Sotrastaurin | Inhibitor Kits

Serous Retinopathy Associated with Mitogen-Activated Protein Kinase Kinase Inhibition (Binimetinib) for Metastatic Cutaneous and Uveal Melanoma

Elon H.C. van Dijk, MD,1 Carla M.L. van Herpen, MD, PhD,2 Marina Marinkovic, MD,1,3 John B.A.G. Haanen, MD, PhD,4 Drake Amundson, BSc,5 Gré P.M. Luyten, MD, PhD,1 Martine J. Jager, MD, PhD,1 Ellen H.W. Kapiteijn, MD, PhD,6 Jan E.E. Keunen, MD, PhD,7 Grazyna Adamus, PhD,5 Camiel J.F. Boon, MD, PhD1

Abstract

Purpose: To analyze the clinical characteristics of a serous retinopathy associated with mitogen-activated protein kinase kinase (MEK) inhibition with binimetinib treatment for metastatic cutaneous melanoma (CM) and uveal melanoma (UM), and to determine possible pathogenetic mechanisms that may lead to this retinopathy.
Design: Prospective observational, cohort-based, cross-sectional study.
Participants: Thirty CM patients and 5 UM patients treated with the MEK inhibitor binimetinib (CM) or a combination of binimetinib and the protein kinase C inhibitor sotrastaurin (UM).
Methods: Extensive ophthalmic examination was performed, including Early Treatment of Diabetic Retinopathy Study best-corrected visual acuity, applanation tonometry, slit-lamp examination, indirect ophthalmoscopy, digital color fundus photography, and optical coherence tomography (OCT). In selected cases, additional examinations were performed, including visual field testing and electro-oculography (EOG). Blood samples were obtained from 3 CM patients and 3 UM patients to analyze the presence of autoantibodies against retinal and retinal pigment epithelium (RPE) proteins.
Main Outcome Measures: Visual symptoms, visual acuity, fundus appearance, characteristics on OCT, fundus autofluorescence (FAF), and EOG.
Results: Six CM patients (20%) and 2 UM patients (40%) reported visual symptoms during the study. The median time to the onset of symptoms, which were all mild and transient, was 3.5 days (range, <1 hour to 3 weeks). On OCT, subretinal fluid (SRF) was detected in 77% of CM patients and 60% of UM patients. In the 26 patients with SRF, the fovea was affected in 85%. After the start of the medication, an EOG was performed in 19 eyes of 11 patients; 16 of these eyes (84%) developed SRF on OCT. Fifteen of these eyes (94%) showed an abnormal Arden ratio (<1.65). A broad pattern of anti-retinal antibodies was found in 3 CM patients and 2 UM patients tested, whereas anti-RPE antibodies were detected in all 6 tested patients.
Conclusions: A time-dependent and reversible serous retinopathy can develop both in patients with metastatic CM and UM treated with binimetinib. A minority of patients develop visual symptoms, which are generally mild and transient. A cause of binimetinib-associated serous retinopathy may be toxicity of medication, but autoantibodies also may be involved. Ophthalmology 2015;122:1907-1916 ª 2015 by the American Academy of Ophthalmology.

Introduction

Cutaneous melanoma (CM) is an increasingly occurring malignant skin tumor. Distant metastases, either via lymphatics or the bloodstream, occur in 15% of patients, with a 5-year survival rate of 15% to 22%.1 Although CM represents only 4% of skin cancer cases, it causes 65% of skin cancererelated deaths.2
Uveal melanoma (UM) is the most common primary intraocular malignancy in adults. Metastasis occurs in 34% of patients within 10 years after diagnosis, and after detection of metastases, the 2-year survival rate is only 8%.3 Metastasis in UM occurs purely hematogenously, and the liver is involved in 95% of patients. The median survival in these cases is 4 to 6 months, whereas the median survival in cases without liver metastases is 19 to 28 months.4
The treatment options for metastatic CM and UM are limited. However, new targeted treatment options include therapies that target the mitogen-activated protein kinase (MAPK) signaling pathway. The MAPK pathway plays a crucial role in intracellular signal transduction and induces transcription of genes, encoding several cellular processes, such as growth, differentiation, migration, inflammation, angiogenesis, and cell death.5,6 The MAPK pathway is regulated by several growth factors, including fibroblast growth factor receptors, which are present on the cell surface. The MAPK cascade can be activated by the rat sarcoma small guanidine phosphatase, which subsequently activates rapidly accelerated fibrosarcoma kinase, enabling the phosphorylation and activation of mitogen-activated protein kinase kinase (MEK) and, finally, extracellular signal-regulated kinase.7
Patients with CM or UM can be subdivided on the basis of underlying genetic mutations in the melanoma cells. In 52% to 86% of patients with UM and in more than 66% of patients with CM, uncontrolled activation and aberrant signaling of the MAPK pathway can be detected.8 In CM, RAS and BRAF gene mutations may cause activation of the MAPK pathway.9 In UM, mutually exclusive mutations in GNAQ and GNA11 often can be found. These mutations cause a permanently activated GTP-bound state, leading to the activation of protein kinase C and consequently the MAPK pathway.8,9
Several targeted therapies, such as the BRAF inhibitors vemurafenib and dabrafenib, the MEK inhibitors trametinib, selumetinib, and binimetinib (MEK162), and combinations of BRAF and MEK inhibitors, have emerged as novel therapeutic options, especially for patients with metastatic BRAF-mutated CM.10e12 In patients with NRAS-mutated CM, MEK inhibition by MEK162 has shown clinical activity.13
Retinopathy, described as a “central serous chorioretinopathy-like event,” an unspecified “chorioretinopathy,” or a “bilateral, multifocal, mild and self-limiting retinopathy,” was described in association with binimetinib treatment in patients with advanced or metastatic CM in 2% to 65% of patients.12,14
Little is known about the clinical characteristics, outcome, and pathogenesis of MEK inhibitoreassociated retinopathy. In this study, we analyzed the clinical characteristics of retinopathy associated with MEK inhibition treatment for metastatic CM and UM. In addition, we studied possible pathogenetic mechanisms that may lead to such retinopathy, and we examined a correlation between the development of retinopathy and the tumor response to MEK inhibition treatment in metastatic CM.

