A, Nongranulomatous anterior uveitis 3 weeks after vemurafenib treatment onset. In the left eye, pigment deposits are seen on the anterior capsule at the site of synechiae, which were subsequently broken. B, Vitritis (anterior chamber cells [4+]) in the left eye; optic disc edema is seen through the dense vitreous haze.
Nongranulomatous anterior uveitis in the left eye 7 months after onset of vemurafenib treatment. The intense inflammation resulted in posterior synechiae, some of which remained despite the mydriatic and corticosteroid local treatment.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Guedj M, Quéant A, Funck-Brentano E, et al. Uveitis in Patients With Late-Stage Cutaneous Melanoma Treated With Vemurafenib. JAMA Ophthalmol. 2014;132(12):1421–1425. doi:10.1001/jamaophthalmol.2014.3024
Copyright 2014 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
This case series highlights the risk of uveitis in patients treated with vemurafenib for unresectable or metastatic cutaneous melanoma.
To assess the occurrence and severity of uveitis as an adverse effect of vemurafenib therapy.
Design, Setting, and Patients
In this observational small case series, data were collected successively from May 1, 2012, through February 31, 2013, from patients with clinical signs of ocular inflammation treated with vemurafenib at the Department of Ophthalmology, Cochin-Hôtel-Dieu Hospital.
Main Outcomes and Measures
Patients’ demographics, vemurafenib dosages, and the intervals between the onset of treatment and the first ocular symptoms were recorded. The characteristics of ocular inflammatory manifestations were analyzed. The effect of the discontinuation of vemurafenib therapy on ocular manifestations was assessed, as well as the effect of rechallenging when vemurafenib was reintroduced.
Seven patients (mean [SD] age, 74.7 [4.0] years) had uveitis. The vemurafenib dose was 960 mg twice per day in 6 patients and a half dose in 1 patient. The mean (SD) time until the appearance of ocular signs was 5.6 (2.3) months (range, 19 days to 7 months), and inflammation ranged from mild or low-grade anterior uveitis to severe explosive panuveitis complicated by retinal detachment. Signs of ocular inflammation were always bilateral. Optical coherence tomography revealed a macular edema in only 1 of the 7 patients. Clinical improvement occurred when vemurafenib therapy was stopped in 5 of 7 patients. The rechallenge at treatment reintroduction was positive in 2 of 7 patients.
Conclusions and Relevance
This small case series highlights that uveitis can be a noteworthy adverse effect of vemurafenib therapy in patients with metastatic cutaneous melanoma. However, these cases of uveitis were usually restricted to the anterior segment and manageable with local corticosteroid therapy, which justified the continuation of vemurafenib therapy because the benefits regarding the patients’ survival were greater than the risk to their vision.
Vemurafenib is a BRAF (OMIM 164757) enzyme inhibitor developed for the treatment of late-stage melanoma with the V600E BRAF mutation, which is thought to be present in half of all melanomas1 and responsible for the proliferation of tumoral cells. The name vemurafenib derives from V600E mutated BRAF inhibition.
Vemurafenib received approval from the Food and Drug Administration for the treatment of late-stage melanoma in August 2011 and from the European Medicines Agency in February 2012 as monotherapy for the treatment of adults with the V600 BRAF mutation and nonresectable or metastatic melanoma. In clinical trials, vemurafenib treatment was reported to improve survival in previously untreated and treated patients with advanced melanoma with BRAF V600 mutations.2,3 The most common grade 3 or higher adverse events in patients receiving vemurafenib (at the maximum tolerated dose of 960 mg twice a day) compared with those receiving conventional chemotherapy were cutaneous squamous cell carcinoma (31%) (which could generally be surgically removed), rashes (52%), liver function abnormalities, arthralgia (58%), and photosensitivity (52%).
To our knowledge, no adverse effects, such as uveitis or intraocular inflammation, have been reported so far in the medical literature. However, in phase 1 and 2 clinical trials, 5 cases of uveitis have been reported.4 The purpose of this study was to describe the detailed ocular features of patients with uveitis related to vemurafenib therapy.
