Claus Garbe14, Jessie

8 Hsu15, Surai Jones15, Haocheng Li16, Edward McKenna15, Athina Voulgari17, and Grant

9 A. McArthur18

12 1Istituto Nazionale Tumori IRCCS Fondazione G. Pascale, Naples, Italy; 2University of

13 Nantes, Nantes, France; 3The Royal Marsden Hospital, London, UK; 4Jonsson

14 Comprehensive Cancer Center at the University of California at Los Angeles, Los

15 Angeles, CA, USA; 5National Institute of Oncology, Budapest, Hungary; 6Azienda

16 Ospedaliera Universitaria Senese, Siena, Italy; 7Papa Giovanni XXIII Hospital,

17 Bergamo, Italy; 8N.N. Blokhin Russian Cancer Research Center, Moscow, Russia;

18 9Moscow City Oncology Hospital 62, Krasnogorsk, Russia; 10Centre Hospitalier Lyon

19 Sud, Pierre-Bénite, France; 11Hospital Universitario Virgen Macarena, Seville, Spain;

20 12Princess Alexandra Hospital, Woolloongabba, Queensland, Australia; Gallipoli Medical

21 Research Foundation, Greenslopes, Queensland, Australia; and University of

22 Queensland, St. Lucia, Queensland, Australia; 13Hôpital Saint André, Bordeaux, France;

23 14University of Tübingen, Tübingen, Germany; 15Genentech, a Member of the Roche

Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer Research. Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

24 Group, South San Francisco, CA, USA; 16F. Hoffmann–La Roche Ltd, Mississauga, ON,

25 Canada; 17Roche Products Ltd, Welwyn Garden City, UK; 18Peter MacCallum Cancer

26 Centre and the University of Melbourne, Melbourne, Victoria, Australia.

28 Running Title: 5-Year Outcomes With Cobimetinib + Vemurafenib in Melanoma

30 Keywords: Cobimetinib, Vemurafenib, Melanoma

32 Funding

33 The study was funded by F. Hoffmann-La Roche Ltd.

35 Corresponding author: Paolo A. Ascierto

36 Melanoma, Cancer Immunotherapy and Innovative Therapy Unit

37 Istituto Nazionale Tumori Fondazione Pascale

38 Via Mariano Semmola, 80131 Napoli, Italy

39 Email: [email protected]

40 Phone: +39 081 5903236 41

42 Target Journal: Clinical Cancer Research

44 Conflicts of Interest

45 P.A. Ascierto reports grants and personal fees from Bristol Myers Squibb, Roche-

46 Genentech, Array BioPharma, and Sanofi, personal fees from Merck Sharp & Dohme,

47 Novartis, Merck Serono, Pierre Fabre, Incyte, MedImmune, AstraZeneca, Syndax, Sun

48 Pharma, Idera, Ultimovacs, Sandoz, Immunocore, 4SC, Alkermes, Italfarmaco,

49 Boehringer Ingelheim, Eisai, Regeneron, Daiichi Sankyo, Pfizer, OncoSec, Nouscom,

50 and Lunaphore, and other from Takis, outside the submitted work. B. Dréno reports

51 grants from Roche during the conduct of the study. J. Larkin reports grants and

52 personal fees from Achilles, Bristol Myers Squibb, Merck Sharp & Dohme, Nektar

53 Therapeutics, Novartis, Pfizer, and Roche, grants from Aveo, and Pharmacyclics, and

54 personal fees from AstraZeneca, Boston Biomedical, Eisai, EUSA Pharma,

55 GlaxoSmithKline, Ipsen, Imugene, Incyte, iOnctura, Kymab, Merck, Pierre Fabre,

56 Secarna, Vitaccess, Covance, and Immunocore, outside the submitted work. A. Ribas

57 reports personal fees from Novartis during the conduct of the study, and grants from

58 Agilent and Bristol Myers Squibb, personal fees from Amgen, Chugai, Novartis, Merck,

59 Sanofi, Vedanta Biosciences, and Nurix Therapeutics, and personal fees and other from

60 Arcus, Highlight Therapeutics, Compugen, CytomX, Five Prime Therapeutics, RAPT

61 Therapeutics, Merus, Rgenix, PACT Pharma, Tango Therapeutics, Advaxis, Kite-

62 Gilead, Apricity Therapeutics, ImaginAb, IsoPlexis, Lutris Pharma, MapKure, and 4C

63 Biomed, outside the submitted work. G. Liszkay reports personal fees from Roche,

64 Merck Sharp & Dohme, Novartis, Bristol Myers Squibb, Sanofi-Genzyme, and Pfizer

65 outside the submitted work. M. Maio reports other from Bristol Myers Squibb,

66 AstraZeneca, Roche, Merck Sharp & Dohme, and Merck, during the conduct of the

67 study, and other from Amgen, Pierre Fabre, and Alfasigma, outside the submitted work.

68 M. Mandalà reports personal fees from Novartis, Bristol Myers Squibb, Merck Sharp &

69 Dohme, Pierre Fabre, and Sanofi, outside the submitted work. L. Demidov reports

70 grants and personal fees from Roche during the conduct of the study, grants and

71 personal fees from Novartis, BIOCAD, and Merck Sharp & Dohme, and grants from

72 Amgen, Bristol Myers Squibb, Array Biopharma, and Pfizer, outside the submitted work.

73 D. Stroyakovskiy reports no conflicts of interest. L. Thomas reports no conflicts of

74 interest. L. de la Cruz-Merino reports grants and personal fees from Bristol Myers

75 Squibb, Roche, and Merck Sharp & Dohme, grants from Celgene, and personal fees

76 from Amgen, Pierre Fabre, and Novartis, outside the submitted work. V. Atkinson

77 reports personal fees and nonfinancial support from Bristol Myers Squibb, personal fees

78 from Merck Sharp & Dohme, Merck, Novartis, Nektar Therapeutics, Pierre Fabre, Q

79 Biotics, and Roche, and nonfinancial support from OncoSec, outside the submitted

80 work. C. Dutriaux reports no conflicts of interest. C. Garbe reports grants and personal

81 fees from Roche during the conduct of the study, grants and personal fees from Bristol

82 Myers Squibb, NeraCare, Novartis, and Sanofi, and personal fees from Amgen, Merck

83 Sharp & Dohme, and Philogen, outside the submitted work. J. Hsu and S. Jones report

84 employment with Genentech. H. Li reports employment with Roche Canada. E.

85 McKenna reports employment and stock/stock options with Genentech. A. Voulgari

86 reports employment with Roche. G.A. McArthur reports institutional research grants

87 from Roche-Genentech during the conduct of the study, and institutional research

88 grants from Array Biopharma outside the submitted work.

90 Counts:

91 Abstract: 248 words (limit: 250)

92 Body: 3,732 words (limit: 5000)

93 Tables/figures: 3/3 (limit: 6 total)

94 References: 19 (limit: 50)

95 Statement of Translational Relevance

96 Extended follow-up of the phase III coBRIM study of cobimetinib plus vemurafenib

97 confirms the positive benefit/risk profile of cobimetinib plus vemurafenib in patients with

98 BRAFV600 mutation–positive metastatic melanoma. With at least 5 years of follow up

99 since the last patient was randomized, results continue to demonstrate substantial

100 progression-free survival and overall survival benefits for cobimetinib plus vemurafenib

101 compared with vemurafenib monotherapy. The long-term safety profile is consistent with

102 previously published reports, with no new safety signals detected. Among patients

103 treated with cobimetinib plus vemurafenib, greatest overall survival benefit is observed

104 in patients who achieve a complete response, and in those with normal lactate

105 dehydrogenase levels and low tumor burden at baseline. Novel treatment strategies are

106 needed to improve long-term outcomes for patients in poor prognosis subgroups,

107 particularly those with elevated lactate dehydrogenase levels at baseline.