Methods

Patient Characteristics

Twenty patients from an academic medical center (Radboud University Medical Center, Nijmegen, The Netherlands) and 10 patients from a comprehensive cancer center (The Netherlands Cancer Institute Antoni van Leeuwenhoek, Amsterdam, The Netherlands, in collaboration with the general hospital Onze Lieve Vrouwe Gasthuis, Amsterdam, The Netherlands) were included. These patients had a histologically confirmed, locally advanced, and unresectable or metastatic CM harboring BRAF V600E or NRAS mutations, confirmed by a central laboratory.12 From another academic medical center (Leiden University Medical Center, The Netherlands), 5 patients with liver biopsy-confirmed metastatic UM were included.
The local ethics committee approved the study. Each patient signed informed consent, and the studies were performed in accordance with the Declaration of Helsinki. The clinical trials were registered in the Clinical Trial Registration as numbers NCT01320085 (Nijmegen and Amsterdam) and NCT01801358 (Leiden). Within these trials, patients were included from July 2011 to April 2014.

Treatment

In patients with metastatic CM, the MEK inhibitor binimetinib was administered in a phase II trial setting. Patients received 45 mg of binimetinib orally twice per day, continuously for 28 days (defined as a 1-treatment cycle, for scheduling purposes). The patients with metastatic UM received a combination of binimetinib and the protein kinase C inhibitor sotrastaurin (AEB071) in a phase Ib/II trial setting. Patients with metastatic UM started with continuous dosing of binimetinib 30 or 45 mg twice daily for 28 days. They also received sotrastaurin twice daily. Patients continued with the study medication until the development of disease progression or unacceptable toxicity.

Ophthalmic Examinations

All patients received a complete ophthalmic examination, including Early Treatment of Diabetic Retinopathy Study (ETDRS) best-corrected visual acuity (BCVA), applanation tonometry, slit-lamp examination, and indirect ophthalmoscopy. When ETDRS BCVA was not available (in 10 patients), Snellen BCVA was determined and a previously established conversion method was used to achieve ETDRS values.15
Dilation of pupils was achieved by topical administration of 1% tropicamide and 5% phenylephrine drops. Optical coherence tomography (OCT), using spectral-domain OCT (Spectralis HRAþOCT; Heidelberg Engineering, Heidelberg, Germany) or the Cirrus OCT device (Carl Zeiss Meditec, Dublin, CA), and digital color fundus photography (Topcon Corp., Tokyo, Japan) were performed at baseline, at day 15 of treatment cycle 1 (CM) or at day 8 and day 22 of treatment cycle 1 (UM), day 2 of each treatment cycle from cycle 2, and at the end of treatment (within 14 days after discontinuation of treatment). In selected cases, enhanced-depth imaging OCT, fundus autofluorescence (FAF), and fluorescein angiography (FA) were performed with spectral-domain OCT. In addition, electro-oculography (EOG) and electroretinography (ERG) were performed according to the guidelines of the International Society for Clinical Electrophysiology of Vision in selected cases, as well as color vision testing using the Desaturated Panel D15 test.16e18
In25 patients (20withCMand5 withUM), centralvisual analysis was performed (Humphrey Field Analyzer, Carl Zeiss Meditec; 24-2 or 30-2 SITA Standard Algorithm). One patient underwent microperimetry using the MAIA microperimeter (CenterVue, SpA, Padova, Italy), examining an area covering the central 15.
Patients who reported new or worsened visual symptoms received an additional ophthalmic examination at their earliest convenience. Additional ophthalmic examinations were performed as clinically indicated. When signs of ocular toxicity occurred, patients received weekly ophthalmologic examinations until the symptoms resolved or stabilized after at least 6 months of follow-up.
In this study, serous retinopathy was defined as retinal lesions on ophthalmoscopy corresponding to a localized separation of the neuroretina and retinal pigment epithelium (RPE) on OCT.

