MEK Inhibitor-Associated Retinopathy (MEKAR) in Metastatic Melanoma: Long-Term Ophthalmic Effects
Keywords: MEK inhibition, Retinopathy, Melanoma, Long-term treatment, Optical coherence tomography, BRAF inhibition
Abstract
Background: Mitogen-activated protein kinase kinase (MEK) inhibitors have generated significant interest in oncology, demonstrating activity in various cancers, especially melanoma. MEK inhibitors are known to induce a transient retinopathy, considered a class effect. However, there is limited data on the retinal effects of long-term MEK inhibition.
Patients and Methods: In this prospective, observational study, patients with advanced melanoma participating in different phase 1/2 or phase 3 clinical trials were treated with the MEK inhibitor binimetinib, a v-Raf murine sarcoma viral oncogene homolog B (BRAF) inhibitor, or combination therapy. Regular ophthalmological examinations were conducted, including evaluation of visual function, biomicroscopy, dilated fundoscopy, and optical coherence tomography (OCT) for up to two years. Retinopathy was diagnosed based on defined OCT criteria.
Results: Sixty-two patients were evaluated between October 1, 2011, and July 31, 2015: 13 received binimetinib alone, 10 received a selective BRAF inhibitor, and 39 received combination therapy. Binimetinib caused dose-related lesions with serous neuroretinal detachments and edema in 92% of monotherapy patients and 100% of those on combination treatment, with changes strongly dependent on the timing after medication. With continued treatment, retinal volume and thickness decreased to levels below baseline, without apparent functional deficits or structural changes.
Conclusions: Binimetinib induces a specific retinopathy with daily fluctuations, persisting for many months. Retinal thinning, possibly indicating early atrophy, was observed after long-term treatment, but without functional relevance to date.
© 2016 Elsevier Ltd. All rights reserved.
Introduction
MEK inhibitors have garnered considerable attention in oncology by targeting a protein kinase in the mitogen-activated protein kinase pathway, which is implicated in the development of various cancers. Tumor control has been demonstrated in numerous clinical studies. Novel strategies such as targeted therapy and immunomodulation have improved survival in metastatic melanoma. The addition of selective MEK inhibitors to BRAF inhibitor treatment has improved clinical outcomes and delayed drug resistance in BRAF-mutated melanoma patients. MEK inhibitors have also shown activity in NRAS-mutated melanoma. Common adverse events include acneiform skin rashes, edema, retinopathy, and diarrhea. Combined therapeutic regimens reduce the frequency of many adverse events seen in BRAF monotherapy, but MEK inhibitor-related adverse events, including retinal events, persist. Some clinical features of MEK inhibitor-associated retinopathy, resembling central serous chorioretinopathy (CSR), have been described in studies involving about 30 patients and in several case series. There is currently no long-term data available. This study investigated the ophthalmological effects of the selective MEK inhibitor binimetinib alone or in combination with other kinases, including the BRAF inhibitor encorafenib, over many months.
Patients and Methods
2.1. Study Design and Participants
This prospective, observational study included patients with advanced BRAF- and NRAS-mutated cutaneous melanoma, focusing on MEK inhibitor-related ocular side effects during treatment with binimetinib alone or combined with RAF inhibitors, for up to two years. Patients were recruited from seven clinical trials, all treated at the Department of Dermatology, Zurich University Hospital, Zurich, Switzerland, and underwent regular ophthalmological examinations. Patients with increased risk of retinal vein thrombosis, history of vascular occlusion, or previous CSR were excluded. According to their treatment in the original trials, patients were allocated to binimetinib monotherapy, combined treatment with RAF inhibitors, or selective BRAF inhibitor monotherapy. Patients with pre-existing eye conditions were analyzed and reported separately.
2.2. Procedures
All patients underwent full ophthalmological examinations at regular intervals, as per the trial protocol. Examinations included an ophthalmological history, medication timing and dose, best-corrected visual acuity, static perimetry (for phase 1/2 studies), slit lamp examination, applanation tonometry, dilated fundoscopy, and multimodal imaging, especially OCT. At each visit, at least an OCT volume scan centered on the fovea was performed using the same acquisition protocol and rescan technology. Segmentation errors were manually corrected, and only high-quality scans from patients with stable fixation were included. Blue laser autofluorescence was recorded in all patients after one year of treatment, if visual acuity changed, or if symptoms were present.