108

109 Abstract

110 Purpose: The randomized phase III coBRIM study (NCT01689519) demonstrated

111 improved progression-free survival (PFS) and overall survival (OS) with addition of

112 cobimetinib to vemurafenib compared with vemurafenib in patients with previously

113 untreated BRAFV600 mutation–positive advanced melanoma. We report long-term follow-

114 up of coBRIM, with at least 5 years since the last patient was randomized.

115 Experimental Design: Eligible patients were randomized 1:1 to receive either oral

116 cobimetinib (60 mg once daily on days 1–21 in each 28-day cycle) or placebo in

117 combination with oral vemurafenib (960 mg twice daily).

118 Results: 495 patients were randomized to cobimetinib plus vemurafenib (n = 247) or

119 placebo plus vemurafenib (n = 248). Median follow-up was 21.2 months for cobimetinib

120 plus vemurafenib and 16.6 months for placebo plus vemurafenib. Median OS was 22.5

121 months (95% CI, 20.3–28.8) with cobimetinib plus vemurafenib and 17.4 months (95%

122 CI, 15.0–19.8) with placebo plus vemurafenib; 5-year OS rates were 31% and 26%,

123 respectively. Median PFS was 12.6 months (95% CI, 9.5–14.8) with cobimetinib plus

124 vemurafenib and 7.2 months (95% CI, 5.6–7.5) with placebo plus vemurafenib; 5-year

125 PFS rates were 14% and 10%, respectively. OS and PFS were longest in patients with

126 normal baseline lactate dehydrogenase levels and low tumor burden, and in those

127 achieving complete response. The safety profile remained consistent with previously

128 published reports.

129 Conclusions: Extended follow-up of coBRIM confirms the long-term clinical benefit and

130 safety profile of cobimetinib plus vemurafenib compared with vemurafenib monotherapy

131 in patients with BRAFV600 mutation–positive advanced melanoma.

132 Introduction

133 In the last several years, combination regimens of BRAF and MEK targeted agents

134 have improved treatment outcomes in patients with BRAFV600 mutation–positive

135 metastatic melanoma (1–3). Long-term follow-up data from studies of BRAF and/or

136 MEK inhibitors show that about a third of participating patients derive prolonged survival

137 benefit from these regimens (4–7). Long-term follow-up of patients with BRAF-inhibitor–

138 naive BRAFV600 mutation–positive metastatic melanoma in the phase Ib BRIM7 study of

139 the cobimetinib plus vemurafenib combination showed that the overall survival (OS) rate

140 plateaued at 39.2% at years 4 and 5 (4). Similarly, a pooled analysis of patients

141 receiving dabrafenib plus trametinib in the COMBI-d and COMBI-v studies showed OS

142 rates of 37% and 34% at 4 and 5 years, respectively (5). It is of interest to characterize

143 and identify prognostic factors associated with the long-term survival benefits of these

144 treatments, both to inform treatment decisions and to identify new unmet needs for this

145 patient population.

146

147 The phase III coBRIM study evaluated the combination of cobimetinib plus vemurafenib

148 compared with placebo plus vemurafenib in previously untreated patients with advanced

149 melanoma harboring BRAFV600 mutations (1). In the primary analysis, with a median

150 follow-up of 7.3 months, median progression-free survival (PFS) was 9.9 versus 6.2

151 months with cobimetinib plus vemurafenib versus placebo plus vemurafenib (hazard

152 ratio [HR], 0.51; 95% confidence interval [CI], 0.39–0.68; P < 0.001); OS data were

153 immature (1). With a median follow-up of 14.2 months, updated PFS analysis showed

154 that median PFS had improved to 12.3 months with cobimetinib plus vemurafenib

155 versus 7.2 months with placebo plus vemurafenib (HR, 0.58; 95% CI, 0.46–0.72; P <

156 0.0001) (8). With a median follow-up of 18.5 months, the final OS analysis showed that

157 cobimetinib plus vemurafenib significantly improved median OS versus placebo plus

158 vemurafenib (22.3 vs. 17.4 months; HR, 0.70; 95% CI, 0.55–0.90; P = 0.005) (8).

159

160 Herein, we present the extended follow-up analysis of the coBRIM study, after patients

161 were followed for at least 5 years since the last patient was randomized, evaluating

162 long-term treatment outcomes and prognostic characteristics of patients who derived

163 long-term benefit from this regimen.

164

165 Materials and Methods

166 Study design and patients

167 The coBRIM study was a multicenter, randomized, double-blind, placebo-controlled,

168 phase III study that evaluated the efficacy and safety of the combination of cobimetinib

169 plus vemurafenib compared with placebo plus vemurafenib. The study was conducted

170 at 135 sites in the United States, Australia, New Zealand, Israel, and Europe. Detailed

171 methods have previously been published, and the protocol is available online (1).

172 Briefly, eligible patients were aged 18 years or older and had histologically confirmed

173 unresectable stage IIIC or stage IV melanoma harboring a BRAFV600 mutation detected

174 using the cobas® 4800 BRAF V600 Mutation Test (Roche Molecular Systems,

175 Branchburg, NJ, USA); measurable disease as per Response Evaluation Criteria In

176 Solid Tumors version 1.1 (RECIST v1.1) (9); Eastern Cooperative Oncology Group

177 performance status (ECOG PS) of 0 or 1; and adequate hematological, hepatic, renal,

178 and cardiac function. Patients with previously treated brain metastases were eligible if

179 they had a history of stable disease for at least 3 weeks.

180

181 This study was registered with ClinicalTrials.gov (NCT01689519) and was conducted in

182 accordance with the provisions of the Declaration of Helsinki and Good Clinical Practice

183 guidelines. The protocol and its amendments were approved by the local institutional

184 review board or independent ethics committee at each study site. All patients provided

185 written informed consent.

186

187 Treatment

188 Patients were randomized in a 1:1 ratio to receive either oral cobimetinib (60 mg once

189 daily for 21 days followed by a 7-day rest period in each 28-day cycle) or placebo in

190 combination with oral vemurafenib (960 mg twice daily), with stratification by disease

191 stage and geographic region. Treatment was continued until disease progression,

192 death, unacceptable toxicity, or withdrawal of consent. Following the primary analysis,

193 the protocol was amended to allow patients in the placebo plus vemurafenib arm to

194 cross over to the cobimetinib plus vemurafenib arm at the discretion of the investigator if

195 they did not have disease progression and had not discontinued vemurafenib.

196 Modification of the cobimetinib or vemurafenib dose was permitted for management of

197 adverse events (AEs). Tumor assessments were performed at baseline and every 8

198 weeks thereafter. Safety was monitored throughout treatment until 28 days after the last

199 dose of study medication; thereafter, safety reporting included deaths, serious AEs

200 (including new primary cancers), and other AEs of concern considered to be related to

201 prior treatment with study medication. Safety data were regularly reviewed by an

202 independent data and safety monitoring committee.

203

204 Outcomes

205 For the long-term follow-up analysis, outcomes of interest were OS (time from

206 randomization until death from any cause), PFS (time from randomization until disease

207 progression per RECIST v1.1 as assessed by the investigator or death from any cause),

208 confirmed objective response rate (proportion of patients with complete response or

209 partial response per RECIST v1.1), response duration (time from first documentation of

210 response until disease progression or death from any cause), and safety.