Dose Modification in Patients with Ophthalmologic Symptoms

Adverse events were assessed according to the Common Terminology Criteria for Adverse Events version 4.03.19 The classification of adverse events into 4 grades and dose changes after the occurrence of adverse events specified in the study protocols can be found in the Supplemental Text 1 (available at www.aaojournal.org).

Analysis of Autoantibodies in Serum

Sera from patients were collected after the start of binimetinib treatment. Sera were examined for anti-retinal and anti-RPE autoantibodies by Western blotting using human retinal and RPE proteins extracted from the retina or RPE, dissected from the human eye. A description of the analyses can be found in the Supplemental Text 2 (available at www.aaojournal.org).

Safety and Efficacy Monitoring

During every visit to the outpatient clinic of the Department of Medical Oncology, a report was completed. After clinical assessment, the dose of study medication could be reduced when unacceptable signs of toxicity or disease progression had developed. Every 2 treatment cycles, a computed tomography scan was made for evaluation of disease. The main efficacy end points of the study were progression-free survival and overall survival.

Statistical Analysis

Both ETDRS BCVA at the time point of most pronounced subretinal fluid (SRF) on OCT and at final follow-up were compared with EDTRS BCVA at screening, using an independent t test in IBM SPSS Statistics, version 20.0 (SPSS Inc., Chicago, IL). The same statistical test was used to assess the differences in progression-free survival and overall survival in patients with CM, in whom EOG abnormalities and SRF on OCT were and were not detected.

Results

Patient Characteristics

The 30 patients with CM and 5 patients with UM (22 male, 13 female) were aged 56.3 years (median, 56 years; range, 32e77 years). The clinical characteristics of the patients with CM and UM are summarized in Table 1.