2.3. Outcomes
The incidence and severity of ocular side effects during long-term binimetinib treatment were evaluated by assessing ocular symptoms, changes in visual function, slit lamp examination, and retinal findings. Retinopathy was diagnosed if localized lesions were visible or if typical bright retinal reflexes were dulled on fundoscopy. OCT-positive retinopathy was defined as serous exudation and/or edema in the outer retinal layers. At each session, all possible adverse retinal events were assessed using consistent criteria for lesion morphology (localized, bullous, or widely extended, flat serous neuroretinal detachments), location (foveal, extrafoveal, or both), presence of edema, and detection of segmentation errors. Edema of the outer layers was defined as an increase in central retinal thickness greater than 10 micrometers compared with baseline values.
2.4. Statistical Analyses
All statistical analyses were performed with R using the ‘lme4’ and ‘lmerTest’ packages. Linear mixed-model analyses assessed influences on visual acuity, central retinal thickness, and retinal volume. The dose of binimetinib, treatment with a RAF inhibitor, duration of treatment, and time since medication were introduced as independent variables, with the subject variable as a random intercept. Baseline values were used as reference. Unless otherwise indicated, figures show mean ± standard deviation. P-values less than 0.05 were considered statistically significant.
Results
Between October 1, 2011, and July 31, 2015, 62 patients were enrolled and evaluated at regular intervals. MEK inhibitor-associated retinopathy was found in 12 (92%) of 13 patients during binimetinib monotherapy and in all 39 (100%) patients on combination therapy. Selective BRAF inhibitor monotherapy in 10 patients caused no retinal changes. The sensitivity of biomicroscopy in detecting typical lesions or evidence of edema was 72.5% compared with OCT.
Five (38%) of the 13 patients on binimetinib monotherapy and 27 (69%) of 39 patients on combination therapy reported symptoms, while 2 (20%) of 10 patients on RAF-inhibitor monotherapy were symptomatic. Visual disturbances occurred especially during the first four weeks. On resuming binimetinib therapy after a break, symptoms were similar but less pronounced. The presence of visual symptoms was highly predictive for retinopathy detection on OCT (odds ratio 2.0, 95% CI 1.4-2.6, p < 0.001). However, only 34 (67%) of 51 patients with OCT-positive retinopathy reported visual symptoms. A dose-dependent attenuation of visual acuity was related to the timing of binimetinib administration (p < 0.001 and p = 0.007). The duration of treatment was not relevant (p = 0.118). Binimetinib transiently attenuates visual acuity after dosing, but prolonged exposure has no long-term effects on this parameter. No significant changes were found on perimetry (mean defect versus baseline, p = 0.979). No abnormalities in Roth 28-hue tests were observed after 6–24 months. Two patients previously treated with ipilimumab had bilateral signs of resolved anterior uveitis and pupillary synechiae at screening. These two, and a third patient with sarcoidosis manifestations, had relapses easily managed with topical treatment. Three patients had intermittently elevated intraocular pressure with generalized edema. In patients on binimetinib alone or with RAF265, fundoscopy revealed bilateral greyish-yellow, round or oval lesions of various sizes, often with a larger lesion over the fovea and others distributed beyond the vascular arcades. Fewer, larger lesions were found on combined binimetinib and encorafenib treatment. After several months, only reduced retinal reflections and slightly blurred optic nerve margins remained as evidence of retinal edema. Scanning laser ophthalmoscopy (SLO) was more acceptable to patients than fundoscopy and allowed questionable retinal lesions to be identified clearly. No abnormalities were found with blue laser autofluorescence. Enhanced depth imaging of choroidal thickness during active retinopathy was within normal limits for age. The morphological appearance of retinal lesions differed between binimetinib monotherapy and combined treatment. Bullous lesions were more common on monotherapy, while combined treatment typically resulted in flat neuroretinal detachments extending over a large part of the posterior pole, seen only on OCT. Patients with bullae tended to report visual disturbances, while flat neuroretinal detachments were less likely to provoke symptoms. OCT findings depended on the time elapsed since drug administration, with maximum retinal changes occurring within four hours of medication. A dose-dependent increase in retinal volume and central retinal thickness was observed, especially during the first few weeks of treatment. These changes resolved gradually over 3–6 months and within a year reached values lower than baseline. Selective BRAF inhibitors had no effects on retinal parameters. Nine of 13 patients on binimetinib monotherapy switched to intermittent treatment due to non-eye-related adverse events. Each time therapy was reintroduced, a mild exudative retinopathy recurred. All patients on combined treatment who continued for six months or more developed flat neuroretinal detachments or just edema, during a short window after medication. In a separate analysis, 17 patients were examined after the end of treatment to determine any residual binimetinib-induced retinopathy. All were