211

212 Statistical analysis

213 Patients were enrolled between January 8, 2013, and January 31, 2014. The study was

214 terminated by the sponsor because all patients had at least 5 years of follow-up from

215 the time of randomization, with a final data cut-off date of July 21, 2019. All efficacy

216 analyses were conducted in the intent-to-treat (ITT) population according to original

217 randomized treatment assignment. Safety was assessed in patients who received at

218 least one dose of study treatment according to the actual treatment received. PFS and

219 OS were estimated using the Kaplan-Meier method. OS analyses were performed with

220 and without censoring of data for patients in the placebo plus vemurafenib arm who

221 crossed over to cobimetinib plus vemurafenib. HRs and two-sided 95% CIs were

222 estimated using a stratified Cox proportional hazards model.

223

224 Exploratory analyses were performed to evaluate PFS and OS outcomes in subgroups

225 defined by baseline lactate dehydrogenase (LDH) level (normal vs. elevated) and

226 confirmed best overall response per RECIST v1.1 (complete response vs. partial

227 response vs. nonresponders). OS outcomes were further analyzed in cobimetinib plus

228 vemurafenib–treated patients according to prognostic subgroups previously identified by

229 Hauschild et al. (normal LDH level and sum of the product of longest diameters of target

230 lesions [SLD] ≤45 mm; normal LDH level and SLD >45 mm; elevated LDH level ≤2×

231 upper limit of normal [ULN] and ECOG PS 0; elevated LDH level ≤2× ULN and ECOG

232 PS ≥1; and elevated LDH level >2× ULN) (10) or Long et al. (normal LDH level and <3

233 organ sites; normal LDH level and ≥3 organ sites; elevated LDH level ≤2× ULN and

234 ECOG PS 0; elevated LDH level ≤2× ULN and ECOG PS ≥1; and elevated LDH level

235 >2× ULN) (11).

236

237 Data Sharing Statement

238 Qualified researchers may request access to individual patient level data through the

239 clinical study data request platform (https://vivli.org/). Further details on Roche's criteria

240 for eligible studies are available here (https://vivli.org/members/ourmembers/). For

241 further details on Roche's Global Policy on the Sharing of Clinical Information and how

242 to request access to related clinical study documents, see here

243 (https://www.roche.com/research_and_development/who_we_are_how_we_work/clinic

244 al_trials/our_commitment_to_data_sharing.htm).

245

246 Results

247 A total of 495 eligible patients were enrolled and randomized to receive either

248 cobimetinib plus vemurafenib (n = 247) or placebo plus vemurafenib (n = 248). Two

249 patients, one in each arm, were randomized into the study but did not receive any study

250 treatment. Sixteen patients from the placebo plus vemurafenib arm crossed over to the

251 cobimetinib plus vemurafenib arm following the primary analysis. Baseline

252 characteristics were well balanced between treatment arms (Supplementary Table

253 S1).

254

255 At data cut off, median follow-up duration was 21.2 months (interquartile range [IQR],

256 10.4–59.0) in the cobimetinib plus vemurafenib arm and 16.6 months (IQR, 7.3–42.5) in

257 the placebo plus vemurafenib arm. The most common reasons for discontinuation from

258 the study were death (64% vs. 67%) and withdrawal by patient (8.5% vs. 8.1%) in the

259 cobimetinib plus vemurafenib arm versus the placebo plus vemurafenib arm (Table 1),

260 whereas the most common reasons for discontinuation of study treatment were

261 progressive disease (58% vs. 76%) and adverse events (17% vs. 9%).

262

263 Among patients who discontinued study treatment for reasons other than death,

264 122/240 patients (51%) in the cobimetinib plus vemurafenib arm and 130/227 patients

265 (57%) in the placebo plus vemurafenib arm received subsequent anticancer treatments

266 (Table 2), including immunotherapy (34% vs. 41%), targeted therapy (16% vs. 18%),

267 and chemotherapy (14% vs. 16%).

268

269 Overall Survival

270 At the time of data cut off, 157/247 patients (64%) in the cobimetinib plus vemurafenib

271 arm and 167/248 patients (67%) in the placebo plus vemurafenib arm had died. Median

272 OS was 22.5 months (95% CI, 20.3–28.8) with cobimetinib plus vemurafenib and 17.4

273 months (95% CI, 15.0–19.8) with placebo plus vemurafenib (Fig. 1A). Landmark OS

274 rates with cobimetinib plus vemurafenib versus placebo plus vemurafenib were 38%

275 (95% CI, 32–45) versus 31% (95% CI, 25–37) at 3 years, 34% (95% CI, 28–40) versus

276 29% (95% CI, 23–35) at 4 years, and 31% (95% CI, 25–37) versus 26% (95% CI, 20–

277 32) at 5 years. Results were similar when patients who crossed over to treatment with

278 cobimetinib plus vemurafenib were censored at the date of crossover (Supplementary

279 Table S2).

280

281 A trend for OS benefit for cobimetinib plus vemurafenib over placebo plus vemurafenib

282 was observed in most of the prespecified subgroups (Fig. 1B). Among patients treated

283 with cobimetinib plus vemurafenib, median OS was 38.5 months (95% CI, 28.0 months

284 to not estimable [NE]) for patients with normal baseline LDH level versus 14.8 months

285 (95% CI, 11.3–18.6 months) for patients with elevated baseline LDH level; 5-year

286 landmark OS rates were 43% and 16%, respectively (Fig. 1C). Among patients treated

287 with cobimetinib plus vemurafenib, median OS was prolonged in patients who achieved

288 a confirmed best response of complete response (not reached; 95% CI, 41.2–NE) or

289 partial response (26.5 months; 95% CI, 19.7–33.4) compared with nonresponders (9.7

290 months; 95% CI, 6.1–12.6), with 5-year landmark OS rates of 55%, 29%, and 13%,

291 respectively (Fig. 1D).

292

293 In cobimetinib plus vemurafenib–treated patients analyzed according to subgroups

294 identified by Hauschild et al. (10) (Fig. 2A), OS was most favorable in patients with

295 normal baseline LDH and SLD ≤45 mm and least favorable in those with elevated

296 baseline LDH >2× ULN (Fig. 2B). Five-year OS rates were 52% in patients with normal

297 LDH and SLD ≤45 mm, 34% in those with normal LDH and SLD >45 mm, 22% in those

298 with elevated LDH ≤2× ULN and ECOG PS 0, 15% in those with elevated LDH ≤2× ULN

299 and ECOG PS ≥1, and could not be evaluated in those with elevated LDH >2× ULN

300 because almost all patients had died by 3 years.

301

302 According to subgroups identified by Long et al. (11) (Fig. 2C), OS outcomes in the

303 cobimetinib plus vemurafenib arm were most favorable in patients with normal baseline

304 LDH and <3 organ sites with metastases and least favorable in those with elevated

305 baseline LDH >2× ULN (Fig. 2D). Five-year OS rates were 68% in patients with normal

306 LDH and <3 organ sites, 33% in those with normal LDH and ≥3 organ sites, 22% in

307 those with elevated LDH ≤2× ULN and ECOG PS 0, 15% in those with elevated LDH

308 ≤2× ULN and ECOG PS ≥1, and could not be evaluated in those with elevated LDH >2×

309 ULN because almost all patients had died by 3 years.

310

311 Progression-Free Survival

312 At data cut off, 195/247 patients (79%) in the cobimetinib plus vemurafenib arm and

313 207/248 patients (83%) in the placebo plus vemurafenib arm had PFS events (disease

314 progression per investigator assessment or death). Median PFS was 12.6 months (95%

315 CI, 9.5–14.8) with cobimetinib plus vemurafenib and 7.2 months (95% CI, 5.6–7.5) with

316 placebo plus vemurafenib (Fig. 3A). Landmark PFS rates with cobimetinib plus

317 vemurafenib compared with placebo plus vemurafenib were 23% (95% CI, 17–28)

318 versus 13% (95% CI, 9–18) at 3 years, 17% (95% CI, 12–22) versus 12% (95% CI, 8–

319 16) at 4 years, and 14% (95% CI, 9–19) versus 10% (95% CI, 6–14) at 5 years.