Ophthalmic Characteristics

During the period of administration of the study medication, 6 patients with CM (20%) and 2 patients with UM (40%) reported visual symptoms. Five patients developed symptoms of blurred vision, which resolved within a few hours in 3 patients and within 1 week in the other 2 patients. In 1 patient, these symptoms recurred after 2 weeks. Dark flecks in the visual field were noticed in 2 patients after taking the study medication. One patient developed persistent binocular vertical diplopia 26 days after the start of the study medication. Ophthalmic characteristics are summarized in Table 1.
Subretinal fluid as seen on OCT developed in 22 binocular patients, 3 monocular patients, and 1 functionally monocular patient (a total of 48 eyes); the (converted) median ETDRS BCVA at screening was 86 letters (range, 72e97 letters). At final follow-up, the median ETDRS BCVA recorded in 42 eyes was 87 letters (range, 65e99 letters). The difference in ETDRS BCVA between final follow-up and baseline was not statistically significant (P ¼ 0.586).
Ophthalmoscopy revealed transparent to yellowish vitelliform fundus lesions consisting of shallow SRF that developed in 27 of 35 eyes of 18 patients with CM (77%) and in 3 of 5 eyes of 5 patients with UM (60%), from whom ophthalmoscopy images were available (Figs 1A and D, 2A and D). Patients showed single or multifocal lesions (Table 1). Residual mild granular RPE changes were seen after the disappearance of the lesions with SRF in 7 of 15 follow-up cases (47%).
Subretinal fluid was detected in 23 of 30 patients with CM (77%) and 3 of 5 patients with UM (60%) (Fig 1A and D).
Subretinal fluid accumulation was bilateral in all patients with CM. Also, the onset of SRF accumulation was virtually identical in both eyes of all 22 binocular patients. On OCT, lesions were detected between a few hours and 26 days (median, 14 days) after the start of binimetinib. In 22 of 26 patients with CM/UM (85%) who showed SRF on OCT, the SRF affected the fovea. In all 18 eyes of 10 patients of a total of 40 eyes of 22 patients (45%) with subfoveal serous SRF on OCT, the aspect of lesions was dome-shaped (Fig 1C and F). On infrared reflectance imaging, these lesions showed a hyperreflective center surrounded by a hyporeflective zone, with a median size of 0.50 optic disc diameter (mean, 0.63; range, 0.33e1.5 disc diameter) (Fig 1B and E). In all 22 eyes of the other 12 patients (55%) with subfoveal SRF on OCT, smaller, round, and relatively welldemarcated foveal lesions were present, which corresponded to a shallow neuroretinal detachment on OCT with a hyperreflective subretinal accumulation (Fig 2I and N).
The median ETDRS BCVA at the time point of most pronounced subfoveal SRF accumulation on OCT was 85 letters (range, 64e97 letters). No statistically significant difference (P ¼ 0.572) was found between ETDRS BCVA at the time point of most pronounced SRF on OCT and at screening.
Subretinal fluid was also detected extrafoveally on OCT in 22 of 26 patients (85%). Four of 22 patients (18%) showed only extrafoveal lesions (Fig 1G and H). The median total number of extrafoveal lesions counted in the posterior pole of 21 patients with CM and UM, based on a 30 infrared reflectance photograph, was 7 (mean, 8.33; range, 0e27 lesions; 2 eyes without extrafoveal lesions) in 39 eyes.
Complete resolution of SRF on OCT within 2 weeks after detection of SRF was observed in 7 of 23 patients (30%) who received ophthalmological follow-up. In 4 of these patients, SRF disappeared spontaneously; in 1 patient, the SRF disappeared after binimetinib dose reduction; and in 2 patients, the study medication was stopped 3 and 6 days before complete resolution of SRF. After the first detection of SRF, 6 of 14 patients (43%) showed a complete resolution at 2 months of follow-up. In all these patients, the resolution occurred spontaneously. Persistent subfoveal or extrafoveal SRF for at least 4 months after the first detection of SRF on OCT was seen in 2 of 5 patients who had sufficient ophthalmological follow-up. None of these patients had visual symptoms. Mild intraretinal fluid accumulation was present on OCT in 1 patient with CM (Fig 2AeC) and 1 patient with UM; the intraretinal fluid had disappeared 12 and 7 days after the first detection, respectively.
In 4 of 26 patients (15%), including 3 with CM and 1 with UM, recurrence of SRF was seen after initial resolution. Subretinal fluid reappeared 9 days (range, 3e16 days) after the restart of treatment. Despite the continuation of medication, this SRF showed spontaneous resolution in all patients. One patient with a recurrence had symptoms of blurred vision, which had also occurred at the first detection of SRF. In 3 patients with recurrence of SRF, the location of the SRF was almost similar to that in the previous episode. In 1 patient with initial bilateral subfoveal SRF, for whom the binimetinib was temporarily stopped, unilateral extrafoveal SRF with subfoveal SRF occurred within 1 week after restarting binimetinib (Fig 2IeU).
The FAF imaging was performed during ophthalmological follow-up in 16 patients, 13 of whom had serous SRF, and abnormalities varied from a mild decrease of FAF to slightly increased FAF (Table 1).
One patient with serous foveal and extrafoveal lesions underwent multiple FA examinations during the course of SRF accumulation and after resolution of SRF. Lesions showed moderate hypofluorescence in the late phase of the angiogram (Fig 2DeH).
Eight patients underwent an EOG before the start of the binimetinib therapy, which showed a normal Arden ratio (>1.8) in 14 eyes of 7 patients, whereas 1 patient had an abnormal Arden ratio (<1.65) in both eyes without any ophthalmological abnormalities, revealing a pathologically decreased light rise. During treatment, a subsequent EOG showed a subnormal Arden ratio (1.65e1.8) in 2 eyes of 2 patients (25%), an abnormal ratio in both eyes of 1 patient (25%), and a normal ratio in the remaining 4 eyes (50%), although 3 of these eyes had developed SRF previously or at the time of testing. The normal Arden ratios were found in 1 eye of a patient who did not develop SRF on OCT and in 3 eyes of 2 patients who did develop SRF.
In19eyesof11cases,anEOGwasperformedonlyafterthestartof the binimetinib medication, and 16 of these eyes (84%) had developed SRF. Fifteen of these eyes (94%) with SRF showed an abnormal Arden ratio, and 1 eye (6%) showed a subnormal Arden ratio.
Color vision testing in 17 patients, including 9 patients with subfoveal SRF on OCT at the time of color vision testing, showed no significant changes in color vision during the time of follow-up. In 22 patients, including 12 patients with subfoveal SRF, central visual field analysis was within normal limits over several treatment cycles. In 1 monocular patient with UM, microperimetry using the MAIA microperimeter was performed when SRF was already decreasing and the patient had no visual symptoms. Microperimetry revealed a mildly decreased retinal sensitivity (Fig 3).
During the period of the administration of study medication, 1 patient with UM and 1 patient with CM developed small retinal hemorrhages without signs of retinal vascular occlusion. One patient with CM developed papillary edema in both eyes.

Correlation between Ophthalmological Findings and Oncologic Response to Binimetinib Treatment in Patients with Cutaneous Melanoma

In the 30 patients with CM, the median progression-free survival was 77 days (range, 20e898 days) and the median overall survival was 189.5 days (range, 24e919 days). The median overall survival was 193 days (range, 24e861 days) in 23 patients who developed SRF, compared with 167 days (range, 63e919 days) in 7 patients without SRF. This finding was not statistically significant (P ¼ 0.595).