symptom-free and visual function tests were unchanged from baseline. No degenerative changes, such as altered pigmentation on fundoscopy or autofluorescence, were found. Retinal volume and thickness remained below pre-treatment values for at least a week after treatment ended. Follow-up suggested possible recovery. High-resolution OCT scans did not detect any irregularities in retinal architecture, particularly in the ellipsoid zone and outer nuclear layer. Seven patients with pre-existing eye disease were analyzed separately. Conditions included unilateral or bilateral macular edema with epiretinal gliosis, extensive drusen with dry age-related retinopathy and glaucomatous optic atrophy but well-controlled intraocular pressure, traumatic optic atrophy, only one functional eye with extreme myopia, and severe panuveitis following ipilimumab. All patients had typical serous neuroretinal detachments, documented with OCT and following a similar course to other participants. Groups of drusen often promoted serous detachments, which took longer to resolve. Discussion This article provides the first description of the long-term ophthalmic effects of binimetinib treatment alone or combined with RAF inhibitors in advanced melanoma. Binimetinib induces a dose-dependent retinopathy in almost all patients. The morphological aspects, with bilateral neuroretinal detachments showing daily fluctuations and edema of the outer retinal layers, are highly specific for MEK inhibitors. Patients on monotherapy, in particular, tend to develop bullous lesions that frequently cause visual disturbances. The addition of the selective BRAF inhibitor encorafenib modulates the retinopathy to give flatter and more extensive detachments. None of the patients experienced persistent functional deficits, even after treatment lasting up to two years. OCT is essential for diagnosis and follow-up, as fundoscopy detected only about 70% of lesions. After an initial increase, retinal thickness gradually decreased, reaching values below screening after six months. Partial recovery of the retina seems to occur once treatment is stopped. MEK inhibitor-induced retinopathy occurs with various MEK inhibitors and is considered a class effect. MEK1/2 inhibitors differ in pharmacokinetics. The maximum plasma concentration of binimetinib is reached in about 1.18 hours, with a half-life of eight hours. Cobimetinib and trametinib have half-lives of approximately two and 4.5 days, respectively. These characteristics may strongly influence the incidence of adverse events, including retinopathy, depending on the timing of examination. The term MEKAR is proposed in analogy to cancer-associated retinopathy (CAR) and melanoma-associated retinopathy (MAR), both rare eye conditions with severe functional deficits in metastatic disease. Autoimmune-based cross-reactions between tumor antigens and retinal tissues have been described. In patients treated with MEK inhibitors, the underlying etiology is suggested to be a dysfunction of the highly active metabolism and fluid regulation of the retinal pigment epithelium (RPE), essential for supplying visual receptors. This is supported by findings of functional deficits in electrophysiological investigations. The pathology is initiated in the outer retinal layers, with the first subtle changes being segmentation errors and RPE edema, followed by changes in the interdigitation zone between the apical processes of the RPE and cone outer segments. Lesions become less pronounced after a few weeks, but a time-dependent retinal response persists. Given that retinopathy was observed in nearly all patients, it is unlikely to be an allergic or autoimmune phenomenon, but rather a toxic reaction to MEK inhibitor exposure. A reduction in retinal measurements to below baseline values was observed, persisting even after discontinuing the MEK inhibitor. The underlying cause is unclear. In inherited retinal degenerative diseases and toxic retinopathy, the functional capacity of a stressed pigment epithelium remains intact for a limited time before atrophy develops. A similar stress reaction and disturbance of homeostasis has been shown in the skin, ultimately leading to chronic changes. The findings suggest partial resolution of retinal thinning after treatment ends. However, persistent degeneration is a slowly evolving process, and the two-year observation period may be too short to detect it. Only binimetinib was investigated in a small group, limiting extrapolation to other MEK inhibitors. Excluding patients with a history of retinal vein occlusion from MEK inhibitor treatment is probably not justified, as the condition described here involves a different retinal compartment. Venous occlusion and macular edema occur in the inner retinal layers, while the pigment epithelium maintains the outer blood-retinal barrier. Classic CSR is thought to be due to pathological changes in choroidal circulation and hormonal influences. Immunomodulation-associated recurrent uveitis may also be explained by MEK inhibitor-induced barrier dysfunction in the otherwise immune-privileged eye. Overall, treatment with binimetinib appears safe regarding ocular adverse effects for up to two years. The appearance of retinal lesions is highly dose- and time-dependent, making the timing of examination important for diagnosis. As about one-third of retinopathies are not detected on fundoscopy,NST-628 OCT is essential for monitoring patients treated with MEK inhibitors.