320

321 Among patients treated with cobimetinib plus vemurafenib, median PFS was longer for

322 patients with normal LDH level at baseline (15.0 months; 95% CI, 12.9–22.0) compared

323 with those with elevated baseline LDH level (8.6 months; 95% CI, 7.3–10.0) (Fig. 3B).

324 Among patients treated with cobimetinib plus vemurafenib, median PFS was longer in

325 patients who achieved a confirmed best response of complete response (41.4 months;

326 95% CI, 28.4–66.8) or partial response (12.9 months; 95% CI, 10.6–14.9) compared

327 with nonresponders (3.7 months; 95% CI, 3.6–3.8) (Fig. 3C).

328

329 Best Overall Response

330 In the cobimetinib plus vemurafenib arm, a best overall response of confirmed objective

331 response was achieved in 172/247 patients (70%), with complete response in 52

332 patients (21%) and partial response in 120 patients (49%) (Supplementary Table S3).

333 In the placebo plus vemurafenib arm, 123/248 patients (50%) achieved confirmed

334 objective response, with complete response in 32 patients (13%) and partial response in

335 91 (37%) (Supplementary Table S3). In both treatment arms, the complete response

336 rate improved with longer follow-up. In the cobimetinib plus vemurafenib arm, complete

337 response rates improved from 10% at the primary analysis to 16% at the updated PFS

338 analysis and 21% in the current analysis. In the placebo plus vemurafenib arm, the

339 complete response rate improved from 4% at the primary analysis to 11% at the

340 updated PFS analysis and 13% at the current analysis. The median duration of

341 response was 14.7 months (95% CI, 12.9–19.3) in the cobimetinib plus vemurafenib

342 arm and 9.2 months (95% CI, 7.5–12.9) in the placebo plus vemurafenib arm.

343

344 Safety

345 At data cut off, median duration of cobimetinib treatment was 275 days (range, 4–2245)

346 and median duration of vemurafenib treatment was 281 days (range, 9–2246) for

347 patients in the cobimetinib and vemurafenib arm; median duration of vemurafenib

348 treatment for patients in the placebo plus vemurafenib arm was 175 days (range, 5–

349 2268).

350

351 Treatment-emergent AEs of any grade occurred in 246/248 patients (99%) in the

352 cobimetinib plus vemurafenib arm and 240/245 patients (98%) in the placebo plus

353 vemurafenib arm. AEs occurring in ≥20% patients in either arm are shown in Table 3.

354 Grade ≥3 AEs occurred in 194/248 patients (78%) in the cobimetinib plus vemurafenib

355 arm and 155/245 patients (63%) in the placebo plus vemurafenib arm. The difference in

356 grade ≥3 AE rates between arms was largely attributable to laboratory abnormalities of

357 gamma glutamyltransferase (GGT) increased (15% vs. 10%), alanine aminotransferase

358 (ALT) increased (12% vs. 6%), and blood creatine phosphokinase (CPK) increased

359 (12% vs. <1%) (Table 3).

360

361 Protocol-defined AEs of special interest (AESIs) were actively monitored during the

362 study. AESIs that were more common in the cobimetinib plus vemurafenib arm

363 compared with the placebo plus vemurafenib arm were retinal detachment or central

364 serous retinopathy (27% vs. 5%), grade ≥3 photosensitivity (4% vs. 0), grade ≥3 liver

365 laboratory abnormalities (25% vs. 15%), grade ≥2 ejection fraction reduction (13% vs.

366 4%), and grade ≥3 CPK elevation (12% vs. 1%). Fewer patients in the cobimetinib plus

367 vemurafenib arm had any cutaneous primary malignancy compared with the placebo

368 plus vemurafenib arm (13% vs. 23%). Grade ≥3 rash (17% vs. 17%) and grade ≥3 QTc

369 interval prolongation (2% vs. 2%) occurred in similar numbers of patients in each

370 treatment arm.

371

372 Serious AEs occurred in 105/248 patients (42%) with cobimetinib plus vemurafenib and

373 71/245 patients (29%) with placebo plus vemurafenib. Serious AEs occurring in ≥2% of

374 patients in either arm were pyrexia (3% vs. 1%), pneumonia (2% vs. 1%), and

375 dehydration (2% vs. 0). Grade 5 AEs occurred in 6/248 patients (2%) in the cobimetinib

376 plus vemurafenib arm (cardiac arrest, pneumonia, death [unknown], coma, Clostridium

377 difficile colitis, and myocardial infarction) and 5/245 patients (2%) in the placebo plus

378 vemurafenib arm (cardiac failure, pulmonary embolism, atelectasis, death [unknown],

379 and hemorrhagic stroke).

380

381 AEs leading to discontinuation of any study drug occurred in 67/248 patients (27%) in

382 the cobimetinib plus vemurafenib arm and 30/245 patients (12%) in the placebo plus

383 vemurafenib arm. AEs leading to discontinuation of cobimetinib or placebo in ≥2% of

384 patients were aspartate aminotransferase (AST) increased (2% vs. <1%), ejection

385 fraction decreased (2% vs. 1%), and rash (2% vs. <1%) in the cobimetinib plus

386 vemurafenib arm versus the placebo plus vemurafenib arm. AEs leading to

387 discontinuation of vemurafenib in ≥2% of patients were AST increased (2% vs. 1%),

388 ALT increased (2% vs. 1%), GGT increased (2% vs. 2%), and rash (2% vs. <1%).

389

390 Discussion

391 This extended follow-up of the phase III coBRIM study confirms both the long-term

392 clinical benefit and the safety profile of cobimetinib plus vemurafenib in patients with

393 BRAFV600 mutation–positive metastatic melanoma. With at least 5 years of follow up

394 since the last patient was randomized, results continue to demonstrate substantial PFS

395 and OS benefits for cobimetinib plus vemurafenib compared with vemurafenib

396 monotherapy. The safety profile was consistent with previously published reports (1, 8),

397 with no new safety signals detected over extended follow-up.

398

399 The 5-year OS rate of 31% for cobimetinib plus vemurafenib is consistent with that

400 reported with the combination of dabrafenib plus trametinib (34%) in a pooled analysis

401 of the COMBI-d and COMBI-v studies (5). Similar to the COMBI-d and COMBI-v pooled

402 analysis, we observed a plateauing of both PFS and OS curves from approximately 3

403 years, suggesting favorable survival outcomes for a substantial proportion of patients

404 treated with the combination. Moreover, the 5-year OS rates for cobimetinib plus

405 vemurafenib in patients with normal and elevated LDH (43% and 16%, respectively)

406 were identical to those achieved with dabrafenib plus trametinib in the pooled analysis

407 of the COMBI-d and COMBI-v studies (5). Similar proportions of patients across these

408 studies received subsequent anticancer treatment with immunotherapy or further

409 targeted therapy. Although the extent to which long-term OS is influenced by

410 subsequent treatment is unclear, recent analyses suggest that patients who progress

411 following treatment with BRAF and/or MEK inhibitors derive an OS benefit from

412 subsequent treatment with immunotherapy or additional targeted therapy (12).