Autoantibody Analysis

Evidence of anti-RPE autoantibodies was found in 3 patients with CM and 3 patients with UM, and anti-retinal antibodies were detected in 3 patients with CM and 2 patients with UM. Multiple anti-retinal autoantibodies and anti-RPE autoantibodies could be detected in 1 patient. Anti-bestrophin antibodies were present in 1 patient with CM and 2 patients with UM (Table 2).

Discussion

In this study, serous retinopathy developed in 77% of 30 patients with CM and in 60% of 5 patients with UM treated with the MEK inhibitor binimetinib and affected the fovea in 85% of these cases. However, visual symptoms developed in only 23% of the patients. Most patients developed symptoms of blurred vision or dark flecks in the visual field within a few hours to 3 weeks after starting the study medication, and these symptoms generally disappeared within 1 week. In smaller round to oval lesions and larger, more indistinct and irregular lesions on IRR imaging (G), occurring 16 days after the start of binimetinib in a 64year-old patient with metastatic CM. Lesions were present only outside the fovea. On OCT (scanning plane depicted with light-green line in G), the smaller round/oval lesions corresponded to a dome-shaped accumulation of SRF (H, black arrow), and the larger irregular lesions corresponded to a shallow neuroretinal detachment (H, white arrow). The space between the small and shallow neuroretinal detachments and retinal pigment epithelium was filled mainly with hyperreflective material on OCT, presumably correspondingtoaccumulationofshedphotoreceptorouter segmentdebris.IandJ, Multiple lesions on IRR imaging (I) and the corresponding fundus autofluorescence (FAF) image (J) in binimetinib-associated serous retinopathy. On FAF, the lesions corresponded to mildly decreased autofluorescence, in some lesions with a border of faintly increased autofluorescence. some patients, visual symptoms recurred within 1 week after restarting the binimetinib. However, the serous SRF collections in general seemed to have little visual impact: The difference between visual acuity at the time of most pronounced SRF and at screening was insignificant. At the end of study follow-up, there was no significant change in visual acuity compared to baseline pretreatment values.
In binocular patients, lesions were always bilateral and fairly symmetric. On ophthalmoscopy, foveal and extrafoveal lesions could have a vitelliform aspect corresponding to domeshaped serous neuroretinal detachments, but some patients had more extensive, irregular, shallow neuroretinal detachments. Multifocal areas of SRF occurred frequently outside the fovea, which is an indication of a more extensive retinal/RPE dysfunction. We did not find a clear correlation between the presence or absence of serous retinopathy and the progressionfree or overall survival of the patients with metastatic CM.
Figure 2. Multimodal imaging illustrating the clinical course of serous retinopathy associated with binimetinib treatment. AeC, Multifocal, variably sized serous neuroretinal detachments on red-free fundus photography (A) and infrared reflectance (IRR) imaging (B) in a 64-year-old patient with metastatic cutaneous melanoma (CM) treated with binimetinib. Bilateral mild intraretinal fluid accumulation and a dome-shaped serous neuroretinal detachment with subretinal fluid (SRF) accumulation were seen on optical coherence tomography (OCT) (C). The intraretinal fluid appeared 20 days after the start of binimetinib and disappeared in 12 days. The binimetinib dose was tapered 7 days after appearance of the intraretinal fluid. Resolution of the intraretinal fluid occurred faster than the resolution of the subfoveal SRF, which was still present at ophthalmological follow-up 11 weeks later. D and H, A 65-year-old patient with metastatic CM with yellowish, foveal, and small, round, and oval vitelliform extrafoveal lesions on fundus photography (D), corresponding to hyporeflective lesions with a moderately hyperreflective center on IRR imaging (E). On OCT (light-green line in E indicating scanning plane), the foveal lesion corresponded to a focal serous neuroretinal detachment (F). On fluorescein angiography (FA), lesions showed hardly any fluorescence changes in the early phase (G), whereas the smaller temporal lesions showed modest fluorescein staining in the late phase of the angiogram, and the foveal lesions remained mostly hypofluorescent (H). No FA abnormalities were present after resolution of the serous SRF. IeU, Evolution of binimetinib-associated serous retinopathy in a 45-year-old patient with metastatic CM. At the baseline examination before the start of binimetinib treatment, the OCT scan showed an entirely normal macular structure (M). Within 1 day after the start of binimetinib treatment (45 mg twice daily), the patient developed bilateral small, round, and relatively welldemarcated foveal lesions without evidence of extrafoveal