This was an observational study conducted within the field of approved indications for vemurafenib in France. Informed consent and institutional review board approval were not required in that context. We collected data from patients who were all followed up for metastatic cutaneous melanoma and treated with vemurafenib and who were sent to the Department of Ophthalmology, Cochin Hospital, from May 1, 2012, through February 31, 2013. Vemurafenib-treated patients from 2 departments of dermatology were referred to us in a systematic manner because of ocular symptoms (pain, redness, vision loss, blurred vision, and floaters) but without systematic-specific screening. Patient demographics, vemurafenib dosages, and the interval between treatment onset and first ocular symptoms were recorded. The results of tests used to diagnose other causes of uveitis were reviewed. Optical coherence tomography (OCT) of the macula was performed in a systematic manner. The effect of the discontinuation of vemurafenib therapy on ocular manifestations was assessed, as well as the effect of rechallenging in cases of vemurafenib reintroduction.
Among 78 vemurafenib-treated patients, 8 were referred for ocular symptoms, which led to the diagnosis of uveitis in 7 patients (1 man and 6 women) and of episcleritis in 1 patient. The mean (SD) age at the onset of uveitis was 74.7 (4.0) years (range, 69-81 years). The vemurafenib dosage at onset was 960 mg twice per day for all patients. For 1 patient, it was initially 480 mg twice per day, then 960 mg twice per day. The mean (SD) time between the treatment onset and the first ocular symptoms was 5.6 (2.3) months (range, 19 days to 7 months).
Ocular findings are summarized in the Table. The intraocular inflammation was characterized by low-grade or mild anterior uveitis, except in 1 patient with severe acute onset panuveitis, which was further complicated by retinal detachment and complete visual loss.
Five patients discontinued treatment; in 2 patients, anterior uveitis subsided despite continued vemurafenib treatment. Clinical improvement occurred when vemurafenib treatment was discontinued in 5 of the 7 patients. On rechallenging with vemurafenib (4 patients), an effect of treatment reintroduction was noted in 2 patients, who relapsed after the rechallenge (Table 2).
Patient 1, a 74-year-old woman who had been treated with vemurafenib, 960 mg twice a day, since June 2012 for a metastatic cutaneous melanoma of the leg, was seen by us in January 2013 (after a 7-month treatment period) and reported having had floaters for a few weeks. Her visual acuity was 20/20 in both eyes. The slitlamp examination revealed a bilateral nongranulomatous anterior uveitis, with thin retrocorneal precipitates, posterior synechiae, anterior chamber (AC) cells (1+). The fundus examination revealed no vitreous haze or vasculitis, and no macular edema was found on OCT. The results of diagnostic tests adjusted to the clinical presentation (HLA-B27 typing, angiotensin-converting enzyme, Treponema pallidum hemagglutination VDRL test, C-reactive protein, blood cell count, QuantiFERON tuberculosis) were negative. Inflammatory signs (AC cells) disappeared completely after discontinuation of vemurafenib treatment and with local corticosteroid therapy, which was tapered in 4 weeks.
Patient 2, a 78-year-old woman who had been treated with vemurafenib at half and then full dose (960 mg twice a day) since June 2012 for a metastatic cutaneous melanoma of the leg, was referred to the Department of Ophthalmology, Cochin Hospital, in November 2012 (after a 5-month treatment period). Her visual acuity was 20/20 OD and 20/40 OS. A bilateral anterior nongranulomatous uveitis was diagnosed, and vemurafenib therapy was discontinued for 2 weeks. As the symptoms improved, vemurafenib was reintroduced 15 days later at a 75% dose. In January 2013, a relapse of the anterior bilateral uveitis occurred, with more than half of the AC cells, no synechiae but pigments on the anterior lens capsule, and anterior vitritis (AC cells [1+]). The blood test results were negative. A favorable response to topical corticosteroid therapy was observed.
Patient 3, a 81-year-old man who had been treated with vemurafenib at full dose (960 mg twice a day) then half dose since May 2012 for a metastatic cutaneous melanoma of the elbow, was referred to the Department of Ophthalmology, Cochin Hospital, in December 2012 (after a 7-month treatment period) for a mild bilateral blurred vision. Her visual acuity was 20/40 OD and 20/28 OS. The slitlamp examination revealed a bilateral anterior nongranulomatous uveitis, with posterior synechiae, slightly elevated flare meter measures at 30 photons per millisecond (N<8) without visible AC cells, and a mild cataract. The fundus examination findings were normal, and no macular edema was revealed on OCT. Blood test results were negative. A favorable response to topical corticosteroid therapy was observed.