413

414 Identification of subgroups of patients likely to achieve long-term treatment benefit is

415 important to inform treatment decisions in the management of patients with metastatic

416 melanoma. Conventional prognostic factors for survival in patients with metastatic

417 melanoma include disease stage, baseline LDH levels, and baseline ECOG PS (13,

418 14). Recent analyses suggest that LDH level at baseline remains the most important

419 prognostic factor for survival in patients treated with targeted therapy or immunotherapy

420 (10, 11, 15, 16). Consistent with these findings, we observed longer PFS and OS for

421 cobimetinib plus vemurafenib–treated patients with normal LDH levels (below ULN) at

422 baseline compared with those who had elevated LDH levels at baseline. In accordance

423 with previous prognostic models in patients with metastatic melanoma treated with

424 BRAF and MEK inhibitor combination therapy, long-term OS was most favorable in

425 patients with normal LDH levels and low tumor burden (defined as either SLD ≤45 mm

426 or <3 organ sites with metastasis) (10, 11). The majority of these patients had long-term

427 survival with cobimetinib plus vemurafenib therapy, with 5-year OS rates of 52% in the

428 subgroup defined by normal baseline LDH and SLD ≤45 mm, and 68% in the subgroup

429 defined by normal baseline LDH and <3 organ sites. The 5-year OS rate of 68% in

430 patients with normal baseline LDH and <3 organ sites treated with cobimetinib plus

431 vemurafenib in the current study compares favorably with the 5-year OS rate of 55% for

432 this subgroup of patients treated with dabrafenib plus trametinib (5). Conversely,

433 patients with elevated baseline LDH >2× ULN had the least favorable long-term OS

434 outcomes, with less than 3% of patients surviving beyond 3 years.

435

436 A previous pooled analysis of four studies in patients with BRAFV600-mutated metastatic

437 melanoma treated with vemurafenib or cobimetinib plus vemurafenib (BRIM-2, BRIM-3,

438 BRIM-7, and coBRIM) demonstrated that greater depth of response was associated

439 with improved survival (17). Among patients treated with cobimetinib plus vemurafenib

440 in the coBRIM study, those who achieved complete response had the best long-term

441 survival outcomes, with 5-year PFS and OS rates of 39% and 55%, respectively,

442 compared with 14% and 31%, respectively, in the overall population. PFS and OS

443 outcomes remained unfavorable in patients who did not achieve objective response,

444 with 5-year PFS and OS rates of only 4% and 13%, respectively, representing an area

445 of unmet need. It was also demonstrated that with at least 5 years of follow-up,

446 complete response rates had increased compared with rates in the primary analysis (1)

447 in both the cobimetinib plus vemurafenib arm (21% vs. 10%) and the placebo plus

448 vemurafenib arm (13% vs. 4%), indicating a beneficial long-term effect with BRAF and

449 MEK inhibitor therapy.

450

451 Novel treatment strategies are needed to improve long-term outcomes for patients in

452 poor prognosis subgroups, particularly those with elevated LDH levels at baseline. One

453 strategy that has been investigated is the addition of immune checkpoint inhibitors to

454 BRAF inhibitor plus MEK inhibitor combination therapy. Recently, the phase III

455 IMspire150 study demonstrated that the combination of atezolizumab plus vemurafenib

456 plus cobimetinib improved investigator-assessed PFS compared with placebo plus

457 vemurafenib plus cobimetinib (median, 15.1 vs. 10.6 months; HR, 0.78; 95% CI, 0.63–

458 0.97; log-rank P = 0.025) in patients with BRAFV600 mutation–positive advanced

459 melanoma (18). Interestingly, exploratory analyses suggested that the treatment benefit

460 observed with atezolizumab plus vemurafenib plus cobimetinib was more pronounced

461 among patients with elevated LDH levels at baseline in the subgroup of patients whose

462 tumors did not express programmed death ligand-1 (PD-L1) (19). However, long-term

463 outcomes are not yet available and further analyses are needed to evaluate subgroups

464 of patients who may benefit from triplet combination therapy.

465

466 This extended analysis of the randomized phase III coBRIM study, with at least 5 years

467 since the last patient was randomized and a relatively low dropout rate, allows robust

468 estimation of long-term OS outcomes. Limitations include the potential confounding

469 effect of postprogression treatments on OS outcomes, the retrospective exploratory

470 nature of analyses according to prognostic subgroups, and the potential influence of

471 immortal time bias on analysis of survival outcomes according to best overall response.

472

473 Long-term follow-up efficacy and safety data continue to support a positive benefit/risk

474 balance for cobimetinib plus vemurafenib compared with placebo plus vemurafenib. The

475 OS results confirmed the long-term OS benefit in patients treated with cobimetinib plus

476 vemurafenib compared with placebo plus vemurafenib in patients with previously

477 untreated, BRAFV600 mutation–positive advanced melanoma. The greatest benefit was

478 observed in patients who achieved a complete response and in those with normal LDH

479 levels and low tumor burden at baseline.

480

481 Authors’ Contributions

482 P. A. Ascierto: Conceptualization, data curation, supervision, investigation, writing-

483 original draft, writing-review and editing. B. Dréno: Conceptualization, resources, data

484 curation, supervision, validation, investigation, visualization, writing-review and editing.

485 J. Larkin: Investigation, writing-review and editing. A. Ribas: Conceptualization, data

486 curation, supervision, investigation, writing-review and editing. G. Liszkay: Resources,

487 validation, investigation, writing-review and editing. M. Maio: Data curation, supervision,

488 investigation, writing-review and editing. M. Mandalà: Resources, data curation, writing-

489 review and editing. L. Demidov: Data acquisition, writing-review and editing. D.

490 Stroyakovskiy: Investigation, writing-review and editing. L. Thomas: Data acquisition,

491 investigation, writing-review and editing. L. de la Cruz-Merino: Validation, visualization,

492 writing-review and editing. V. Atkinson: Supervision, formal analysis, writing-review

493 and editing. C. Dutriaux: Resources, investigation, writing-review and editing. C.

494 Garbe: Data curation, formal analysis, supervision, writing-review and editing. J. Hsu:

495 Formal analysis, methodology, writing-review and editing. S. Jones: Validation, formal

496 analysis, writing-review and editing. H. Li: Methodology, formal analysis, writing-review

497 and editing. E. McKenna: Methodology, formal analysis, writing-original draft, writing-

498 review and editing. A. Voulgari: Conceptualization, resources, data curation, formal

499 analysis, supervision, writing-original draft, writing-review and editing. G. A. McArthur:

500 Conceptualization, supervision, investigation, methodology, writing-review and editing.

501

502 Acknowledgments

503 The authors thank all the patients and their families, and the investigators and research

504 teams who participated in this study. The study was funded by F. Hoffmann-La Roche

505 Ltd. Medical writing and editorial support for this manuscript were provided by Melanie

506 Sweetlove, MSc (ApotheCom, San Francisco, CA, USA), and were funded by F.

507 Hoffmann-La Roche Ltd.

508

509

510

511

512 References

513 1. Larkin J, Ascierto PA, Dreno B, Atkinson V, Liszkay G, Maio M, et al. Combined

514 vemurafenib and cobimetinib in BRAF-mutated melanoma. N Engl J Med

515 2014;371(20):1867–76. doi: 10.1056/NEJMoa1408868.

