Background
Aberrant activation of the mitogen-activated protein kinase (MAPK) pathway is commonly observed in human cancers [
1]. Approximately 50% of cutaneous melanomas harbor mutations in the
BRAF gene, resulting in constitutive activation of the MAPK pathway [
2,
3]. Combined BRAF and MEK inhibition enhances antitumor activity and may prevent or delay development of acquired resistance by providing more potent inhibition of the MAPK pathway [
4,
5]. In the coBRIM study, the combination of cobimetinib, a MEK inhibitor, with vemurafenib, a BRAF inhibitor, significantly improved progression-free survival (PFS) [
6] and overall survival (OS) [
7] compared with placebo and vemurafenib (hereafter referred to as vemurafenib) in advanced
BRAF
V600-mutated melanoma.
A unique ocular adverse event (AE) resembling central serous retinopathy has been described with MEK inhibitors [
4,
8‐
13]. MEK-associated serous retinopathy manifests bilaterally within days after inhibitor initiation [
9,
12,
14]. Patients may present with blurred or impaired vision, but many have no symptoms, with problems detected on ophthalmologic examination [
9,
11]. Funduscopic examination reveals a blunted foveal light reflex and often multiple foveal lesions [
9]. Optical coherence tomography (OCT) is used to detect bilateral subfoveal neurosensory retinal detachments not associated with fluorescein angiography or indocyanine green angiography, indicating unremarkable chorioretinal vasculature [
9,
12].
In contrast, classic central serous retinopathy commonly manifests as unilateral metamorphopsia in middle-aged men, frequently associated with recent corticosteroid use or psychological stress [
12,
15]. On funduscopic examination, typical findings include round, well-delineated, shallow, serous macular neurosensory detachment, often surrounded by a halo light reflex. Vascular abnormalities are usually found on fluorescein and indocyanine green angiography, exhibiting single or multiple discrete leakage points that evenly distribute dye throughout the subretinal fluid [
15]. OCT reveals neurosensory detachment in addition to any associated retinal pigment epithelium detachment. The condition is typically self-limiting, and recovery of visual acuity is usually observed within 1–4 months without the need for treatment [
15].
It is clear that the pathophysiology of MEK inhibitor-associated serous retinopathy is distinct from that of classic central serous retinopathy. Therefore, it is important to further define this AE. This article describes the clinical features of serous retinopathy observed with cobimetinib in patients treated in the coBRIM study.
Methods
Study design and treatment
coBRIM was a multicenter, randomized, double-blind, parallel, placebo-controlled Phase III study designed to evaluate the safety and efficacy of cobimetinib combined with vemurafenib, compared with vemurafenib, in patients with
BRAF
V600 mutation—positive unresectable locally advanced or metastatic melanoma. This trial was registered on
clinicaltrials.gov as NCT01689519. The primary end point was investigator-assessed PFS. Secondary end points included OS, objective response rate, duration of response, PFS as assessed by independent review, safety, pharmacokinetics, and health-related quality of life (HRQOL). Complete methodology of the study and primary efficacy and safety results have previously been published and the protocol is available online [
6]. The study was approved by the institutional review board or ethics committee at each participating institution and was conducted in accordance with the provisions of the Declaration of Helsinki and the International Conference on Harmonisation guidelines for Good Clinical Practice. All the patients provided written informed consent.
Key eligibility criteria were age ≥18 years, histologically confirmed unresectable locally advanced stage IIIC or IV melanoma, BRAF
V600 mutation detected using the cobas
®
4800 BRAF V600 Mutation Test (Roche Molecular Systems Inc. USA), no history of systemic therapy for advanced disease, measurable disease according to Response Evaluation Criteria In Solid Tumors version 1.1, and Eastern Cooperative Oncology Group performance status 0–1. Because of the known ocular toxicities associated with MEK inhibitors, patients with a history or ophthalmic examination evidence of a retinal abnormality considered a risk factor for neurosensory retinal detachment/central serous retinopathy, retinal vein occlusion, or neovascular macular degeneration were excluded. Patients also were excluded if they had risk factors for retinal vein occlusion, including uncontrolled glaucoma with intraocular pressure >21 mmHg, grade ≥2 serum cholesterol, hypertriglyceridemia, or fasting hyperglycemia.
Patients were randomly assigned in a 1:1 ratio using an interactive response system [Perceptive Informatics (now Parexel International), USA] to receive oral vemurafenib (960 mg twice daily) in combination with either placebo or oral cobimetinib (60 mg once daily for 21 days followed by 7 days off). Treatment was administered in 28-day cycles and continued until disease progression, unacceptable toxicity, or withdrawal of consent. Patients were stratified according to the American Joint Committee on Cancer stage and geographic region. Dose modifications for management of specific adverse events were mandated by the protocol. For grade ≥2 visual symptoms, a complete ophthalmic examination was to be performed and treatment with cobimetinib combined with vemurafenib was to be interrupted until resolution to grade ≤1. Dose reduction of the implicated agent was employed if grade ≥2 visual symptoms recurred. Treatment was to be permanently discontinued in the case of retinal vein occlusion, lack of resolution of visual symptoms to grade ≤1 within 28 days, or recurrence of grade ≥2 visual symptoms despite dose reduction.