Patient 4, a 75-year-old woman who had been treated with vemurafenib, 960 mg twice a day, since January 2012 for a metastatic cutaneous melanoma of the back, reported acute pain and vision loss in her left eye at her first visit in the Department of Ophthalmology, Cochin Hospital, in August 2012 (after a 7-month treatment period). Her visual acuity was 20/33 OD and 20/200 OS. The slitlamp examination revealed thin retrocorneal precipitates in both eyes, AC cells (1-2+), and synechiae on 360° with pupillary seclusion in the left eye. Flare meter values were increased to 30 photons per millisecond in the right eye and to 200 photons per millisecond in the left eye. The fundus examination findings were normal in the right eye and could not be seen in the left eye, but the ocular echography revealed no retinal damage. Blood test results were negative. Vemurafenib treatment was immediately discontinued, and topical corticosteroid therapy was initiated. As the clinical examination findings improved, vemurafenib was reintroduced in October 2012, at 75% and then full dose after 6 weeks. Two months later, the reappearance of inflammatory ocular signs (pain, redness, and vision loss in the right eye) led to the diagnosis of a relapse of an anterior and intermediate uveitis of the right eye complicated by macular edema. Visual acuity was 20/50 OD, and thin retrocorneal precipitates, posterior synechiae, and vitreous haze (AC cells [2+]) were observed, as well as cystoid macular edema on OCT with a central macular thickness of 500 µm. These symptoms led for the second time to the discontinuation of vemurafenib treatment.
Patient 5, a 69-year-old woman who had been treated with vemurafenib, 960 mg twice a day, since April 2012 for a metastatic cutaneous melanoma of the ear, was referred to the Department of Ophthalmology, Cochin Hospital, in May 2012 (after a 3-week treatment period) for severe bilateral vision loss with pain and redness. Slitlamp examination revealed a bilateral nongranulomatous acute anterior uveitis, with posterior synechiae and AC cells (3+) (Figure 1A). A topical corticosteroid therapy (1 drop per hour) was initiated, and vemurafenib treatment was stopped immediately. However, despite the treatment discontinuation, a worsening of vision loss and intraocular inflammation was observed when the patient was referred once again to the Department of Ophthalmology, Cochin Hospital, in June 2012. Her visual acuity was then restricted to hand movements in both eyes. The fundus examination revealed bilateral vitritis, which was severe in the left eye (AC cells [4+]) with optic disc edema (Figure 1B). An aqueous tap was performed to rule out an endogenous endophthalmitis. Three intravenous methylprednisolone pulses were administered but did not lead to an improvement of the vitreous haze. The vitritis became complicated with a retinal detachment in the right eye. Consequently, a vitrectomy with silicone oil tamponade and endolaser was performed, but visual acuity remained restricted to the vision of hand movements. Vemurafenib treatment was not reintroduced.
Patient 6, a 71-year-old woman who had been treated with vemurafenib, 960 mg twice a day, since August 2012 for a metastatic cutaneous melanoma, was referred to the Department of Ophthalmology, Cochin Hospital, in February 2013 (after a 7-month treatment period) for pain, redness, and vision loss in the left eye. Visual acuity was 20/28 OD and 20/40 OS, and slitlamp examination revealed bilateral retrocorneal precipitates and AC inflammation (AC cells in the right eye [1+] and in the left eye [3+] with flare meter values at 41 photons per millisecond in the right eye and 259 photons per millisecond in the left eye) (Figure 2). The fundus examination findings were normal except for a mild anterior vitritis by contiguity in the left eye. Local evolution was satisfactory after the cessation of vemurafenib and topical corticosteroid therapy.