lesions (I). These lesions corresponded to a shallow neuroretinal detachment on OCT (N). Binimetinib was stopped to assess the effect of treatment discontinuation on the retinopathy. Five days after the discontinuation of binimetinib, SRF resolution was seen on OCT (O). Study medication was restarted at a lower dose (30 mg twice daily), and a shallow unilateral serous neuroretinal detachment reoccurred between 3 (P) and 4 days (Q) after the restart. In addition, small, round, and oval extrafoveal lesions (J), which were dome-shaped on OCT (K), gradually developed in these 5 days after restarting binimetinib. Fundus autofluorescence of the foveal lesion showed no obvious abnormalities in the fovea and only minor fundus autofluorescence (FAF) changes corresponding to the focal lesions superior in the macula (L). Despite the fact that binimetinib treatment was continued after this reoccurrence of lesions, the foveal SRF resolved 2 days after the restart of binimetinib (R), and this resolution of the lesion persisted 6 (S), 21 (T), and 43 days later (U). Resolution of the extrafoveal lesions occurred between 6 and 21 days later.
Previous studies have described the aforementioned retinal lesions associated with MEK inhibition treatment as central serous-like chorioretinopathy,12 an eye condition resembling central serous chorioretinopathy,14 bilateral subfoveal neurosensory retinal detachments,20 and MEK inhibitoreassociated retinopathy.21 In our study, there was no evidence of choroidal abnormalities or leakage from focal RPE defects in the patients with serous SRF. Instead, we found that the most striking abnormality besides the usually transient serous retinopathy was the profoundly abnormal EOG. This indicates that the primary abnormality is a panretinal dysfunction of the RPE, which can lead to a serous retinopathy supposedly due to failure of the RPE pump. Previous studies and OCT, FA, and FAF findings as in our study did not show visible structural RPE changes or “hot spots” of leakage that would have been suggestive of central serous chorioretinopathy.20,21 Therefore, we suggest describing the phenotype as serous retinopathy. Disappearance of SRF was observed in 43% of patients within 8 weeks after the start of binimetinib treatment, despite continuation of the study medication. A possible reason for this disappearance could be a gradual reduced sensitivity of the RPE to medication. However, the EOG remained abnormal despite SRF resolution, indicating prolonged RPE dysfunction despite SRF resolution.
The pathogenesis of serous retinopathy associated with MEK inhibitor treatment is unclear. Serum samples of all 3 patients with CM and 3 patients with UM tested for autoantibodies showed various anti-retinal and anti-RPE antibodies. Data on the possible prevalence of these autoantibodies in individuals are scarce.22 In one study, anti-retinal autoantibodies were found in 16% of 58 healthy subjects, which was significantly lower compared with patients with macular telangiectasia type 2.23 Autoantibodies against bestrophin were present in 3 of 6 patients. These autoantibodies were also reported earlier in a patient with choroidal malignant melanoma in the left eye and vitelliform lesions in the right eye.24 It has been proposed that dysfunction of bestrophin results in abnormal fluid and ion transport by the RPE.25 It is possible that treatment with binimetinib triggers (tumor) cells to stimulate the generation or release of antibodies that could play a role in the pathogenesis of serous retinopathy. Moreover, an autoantibody attack against certain RPE epitopes could result in compromised RPE ion homeostasis, resulting in an abnormal EOG (which reflects changes in the standing potential of the entire RPE), as well as a disturbed RPE pumping function of fluid from the subretinal space, resulting in serous SRF accumulation. A mechanism of anti-RPE antibodies against the RPE protein bestrophin, which is also affected in Best vitelliform macular dystrophy as a result of BEST1 gene mutations, has been described in a patient with metastatic choroidal melanoma with vitelliform paraneoplastic retinopathy who also had a pathologically reduced Arden ratio on the EOG.24,25 In a patient with metastatic melanoma with acute exudative polymorphous vitelliform maculopathy, which may resemble cases with the more extensive multifocal serous retinopathy in our study, anti-RPE antibodies to peroxiredoxin 3 (26 kDa) were found in combination with an abnormal EOG. The autoantibodies became undetectable after the exudation had resolved.26