Ophthalmic assessments
Complete ophthalmic examinations were performed by a qualified ophthalmologist on all patients at screening and day 1 of cycle 2, then every three cycles until cycle 11, every four cycles until cycle 23, every six cycles thereafter or when clinically indicated during the study, and at the end of study treatment visit (Additional file
1). Examinations included visual acuity testing, intraocular pressure measurements by tonometry, slit-lamp ophthalmoscopy, indirect ophthalmoscopy, and OCT (time or spectral-domain).
Analysis
To ensure collection of all potential events, patients who experienced serous retinopathy were identified using a broad group of preferred terms from the Medical Dictionary for Regulatory Activities (MedDRA) (Additional file
2) and record review from ophthalmologic examinations. Events were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) version 4.0 scale for eye disorders-other (Additional file
3). NCI CTCAE version 4.0 does not have a severity grading scale for serous retinopathy. Results were presented and tabulated with descriptive statistics. The data cutoff date for this analysis was 19 Sept 2014.
Discussion
This article describes clinical characteristics of serous retinopathy in patients treated with cobimetinib combined with vemurafenib in the coBRIM study. Because all patients were prospectively screened for visual disturbances throughout the study, per protocol, most events were diagnosed early in the course of treatment, and most patients had no or mild symptoms at diagnosis.
The clinical relevance of surveillance ophthalmic examinations is unclear. Most patients with serous retinopathy had no symptoms, did not require drug discontinuation or dose alteration, and did not experience greater severity of the condition over time. Patients with mildly symptomatic (grade 2) serous retinopathy had dose reduction, resulting in resolution in the majority. For all but one patient (rhegmatogenous retinal detachment) with grade ≥3 serous retinopathy, dose interruption or withdrawal led to improvement without surgery. Few patients experienced recurrence despite continuing or restarting cobimetinib. All recurrent serous retinopathy events were either asymptomatic or mildly symptomatic, and half resolved by the time of data cutoff.
Based on the limited scientific literature available, characteristics of serous retinopathy observed with cobimetinib were similar to those of retinal changes observed with other MEK inhibitors [
9,
11,
12,
14], suggesting that this is a class effect [
11]. Symptomatic ocular toxicity, most commonly blurred vision, has been reported in 0–17% of patients treated with single-agent MEK inhibitors [
16], whereas blurred vision and chorioretinopathy were reported in 0–2% and 1% of patients, respectively, treated with the combination of trametinib and dabrafenib [
17,
18]. Higher incidences of ocular toxicity have been observed in trials incorporating prospective screening, but the reported events are frequently asymptomatic. Prospective screening identified retinopathy in 59% of patients treated with binimetinib, alone or in combination with RAF265 or encorafenib, with symptomatic events reported in 25% of patients [
9], and 26% of patients treated with cobimetinib combined with vemurafenib in the current study, with symptomatic (grade ≥ 2) events in 12% of patients. Given the lack of prospective screening for retinopathy in other trials, incidence cannot be compared across MEK inhibitors. Because of a lack of randomized trials and differences in the description and reporting of ocular toxicity across trials, it is unclear whether the combination of a BRAF inhibitor with a MEK inhibitor alters the incidence of ocular toxicities compared with MEK inhibitor monotherapy. However, in one small study, addition of the pan-RAF inhibitor RAF265 or the selective BRAF inhibitor encorafenib to the MEK inhibitor binimetinib did not appear to influence the incidence of retinopathy [
9]. Furthermore, anecdotal data with binimetinib suggest a temporal relationship between ocular toxicity and dose administration/blood binimetinib levels [
9]. Further research is necessary to determine whether a similar relationship exists for other MEK inhibitors.
Although the MAPK pathway seems to play an important role in maintenance, protection, and repair of the retina, pathophysiologic mechanisms underlying MEK inhibitor-induced retinopathy are not well understood [
11]. The retinal pigment epithelium forms the outer component of the blood–retina barrier and controls solute and fluid permeability, a function critical for maintenance of the specialized environment of the neural retina [
19]; fluid accumulation between these layers can result in retinal detachment [
11]. There is evidence that MAPK signaling regulates density of fluid transport channels (aquaporins) between retinal pigment epithelial cells [
20]. Thus, MEK inhibition may alter permeability of the retinal pigment epithelium, allowing accumulation of subretinal fluid. Further study is needed to test this hypothesis.