Patient 7, a 75-year-old woman who had been treated with vemurafenib, 960 mg twice a day, since June 2012 for a metastatic cutaneous melanoma, was referred to the Department of Ophthalmology, Cochin Hospital, for bilateral floaters that occurred 4 months after the onset of her treatment. Her visual acuity was 20/20 in both eyes; slitlamp examination revealed bilateral retrocorneal precipitates, AC cells (1+) in the left eye, and laser flare meter values of 3 photons per millisecond in the right eye and 17 photons per millisecond in the left eye, with synechiae in both eyes. The fundus examination findings were normal. Quiescent eyes were observed after topical corticosteroid therapy, and vemurafenib treatment was continued.
To our knowledge, uveitis has not been previously reported as an adverse effect of treatment with vemurafenib. A combination of chronologic and semiologic criteria is required to attribute an adverse event to a drug with a causal link.5,6 Three chronologic criteria are taken into account: the time between the observed effect and the introduction of the treatment, the evolution after cessation of the treatment, and the result of rechallenging. Semiologic criteria include 4 items: suggestive clinical signs, predisposing factors, reliable additional tests, and elimination of another possible cause. In our study, the chronologic criteria were consistent because the mean time from initiation of vemurafenib treatment until the appearance of ocular signs was 5.4 months (range, 19 days to 7 months), and the discontinuation of vemurafenib treatment coincided with a clinical improvement in 5 of the 7 patients. Moreover, the rechallenge by treatment reintroduction was positive in 2 of 6 patients (patients 2 and 4). Among the semiologic criteria, vemurafenib has been described in association with uveitis in phase 1 and 2 clinical trials. Flare meter measurements confirmed the presence of an intra-ocular inflammation in all cases, in particular when the AC cells were not observed during visits. The additional tests excluded other non–drug-related causes in 6 of the 7 patients, especially in patient 5, whose anterior chamber tap test result was negative, excluding an endogenous endophthalmitis. However, patient 6 had a previous history of psoriatic arthritis, compatible with non–vemurafenib-related uveitis. Ocular symptoms varied from low to mild anterior uveitis in 5 of the 7 patients.
Only 1 of these 7 patients (patient 5) had very severe ocular damage (severe explosive panuveitis complicated with retinal detachment), worsening even after the discontinuation of treatment. However, most of these cases of uveitis were restricted to the anterior segment (4 of 7) and responded to topical corticosteroid therapy. Thus, the permanent discontinuation of vemurafenib therapy may not be justified because the benefits of the treatment on patient survival are overall greater than the risk on their vision. However, the benefit-risk balance could be more difficult to assess in patients treated with vemurafenib as an adjuvant therapy, as in an ongoing phase 3 study. The withdrawal of vemurafenib should thereby be discussed jointly on a case-by-case basis by ophthalmologists, dermatologists, and the patients.
This small case series highlights that uveitis can be a noteworthy adverse effect of vemurafenib therapy in patients with unresectable or metastatic cutaneous melanoma. Heterogeneous presentations of the inflammation may be observed in this context. The mechanisms that lead to vemurafenib-induced uveitis remain unknown. The continuation of treatment or its discontinuation require careful balancing between the expected benefits of treatment vs its ocular adverse effects.
Submitted for Publication: September 20, 2013; final revision received March 11, 2014; accepted March 12, 2014.
Corresponding Author: Astrid Quéant, MD, Faculté de Médecine, Université Paris Descartes, 27 Rue du Faubourg Saint-Jacques, 75014 Paris, France (email@example.com).
Published Online: August 14, 2014. doi:10.1001/jamaophthalmol.2014.3024.
Author Contributions: Dr Guedj had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Guedj, Quéant, Lellouch, Gantzer, Brézin.
Acquisition, analysis, or interpretation of data: Guedj, Quéant, Funck-Brentano, Kramkimel, Monnet, Longvert, Brézin.
Drafting of the manuscript: Guedj, Quéant, Lellouch, Gantzer, Brézin.
Critical revision of the manuscript for important intellectual content: Quéant, Funck-Brentano, Kramkimel, Monnet, Longvert, Brézin.
Statistical analysis: Guedj.
Administrative, technical, or material support: Guedj, Funck-Brentano.
Study supervision: Guedj, Quéant, Funck-Brentano, Monnet, Gantzer, Brézin.
Conflict of Interest Disclosures: None reported.