Subretinal fluid was present on OCT in the 1 patient with CM and 2 patients with UM who showed the highest number of different antibodies. Multiple antibodies could have a higher pathogenic potential than a single antibody by enhancing pathology. However, only 1 of the 6 patients tested for autoantibodies had visual symptoms. In patients with cancer-associated retinopathy (CAR), the presence of anti-a-enolase autoantibodies has been described.27 These anti-a-enolase autoantibodies were found in only 1 patient in our study, who did not have visual symptoms. In our study, neither a clear relationship between the presence of anti-bestrophin antibodies and EOG nor a consistent pattern of specific autoantibodies in the tested patients was found. Therefore, the identification of a broad spectrum of anti-retinal and anti-RPE antibodies does not necessarily mean that binimetinib-associated serous retinopathy is primarily caused by this autoantibody attack, despite the fact that the presence of these autoantibodies is an indication of autoimmune response.
Bearing this in mind, we hypothesize that direct toxicity of MEK inhibitor treatment to the RPE also can be an important factor in the pathogenesis of binimetinibassociated serous retinopathy and RPE dysfunction. The potentially rapid onset of visual symptoms and serous retinopathy would support direct RPE toxicity as an important contributing factor.
Analysis of anti-retinal, antieretinal pigment epithelium (RPE), and anti-bestrophin antibodies in 3 patients with UM (1 female, 2 male) and 3 patients with CM (3 male). Of these patients, 2 with CM and 2 with UM showed subretinal fluid on optical coherence tomography and 2 with CM and 3 with UM showed abnormal Arden ratios on electro-oculography (EOG) at ophthalmological examination during the study. Anti-retinal antibodies were detected in 3 patients with CM and 2 patients with UM. None of these patients had dark adaptation difficulties or night blindness, and full-field electroretinography was performed in 3 of these patients. Anti-RPE antibodies were detected in all 6 patients. Anti-bestrophin antibodies were detected in 1 patient with CM and 2 patients with UM. These anti-bestrophin autoantibodies were analyzed because they were detected previously in a patient with choroidal malignant melanoma in the left eye and a vitelliform macular lesion in the right eye with a markedly abnormal EOG.24 In 2 patients in whom anti-bestrophin antibodies were detected and in 3 patients in whom antibodies were not detected, EOG did show an abnormal Arden ratio. In 1 patient with anti-bestrophin antibodies, EOG showed a normal Arden ratio in 1 eye and a subnormal Arden ratio in the other eye.
Serous retinopathy associated with MEK inhibitor treatment can be confused with several other clinical entities. Our study indicates that the clinical picture of the retinopathy can be discerned from central serous chorioretinopathy on the basis of the occurrence of symptoms within days after administration of the medication, the absence of associated RPE detachments, the absence of “hot spots” on FA, and an abnormal EOG.20,2 Melanoma-associated retinopathy (MAR) occurs most frequently in patients with metastasized CM. Patients with MAR characteristically have a sudden start of photopsia and may experience night blindness, scotomas, and visual field loss, with rod function being primarily affected. 8 In CAR, both rod and cone function are affected. An association with autoantibodies against retinal bipolar cells has been found both in MAR and CAR,29 as well as autoantibodies against a 22-kDa neuronal antigen, a 34-kDa protein in Müller cells, transducin-b, mitofilin, and titin in cases of MAR. 0,31 Paraneoplastic vitelliform retinopathy resembling Best vitelliform macular dystrophy, acute exudative polymorphous vitelliform maculopathy, and adult-onset foveomacular vitelliform macular dystrophy have been described in patients with metastatic CM and UM.32e37 In Best vitelliform macular dystrophy, a disease caused by autosomal dominantly inherited BEST1 gene mutations that lead to an aberrant bestrophin protein in the RPE, a vitelliform lesion similar to a central serous retinal detachment is described.2 ,3 An abnormal EOG and a normal full-field ERG are typically found in Best vitelliform macular dystrophy. In autosomal recessive bestrophinopathy, also caused by BEST1 gene mutations, the full-field ERG also can be abnormal in addition to the abnormal EOG.3
In summary, we show that the MEK inhibitor binimetinib is associated with a high incidence of serous retinopathy due to dysfunction of the RPE. Possible pathogenetic mechanisms include anti-RPE and anti-retinal autoantibodies or direct RPE toxicity of MEKinhibitors, which is a subject for furthers tudies. Binimetinib-associated serous retinopathy was time-dependent and reversible despite continuation of the study medication. Ophthalmological Sotrastaurin monitoring is warranted in patients using MEK inhibition treatment, but discontinuation of binimetinib administrationgenerallydoesnotseemnecessarybecauseofthe relativelylowvisualimpactandtransientnatureoftheassociated serous retinopathy. Future studies should assess whether other MEK inhibitors besides binimetinib are also associated with the development of a similar clinical picture.