516 2. Long GV, Stroyakovskiy D, Gogas H, Levchenko E, de Braud F, Larkin J, et al.

517 Combined BRAF and MEK inhibition versus BRAF inhibition alone in melanoma.

518 N Engl J Med 2014;371(20):1877–88. doi: 10.1056/NEJMoa1406037.

519 3. Robert C, Karaszewska B, Schachter J, Rutkowski P, Mackiewicz A, Stroiakovski

520 D, et al. Improved overall survival in melanoma with combined dabrafenib and

521 trametinib. N Engl J Med 2015;372(1):30–9. doi: 10.1056/NEJMoa1412690.

522 4. Ribas A, Daud A, Pavlick AC, Gonzalez R, Lewis KD, Hamid O, et al. Extended

523 5-year follow-up results of a phase Ib study (BRIM7) of vemurafenib and

524 cobimetinib in BRAF-mutant melanoma. Clin Cancer Res 2020;26(1):46–53. doi:

525 10.1158/1078-0432.CCR-18-4180.

526 5. Robert C, Grob JJ, Stroyakovskiy D, Karaszewska B, Hauschild A, Levchenko E,

527 et al. Five-year outcomes with dabrafenib plus trametinib in metastatic

528 melanoma. N Engl J Med 2019;381(7):626–36. doi: 10.1056/NEJMoa1904059.

529 6. Hauschild A, Ascierto PA, Schadendorf D, Grob JJ, Ribas A, Kiecker F, et al.

530 Long-term outcomes in patients with BRAF V600-mutant metastatic melanoma

531 receiving dabrafenib monotherapy: analysis from phase 2 and 3 clinical trials. Eur

532 J Cancer 2020;125:114–20. doi: 10.1016/j.ejca.2019.10.033.

533 7. Robert C, Flaherty K, Nathan P, Hersey P, Garbe C, Milhem M, et al. Five-year

534 outcomes from a phase 3 METRIC study in patients with BRAF V600 E/K-mutant

535 advanced or metastatic melanoma. Eur J Cancer 2019;109:61–9. doi:

536 10.1016/j.ejca.2018.12.015.

537 8. Ascierto PA, McArthur GA, Dréno B, Atkinson V, Liszkay G, Di Giacomo AM, et

538 al. Cobimetinib combined with vemurafenib in advanced BRAFV600-mutant

539 melanoma (coBRIM): updated efficacy results from a randomised, double-blind,

540 phase 3 trial. Lancet Oncol 2016;17:1248–60. doi: 10.1016/S1470-

541 2045(16)30122-X.

542 9. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, et al.

543 New response evaluation criteria in solid tumours: revised RECIST guideline

544 (version 1.1). Eur J Cancer 2009;45(2):228–47. doi: 10.1016/j.ejca.2008.10.026.

545 10. Hauschild A, Larkin J, Ribas A, Dréno B, Flaherty KT, Ascierto PA, et al.

546 Modeled prognostic subgroups for survival and treatment outcomes in BRAF

547 V600–mutated metastatic melanoma: pooled analysis of 4 randomized clinical

548 Trials. JAMA Oncol 2018;4(10):1382–8. doi: 10.1001/jamaoncol.2018.2668.

549 11. Long GV, Grob JJ, Nathan P, Ribas A, Robert C, Schadendorf D, et al. Factors

550 predictive of response, disease progression, and overall survival after dabrafenib

551 and trametinib combination treatment: a pooled analysis of individual patient data

552 from randomised trials. Lancet Oncol 2016;17(12):1743–54. doi: 10.1016/S1470-

553 2045(16)30578-2.

554 12. Ascierto PA, Ribas A, Larkin J, McArthur GA, Lewis KD, Hauschild A, et al.

555 Impact of initial treatment and prognostic factors on postprogression survival in

556 BRAF-mutated metastatic melanoma treated with dacarbazine or

557 vemurafenib ± cobimetinib: a pooled analysis of four clinical trials. J Transl Med

558 2020;18(1):294. doi: 10.1186/s12967-020-02458-x.

559 13. Dickson PV, Gershenwald JE. Staging and prognosis of cutaneous melanoma.

560 Surg Oncol Clin 2011;20(1):1–17. doi: 10.1016/j.soc.2010.09.007.

561 14. Balch CM, Gershenwald JE, Soong S-j, Thompson JF, Atkins MB, Byrd DR, et al.

562 Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol

563 2009;27(36):6199. doi: 10.1200/JCO.2009.23.4799.

564 15. Long G, Blank C, Ribas A, Mortier L, Carlino M, Lotem M, et al. Impact of

565 baseline serum lactate dehydrogenase concentration on the efficacy of

566 pembrolizumab and ipilimumab in patients with advanced melanoma: data from

567 KEYNOTE-006. Eur J Cancer 2017;72:S122-–S3.

568 16. Schadendorf D, Long GV, Stroiakovski D, Karaszewska B, Hauschild A,

569 Levchenko E, et al. Three-year pooled analysis of factors associated with clinical

570 outcomes across dabrafenib and trametinib combination therapy phase 3

571 randomised trials. Eur J Cancer 2017;82:45–55. doi: 10.1016/j.ejca.2017.05.033.

572 17. Lewis KD, Larkin J, Ribas A, Flaherty KT, McArthur GA, Ascierto PA, et al.

573 Impact of depth of response on survival in patients treated with

574 cobimetinib ± vemurafenib: pooled analysis of BRIM-2, BRIM-3, BRIM-7 and

575 coBRIM. Br J Cancer 2019;121(7):522–8. doi: 10.1038/s41416-019-0546-y.

576 18. Gutzmer R, Stroyakovskiy D, Gogas H, Robert C, Lewis K, Protsenko S, et al.

577 Atezolizumab, vemurafenib, and cobimetinib as first-line treatment for

578 unresectable advanced BRAFV600 mutation-positive melanoma (IMspire150):

579 primary analysis of the randomised, double-blind, placebo-controlled, phase 3

580 trial. Lancet 2020;395(10240):1835–44. doi: 10.1016/S0140-6736(20)30934-X.

581 19. Ascierto PA, Robert C, Lewis K, Gutzmer R, Stroyakovskiy D, Gogas HJ, et al.

582 1102P Clinical benefit in BRAFV600 mutation-positive melanoma defined by

583 programmed death ligand 1 (PD-L1) and/or lactate dehydrogenase (LDH) status:

584 exploratory analyses from the IMspire150 study. Ann Oncol 2020;31:S745. doi:

585 10.1016/j.annonc.2020.08.1225.

586

587

588 Tables

589 Table 1. Patient disposition.

Patient disposition, n (%) Placebo + Cobimetinib +

vemurafenib vemurafenib

Study disposition n = 248 n = 247

Discontinued from studya 248 (100) 247 (100)

Death 167 (67) 157 (64)

Lost to follow-up 8 (3) 4 (2)

Withdrawal by patient 20 (8) 21 (9)

Physician decision 1 (<1) 4 (2)

Other 4 (2) 2 (1)

Treatment disposition n = 245 n = 248

Discontinued study treatment 245 (100) 248 (100)

Progressive disease 185 (76) 145 (58)

Adverse event 21 (9) 43 (17)

Death 3 (1) 4 (2)

Other 19 (8) 56 (23)

590 aForty-eight patients in the placebo + vemurafenib arm and 59 patients in the

591 cobimetinib + vemurafenib arm remained on study at the time of study termination.