In other retinal diseases, the long-term presence of subretinal fluid can lead to photoreceptor death and permanent vision loss; therefore, persistent events may warrant treatment [
15]. Understanding of the pathogenesis of classic central serous chorioretinopathy emphasizes the role of the choroid, which seems to be extremely permeable in this disease [
15]. However, chorioretinal abnormalities are not observed in MEK inhibitor-associated serous retinopathy, in which pathophysiology seems to result instead from altered permeability of the retinal pigment epithelium. Therefore, treatments that might promote subretinal fluid resolution in classic central serous chorioretinopathy [
15] may not be useful for subretinal fluid resolution in MEK inhibitor-associated serous retinopathy. It is reassuring that almost all cases of MEK inhibitor-associated serous retinopathy in the coBRIM study ultimately resolved either spontaneously or through dose modification.
In our experience, management of cobimetinib-associated serous retinopathy necessitates close collaboration with an ophthalmologist, and early involvement should be considered because most events occur within 1 month of treatment initiation. However, prospective screening is probably not warranted because surveillance ophthalmologic examination typically identifies asymptomatic serous retinopathy. Ophthalmologic examination should be performed periodically during cobimetinib treatment and as necessary for new or worsening visual disturbances. Temporarily withholding or reducing the dose of cobimetinib should be considered upon diagnosis of serous retinopathy.
There are several limitations to our results. First, reliance on AE reporting rather than ophthalmic (OCT) confirmation might have resulted in underreporting of asymptomatic or mildly symptomatic cases. Second, ophthalmic images were not prospectively centrally reviewed. Third, visual acuity was not measured in a standardized fashion across sites. Despite these limitations, the large difference in rates of serous retinopathy between the cobimetinib combined with vemurafenib arm and the vemurafenib arm make it unlikely that results were caused by study limitations rather than the real effects of MEK inhibition with cobimetinib.
Authors’ contributions
LCM, LG, JJG, AV, JL, GM, AR, TE, AGE, and BD conceived the study and wrote the manuscript. GB, SE, JS, and AU analyzed the clinical data. All authors critically reviewed the final manuscript. All authors read and approved the final manuscript.
Competing interests
This study was supported by F. Hoffmann-La Roche Ltd, and several co-authors are employees of Roche/Genentech (GB, SE, JJH, AU). LC-M has received research funding from, acted as a consultant/advisor for, and received travel funding from Roche. LDG has received honoraria and travel funds from Roche, Bristol-Myers Squibb, and GlaxoSmithKline; and acted as a consultant/advisor to Italfarmaco. JG has received honoraria and travel funds from Roche, Bristol-Myers Squibb, and GlaxoSmithKline; acted as a consultant/advisor to GlaxoSmithKline, Bristol-Myers Squibb, Roche, Amgen, Merck, and Novartis; received research funding from Roche and Bristol-Myers Squibb; and received travel funding from Roche. GAM has received research funding from Pfizer, Celgene, and Ventana; acted as a consultant/advisor for Provectus; and received travel funding from Roche and Novartis. AR has stock or other ownership in Compugen, CytomX, Five Prime, and Kite Pharma; and has worked in a consulting/advisory role with Pfizer, Merck, Amgen, and Roche. PAA has received research funding from Array, Bristol-Myers Squibb, Roche, and Ventana and acted as a consultant/advisor to Amgen, Array, Bristol-Myers Squibb, Novartis, Merck Sharp & Dohme, Roche, and Ventana. TRJE has served on advisory boards for Bristol-Myers Squibb, Baxter, Clovis, Eisai, GlaxoSmithKline, Celgene, Bayer, Roche, Genentech, Otsuka, and Karus Therapeutics; consulted for Bristol-Myers Squibb, Celgene, Clovis, Baxter, GlaxoSmithKline, Eisai, Roche, Genentech, and Karus Therapeutics; received research funding from Astra Zeneca, Bristol-Myers Squibb, Baxter, Bayer, Clovis, Celgene, Eisai, GlaxoSmithKline, Novartis, Otsuka, Roche Genentech, Merck, Verastem, Vertex, Immunocore, Basilea, Lilly, and Pharmamar; been on speakers bureaus for Bristol-Myers Squibb, Celgene, GlaxoSmithKline, Roche, and Bayer; and received travel funding from Bristol-Myers Squibb and Celgene. AG-E, AV, and JL have no competing interests. BD has received research funding from Bristol-Myers Squibb and GlaxoSmithKline; worked in a consulting/advisory role for Bristol-Myers Squibb, GlaxoSmithKline, Roche, and Novartis; been on speakers’ bureaus for Bristol-Myers Squibb, GlaxoSmithKline, and Roche; and received travel funding from Bristol-Myers Squibb and Roche.