References

10. Chapman PB, Hauschild A, Robert C, et al. Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med 2011;364:2507–16.
11. Flaherty KT, Robert C, Hersey P, et al. Improved survival with MEK inhibition in BRAF-mutated melanoma. N Engl J Med 2012;367:107–14.
12. Ascierto PA, Schadendorf D, Berking C, et al. MEK162 for patients with advanced melanoma harbouring NRAS or Val600 BRAF mutations: a non-randomised, open-label phase 2 study. Lancet Oncol 2013;14:249–56.
13. van Herpen CM, Agarwala SS, Hauschild A, et al. Overall survival and biomarker results from a phase 2 study of MEK1/ 2 inhibitor binimetinib (MEK162) in patients with advanced NRAS-mutant melanoma. Ann Oncol 2014;25:v1–41.
14. Flaherty KT, Infante JR, Daud A, et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med 2012;367:1694–703.
15. Gregori NZ, Feuer W, Rosenfeld PJ. Novel method for analyzing Snellen visual acuity measurements. Retina 2010;30:1046–50.
16. Marmor MF, Fulton AB, Holder GE, et al. ISCEV Standard for full-field clinical electroretinography (2008 update). Doc Ophthalmol 2009;118:69–77.
17. Marmor MF, Brigell MG, McCulloch DL, et al. ISCEV standard for clinical electro-oculography (2010 update). Doc Ophthalmol 2011;122:1–7.
18. Hood DC, Bach M, Brigell M, et al. ISCEV standard for clinical multifocal electroretinography (mfERG) (2011 edition). Doc Ophthalmol 2012;124:1–13.
19. US Department of Health and Human Services. Common Terminology Criteria for Adverse Events (CTCAE) version 4. 03. 2010; US Government Publishing Office, Washington, DC.
20. McCannel TA, Chmielowski B, Finn RS, et al. Bilateral subfoveal neurosensory retinal detachment associated with MEK inhibitor use for metastatic cancer. JAMA Ophthalmol 2014;132:1005–9.
21. Urner-Bloch U, Urner M, Stieger P, et al. Transient MEK inhibitor-associated retinopathy in metastatic melanoma. Ann Oncol 2014;25:1437–41.
22. Adamus G. Autoantibody targets and their cancer relationship in the pathogenicity of paraneoplastic retinopathy. Autoimmun Rev 2009;8:410–4.
23. Zhu L, Shen W, Zhu M, et al. Anti-retinal antibodies in patients with macular telangiectasia type 2. Invest Ophthalmol Vis Sci 2013;54:5675–83.
24. Eksandh L, Adamus G, Mosgrove L, Andreasson S. Autoantibodies against bestrophin in a patient with vitelliform Footnotes and Financial Disclosures paraneoplastic retinopathy and a metastatic choroidal malignant melanoma. Arch Ophthalmol 2008;126:432–5.
25. Boon CJ, Klevering BJ, Leroy BP, et al. The spectrum of ocular phenotypes caused by mutations in the BEST1 gene. Prog Retin Eye Res 2009;28:187–205.
26. Koreen L, He SX, Johnson MW, et al. Anti-retinal pigment epithelium antibodies in acute exudative polymorphous vitelliform maculopathy: a new hypothesis about disease pathogenesis. Arch Ophthalmol 2011;129:23–9.
27. Adamus G, Aptsiauri N, Guy J, et al. The occurrence of serum autoantibodies against enolase in cancereassociated retinopathy. Clin Immunol Immunopathol 1996;78:120–9.
28. Heckenlively JR, Ferreyra HA. Autoimmune retinopathy: a review and summary. Semin Immunopathol 2008;30:127–34.
29. Milam AH, Saari JC, Jacobson SG, et al. Autoantibodies against retinal bipolar cells in cutaneous melanoma-associated retinopathy. Invest Ophthalmol Vis Sci 1993;34:91–100.
30. Potter MJ, Adamus G, Szabo SM, et al. Autoantibodies to transducin in a patient with melanoma-associated retinopathy. Am J Ophthalmol 2002;134:128–30.
31. Pfohler C, Preuss KD, Tilgen W, et al. Mitofilin and titin as target antigens in melanoma-associated retinopathy. Int J Cancer 2007;120:788–95.
32. Sotodeh M, Paridaens D, Keunen J, et al. Paraneoplastic vitelliform retinopathy associated with cutaneous or uveal melanoma and metastases. Klin Monbl Augenheilkd 2005;222:910–4.
33. Zacks DN, Pinnolis MK, Berson EL, Gragoudas ES. Melanoma-associated retinopathy and recurrent exudative retinal detachments in a patient with choroidal melanoma. Am J Ophthalmol 2001;132:578–81.
34. Palmowski AM, Haus AH, Pfohler C, et al. Bilateral multifocal chorioretinopathy in a woman with cutaneous malignant melanoma. Arch Ophthalmol 2002;120:1756–61.
35. Jampol LM, Kim HH, Bryar PJ, et al. Multiple serous retinal detachments and subretinal deposits as the presenting signs of metastatic melanoma. Retina 2004;24:320–2.
36. Chan CK, Gass JD, Lin SG. Acute exudative polymorphous vitelliform maculopathy syndrome. Retina 2003;23:453–62.
37. Nieuwendijk TJ, Hooymans JM. Paraneoplastic vitelliform retinopathy associated with metastatic choroidal melanoma. Eye (Lond) 2007;21:1436–7.
38. Boon CJ, Klevering BJ, den Hollander AI, et al. Clinical and genetic heterogeneity in multifocal vitelliform dystrophy. Arch Ophthalmol 2007;125:1100–6.
39. Boon CJ, van den Born LI, Visser L, et al. Autosomal recessive bestrophinopathy: differential diagnosis and treatment options. Ophthalmology 2013;120:809–20.