592 Table 2. Subsequent anticancer treatments.a

Subsequent treatment, n (%) Placebo + Cobimetinib +

vemurafenib vemurafenib

(n = 227) (n = 240)

≥1 subsequent treatment 130 (57) 122 (51)

Chemotherapy 36 (16) 33 (14)

Targeted therapy 40 (18) 38 (16)

BRAF inhibitor 25 (11) 21 (9)

Combined BRAF and MEK 16 (7) 19 (8) inhibitor

MEK inhibitor 1 (<1) 3 (1)

Immunotherapy 94 (41) 82 (34)

Ipilimumab 83 (37) 53 (22)

Anti–PD-1 agents 42 (19) 44 (18)

Ipilimumab plus anti–PD-1 4 (2) 6 (3)

Other 1 (<1) 0

593 aIn patients who discontinued study treatment for reasons other than death.

594

595 Table 3. Most common AEs,a regardless of attribution to study treatment.

AEs, n (%) Placebo + Cobimetinib +

vemurafenib vemurafenib

(n = 245) (n = 248)

Any grade Grade ≥3 Any grade Grade ≥3

Diarrhea 81 (33) 3 (1) 152 (61) 17 (7)

Nausea 65 (27) 2 (1) 108 (44) 3 (1)

Rash 95 (39) 14 (6) 104 (42) 14 (6)

Arthralgia 103 (42) 12 (5) 96 (39) 7 (3)

Fatigue 83 (34) 8 (3) 93 (38) 11 (4)

Blood CPK increased 7 (3) 1 (<1) 91 (37) 30 (12)

Photosensitivity reaction 47 (19) 0 86 (35) 8 (3)

Pyrexia 60 (24) 0 79 (32) 4 (2)

Vomiting 34 (14) 2 (1) 69 (28) 4 (2)

ALT increased 44 (18) 14 (6) 68 (27) 29 (12)

AST increased 30 (12) 4 (2) 66 (27) 23 (9)

GGT increased 43 (18) 25 (10) 59 (24) 38 (15)

Decreased appetite 49 (20) 1 (<1) 55 (22) 0

Asthenia 43 (18) 3 (1) 52 (21) 5 (2)

Anemia 22 (9) 8 (3) 51 (21) 10 (4)

Hypertension 24 (10) 7 (3) 51 (21) 21 (8)

Pruritus 47 (19) 1 (<1) 51 (21) 3 (1)

Headache 40 (16) 4 (2) 50 (20) 2 (1)

Blood alkaline phosphatase 25 (10) 4 (2) 47 (19) 14 (6) increased

Alopecia 75 (31) 1 (<1) 42 (17) 1 (<1)

Myalgia 31 (13) 6 (2) 42 (17) 1 (<1)

Rash maculopapular 38 (16) 13 (5) 39 (16) 18 (7)

Dermatitis acneiform 22 (9) 3 (1) 37 (15) 6 (2)

Ejection fraction decreased 11 (4) 3 (1) 32 (13) 5 (2)

Pain in extremity 39 (16) 6 (2) 32 (13) 3 (1)

Hyperkeratosis 75 (31) 6 (2) 31 (13) 1 (<1)

Basal cell carcinoma 6 (2) 6 (2) 16 (6) 15 (6)

Hyponatremia 3 (1) 1 (<1) 13 (5) 7 (3)

Dehydration 2 (1) 0 13 (5) 6 (2)

Squamous cell carcinoma of 33 (13) 33 (13) 10 (4) 9 (4)

skin

Lipase increased 4 (2) 2 (1) 9 (4) 8 (3)

Pneumonia 5 (2) 2 (1) 7 (3) 6 (2)

Keratoacanthoma 23 (9) 23 (9) 5 (2) 5 (2)

596 aAEs of any grade occurring in ≥20% of patients and AEs of grade ≥3 occurring in ≥2%

597 of patients in either treatment arm.

598 Figure Legends

599

600 Figure 1.

601 Overall survival. Kaplan-Meier curve of overall survival in the ITT population (A), Forest

602 plot of hazard ratios for overall survival across patient subgroups (B), overall survival in

603 cobimetinib plus vemurafenib–treated patients with normal versus elevated LDH levels

604 at baseline (C), and cobimetinib plus vemurafenib–treated patients who achieved

605 complete response or partial response versus nonresponders (D).

606

607 Figure 2.

608 Overall survival outcomes by prognostic subgroups. Decision tree for prognostic

609 subgroups identified by Hauschild et al. (10) in cobimetinib plus vemurafenib–treated

610 patients (A), Kaplan-Meier curves of overall survival across prognostic subgroups

611 identified by Hauschild et al. (10) in cobimetinib plus vemurafenib–treated patients (B),

612 decision tree for prognostic subgroups identified by Long et al. (11) in cobimetinib plus

613 vemurafenib–treated patients (C), and Kaplan-Meier curves of overall survival across

614 prognostic subgroups identified by Long et al. (11) in cobimetinib plus vemurafenib–

615 treated patients (D).

616

617 Figure 3.

618 Progression-free survival. Kaplan-Meier curves of progression-free survival in the ITT

619 population (A), cobimetinib plus vemurafenib–treated patients with normal versus

620 elevated LDH levels at baseline (B), and cobimetinib plus vemurafenib–treated patients

621 who achieved complete response or partial response versus nonresponders (C).

Figure 1A Cobimetinib Placebo 100 + vemurafenib + vemurafenib n = 247 n = 248 Median OS, months 22.5 17.4 80 (95% CI) (20.3–28.8) (15.0–19.8)

60 49%

38% 40 34% 31% 39% Overall survivalOverall(%) 31% 29% 20 Cobimetinib + vemurafenib 26% Placebo + vemurafenib + + Censored 0 0 12 24 36 48 60 72 Time (months) Patients at risk

Cobimetinib + vemurafenib 247 210 169 139 109 91 83 76 70 66 60 33 3 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer Placebo + vemurafenib 248 194 142 104 84 72Research.65 62 59 56 52 25 3 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 1B Placebo + vemurafenib Cobimetinib + vemurafenib n = 248 n = 247 n Events Median n Events Median Hazard 95% Wald Favors cobimetinib Favors placebo (months) (months) Ratio Cl + vemurafenib + vemurafenib

All patients 248 167 17.4 247 157 22.5 0.80 (0.64–0.99) Disease stage IIIc 13 6 19.1 21 8 NE 0.52 (0.18–1.51) M1a 40 17 58.6 40 18 54.8 1.06 (0.54–2.05) M1b 42 26 23.3 40 28 19.4 1.19 0.70–2.03) M1c 153 118 14.8 146 103 18.9 0.73 (0.56–0.96) Disease stage group M1c 153 118 14.8 146 103 18.9 0.73 (0.56–0.96) Unresectable Stage lllc, M1a, M1b 95 49 34.6 101 54 36.6 0.95 (0.64–1.39) Age <65 years 179 118 18.3 183 117 22.1 0.87 (0.67–1.12) ≥65 years 69 49 14.7 64 40 25.8 0.64 (0.42–0.98) Race White 235 158 17.4 227 142 23.8 0.78 (0.62–0.98) Non-white 13 9 19.3 20 15 21.3 1.04 (0.45–2.38) Sex Female 108 70 20.5 101 60 30.0 0.80 (0.57–1.13) Male 140 97 14.9 146 97 21.1 0.78 (0.58–1.03) Geographic region Australia, New Zealand, Others 38 19 23.0 40 20 22.5 0.96 (0.51–1.80) Europe 184 131 16.1 182 116 23.8 0.74 (0.58–0.96) North America 26 17 21.4 25 21 19.2 1.01 (0.53–1.93) ECOG PS Unknown 2 2 NE 4 3 15.7 1.59 (0.16–16.07) 0 166 166 19.8 184 117 23.8 0.94 (0.72–1.23) 1 80 80 11.7 58 37 21.8 0.57 (0.38–0.86) 2 1 0 NE NE (NE–NE) Baseline LDH Unknown 4 4 3.1 4 1 NE <0.01 (0.00–NE) Elevated 104 77 11.2 112 90 14.8 0.87 (0.64–1.18) Normal 140 86 23.8 131 66 38.5 0.71 (0.52–0.98) Previously treated brain metastasis Yes 2 1 NE 1 0 NE <0.01 (0.00–NE) No 246 166 17.4 246 157 22.5 0.80 (0.64–1.00) Prior adjuvant therapy Yes 24 17 19.1 24 16 34.6 0.91 (0.46–1.80) No 224 150 17.4 223 141 22.3 0.79 (0.62–0.99) BRAFV600 mutation status V600E 174 122 17.5 170 113 21.8 0.82 (0.63–1.06) V600K 32 19 16.7 24 13 24.1 0.81 (0.40–1.63)

Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer Research. 1/10 1/5 1/2 1 2 5 10 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 1C

100 LDH normal LDH elevated n = 131 n = 112 Median OS, months 38.5 14.8 80 (95% CI) (28.0–NE) (11.3–18.6)

60 65% 50% 46% 40 43%

Overall survivalOverall(%) 29% 20 LDH normal 23% LDH elevated 19% 16% + + Censored 0 0 12 24 36 48 60 72 Time (months) Patients at risk

LDH normal 131 117 104 90 75 62 56 54 49 46 44 25 2 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer LDH elevated 112 89 61 45 30 25Research.23 18 18 17 13 6 1 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 1D Complete Partial 100 response response Nonresponders n = 52 n = 120 n = 75 Median OS, NR 26.5 9.7 80 months (95% CI) (41.2–NE) (19.7–33.4) (6.1–12.6)

76% 68% 60 61% 55% 53% 40 38% 32%

Overall survivalOverall(%) 29% 20 18% 14% 14% 13% 0 0 12 24 36 48 60 72 Time (months) Patients at risk

Complete response 52 52 50 46 37 33 33 31 29 27 24 15 1 Partial response 120 119 95 77 62 50 42 37 33 31 29 16 2 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer Nonresponders 75 39 24 16 10 Research.88 8 8 8 7 2 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 2A

ITT population N = 247

Baseline LDH normal Baseline LDH abnormal n = 131 n = 112

Baseline LDH >ULN SLD ≤45 mm SLD >45 mm Baseline LDH >2× ULN to ≤2× ULN n = 63 n = 68 n = 33 n = 79

ECOG PS 0 ECOG PS ≥1 n n Downloaded from clincancerres.aacrjournals.org= 62on September 26, 2021. © 2021 = American 16 Association for Cancer Research. Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 2B 100

80 74%

61% 60 57% 58% 52%

40 38% 40% 36% 34% 38% 33% 26%

Overall survivalOverall(%) 22% 20 23% 23% 7% 15% 3% 0 0 12 24 36 48 60 72 Time (months) Patients at risk

LDH normal + 63 58 55 50 42 37 33 32 30 27 26 15 2 SLD ≤45 mm LDH normal + 68 59 49 40 33 25 23 22 19 19 18 10 NE SLD >45 mm LDH elevated ≤2× ULN + 62 55 40 33 22 1921 15 15 14 11 5 1 ECOG PS 0 LDH elevated ≤2× ULN + 16 12 9 6 6 33 3 3 3 2 1 NE ECOG PS ≥1 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer LDH elevated >2× ULN 33 21 11 6 2 Research.11 NE NENE NE NE NE Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 2C

ITT population N = 247

Baseline LDH normal Baseline LDH abnormal n = 131 n = 112

<3 organ sites ≥3 organ sites Baseline LDH >ULN Baseline LDH >2× ULN with metastasis with metastasis to ≤2× ULN n = 33 n = 35 n = 96 n = 79

Figure 2D 100

80 85% 78% 75% 68% 60 57%

40 38% 39% 35% 33% 38% 33% 26%

Overall survivalOverall(%) 23% 22% 20 23% 7% 15% 3% 0 0 12 24 36 48 60 72 Time (months) Patients at risk

LDH normal + 35 33 32 32 28 26 24 23 22 21 20 13 NE <3 organ sites LDH normal + 96 84 72 58 47 36 32 31 27 25 24 12 2 ≥3 organ sites LDH elevated ≤2× ULN + 62 55 40 33 22 1 192 15 15 14 11 5 1 ECOG PS 0 LDH elevated ≤2× ULN + 16 12 9 6 6 33 3 3 3 2 1 NE ECOG PS ≥1 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer LDH elevated >2× ULN 33 21 11 6 2 Research.11 NE NENE NE NE NE Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 3A Cobimetinib Placebo 100 + vemurafenib + vemurafenib n = 247 n = 248 Median PFS, months 12.6 7.2 80 (95% CI) (9.5–14.8) (5.6–7.5)

Cobimetinib + vemurafenib 60 Placebo + vemurafenib + + Censored

40 32%

23% 17% 20 14%

Progression-freesurvival(%) 16% 13% 12% 10% 0 0 12 24 36 48 60 72 Time (months) Patients at risk

Cobimetinib + vemurafenib 247 175 117 85 69 54 44 36 29 24 19 10 2 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer Placebo + vemurafenib 248 122 65 46 33 Research.27 24 23 21 19 15 4 1 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 3B

100 LDH normal LDH elevated n = 131 n = 112 Median PFS, months 15.0 8.6 80 (95% CI) (12.9–22.0) (7.3–10.0)

40% 40 28% 23% 18% 20 LDH normal 19%

Progression-freesurvival(%) LDH elevated 14% + + Censored 8% 7% 0 0 12 24 36 48 60 72 Time (months) Patients at risk

LDH normal 131 100 74 54 46 35 29 25 20 15 11 6 1 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer LDH elevated 112 71 39 27 19 16Research.12 8 7 7 6 3 1 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Figure 3C

Complete Partial 100 response response Nonresponders n = 52 n = 120 n = 75 Median PFS, 41.4 12.9 3.7 80 months (95% CI) (28.4–66.8) (10.6–14.9) (3.6–3.8)

60 68%

57% 40 46% 29% 39%

20 19% 12% 9%

Progression-freesurvival(%) 6% 4% 4% 4% 0 0 12 24 36 48 60 72 Time (months) Patients at risk

Complete response 52 51 46 38 33 27 23 18 16 13 9 6 1 Partial response 120 108 64 42 33 26 20 17 12 10 9 3 1 Downloaded from clincancerres.aacrjournals.org on September 26, 2021. © 2021 American Association for Cancer Nonresponders 75 16 7 5 3 Research.11 1 1 1 1 1 Author Manuscript Published OnlineFirst on June 22, 2021; DOI: 10.1158/1078-0432.CCR-21-0809 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

5-Year Outcomes with Cobimetinib Plus Vemurafenib in BRAF V600 Mutation-Positive Advanced Melanoma: Extended Follow-Up of the coBRIM Study

Paolo A Ascierto, Brigitte Dreno, James Larkin, et al.

Clin Cancer Res Published OnlineFirst June 22, 2021.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-21-0809

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2021/06/17/1078-0432.CCR-21-0809.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet been Manuscript edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2021/06/17/1078-0432.CCR-21-0809. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.