Efficacy of azathioprine as a first-line therapy in new onset of acute Vogt-Koyanagi-Harada disease in the Indian population

Abstract

PURPOSE:

The purpose of this study is to evaluate the efficacy of azathioprine as a first-line therapy in new onset of acute Vogt-Koyanagi-Harada (VKH) disease in the Indian population.

METHODS:

This retrospective study included 17 (34 eyes) patients who received azathioprine along with systemic steroids for new onset of acute VKH disease and had minimum follow-up of 24 months. Uveitis activity was assessed clinically and by indocyanine green angiography.

RESULTS:

The mean follow-up period was 27.53 (range: 24–48) months. Corticosteroid-sparing effect was achieved in 14 (82.35%) patients. The median time to achieve the steroid-sparing effect was 3 months (range: 1.5–5 months). Corrected distance visual acuity better than 0.3 logMAR was achieved in 28 (82.35%) eyes. Sunset glow fundus developed in 2 (5.88%) eyes while on azathioprine therapy. None of the eyes developed chronic granulomatous uveitis phase or severe systemic side effect during treatment with azathioprine.

CONCLUSION:

Azathioprine as first-line therapy in patients with new onset of acute VKH disease has effective steroid-sparing effect and prevents progression to chronic recurrent granulomatous stage.

INTRODUCTION

The goals of management of acute Vogt-Koyanagi-Harada (VKH) disease are to suppress the intraocular inflammation in the acute phase and to prevent its progression to the chronic recurrent phase.[1,2,3,4,5,6,7,8,9,10] Despite early administration of high-dose corticosteroids in the initial acute phase followed by a gradual tapering in doses, up to 79% of patients progressed to the chronic recurrent phase, and 38% developed subretinal fibrosis.[11] Administration of long-term high dosage of corticosteroids is associated with systemic and ocular complications in 49.4% of the patients.[12] In such a situation, nonsteroidal immunomodulator therapy (IMT) plays an important role in the long-term control of choroidal inflammation with minimal side effects. There is increasing evidence supporting the fact that dual corticosteroid plus IMT should be introduced since the beginning for new onset of acute VKH disease.[13,14] In addition, the International Uveitis Study Group categorized VKH disease as an absolute indication of administering IMT.[15]

Various IMTs mainly cyclosporine, azathioprine, methotrexate, and mycophenolate mofetil have shown good clinical outcomes when used to treat VKH disease.[16,17,18,19,20] Most studies focusing on the use of IMTs in VKH disease evaluate their efficacy in the chronic recurrent phase when irreversible damage has already occurred. Second, they are comparatively expensive and may cause severe systemic side effects.[21] Azathioprine, an inhibitor of purine synthesis, has been shown to be effective for treatment in a variety of chronic noninfectious uveitis.[22,23,24] Limited studies have evaluated the efficacy of azathioprine in VKH disease while recommending it as a second line of medication for VKH disease.[7,25] In this study, we evaluated the efficacy of azathioprine as a first-line therapy in the new onset of acute VKH disease in the Indian population.

METHODS

The Institutional Ethics Committee approved the study, and it was conducted in full accord with the tenets of the Declaration of Helsinki. The medical records of patients diagnosed with acute VKH disease using the revised diagnostic criteria between January 2016 and July 2018 were reviewed.[2] We included patients with new onset of acute VKH disease who received azathioprine and systemic corticosteroids and had a follow-up history of at least 2 years. Acute VKH disease was defined as the presence of bilateral diffuse choroiditis and serous retinal detachment.[2] Data collected included age, gender, symptom profile, corrected distance visual acuity (CDVA), applanation tonometry, slit-lamp examination of the anterior segment, fundus biomicroscopy, and indirect ophthalmoscopy, treatment received, complications, and recurrences. Visual acuity was measured in logMAR value. Anterior segment inflammatory cells were graded using the Standardization of Uveitis Nomenclature guidelines.[26] Fundus fluorescein angiography (FFA) and indocyanine green angiography (ICGA) were performed in all cases, while ocular ultrasonography and optical coherence tomography (OCT) were performed in 22 and 16 eyes, respectively, before the initiation of therapy. The clinical presentation and characteristics of FFA and ICGA findings formed the basis for the diagnosis of VKH disease.[1,2,3,4,5,6,7] Revised diagnostic criteria for VKH disease were used to classify the cases into three categories: complete, incomplete, or probable VKH disease.[2]

A uniform protocol was followed to treat patients presenting with new onset of acute VKH disease. All patients were treated with high-dose systemic corticosteroids along with azathioprine. If not contraindicated, patients received pulse intravenous methylprednisolone (IVMP) (1 g/day) for three consecutive days. This was followed by oral prednisolone at a dose of 1–1.5 mg/kg/day. In case, IVMP was contraindicated, oral prednisolone was prescribed at a dose of 1.5 mg/kg/day. The prednisolone dose was gradually tapered following improvement in the intraocular inflammation to a maintenance dose of 5–10 mg/day. In general, prednisolone was tapered using the following protocol: 10 mg every 2 weeks, until a daily dose of 40 mg was achieved. Afterward, prednisone was tapered at 5 mg every 2 weeks, until 5–10 mg/day was achieved. Simultaneously, oral azathioprine of 2–3 mg/kg/day was also started and continued for at least 12 months under the guidance of an internist. Starting dose of azathioprine was 25 mg, and subsequently, the dose was increased by 25 mg every 2 weeks till the desired dose was achieved. Systemic evaluation and laboratory results, including complete blood count and liver enzymes, were carried out by an internist every 6 weeks during administration of the azathioprine therapy. The predetermined indications for withdrawal of azathioprine were leukocyte count <3500/mm,[3] platelet counts <10⁵/mm³, hemoglobin <7 g/dL, or an increase in liver enzymes to more than twice the baseline reading. In case of anterior segment inflammation, topical corticosteroids and cycloplegic eye drops were administered as part of the initial treatment regime. At 3 months of follow-up, repeat ICGA and FFA were performed in all cases to look for any disc dye leakage or active choroidal lesion suggestive of disease activity. Once disease inactivity was documented, patients were followed every 8 weeks. Thereafter, ICGA and FFA were performed on suspicion of recurrence of uveitis. Complications documented included cataract, sunset glow fundus, nummular scars, foveal atrophy, subretinal fibrosis, and ocular hypertension.[27,28]

Main outcome measures were corticosteroid-sparing control of inflammation and time required to achieve it. Corticosteroid-sparing control of inflammation was defined by the following: (1) ≤0.5 + anterior chamber cells, ≤0.5 + vitreous cells, ≤0.5 + vitreous haze, and no active choroidal lesions on ICGA and FFA, (2) ≤10 mg of oral prednisone daily and ≤2 drops of prednisolone acetate 1% (or equivalent) daily, and (3) no declaration of treatment failure because of intolerance or safety concerns. Secondary outcome measures were CDVA at 2 year, progression to chronic recurrent granulomatous uveitis, development of sunset glow fundus, and drug-related side effects. Furthermore, patients were divided into two groups based on final CDVA: “favorable outcome” when final CDVA ≤0.3 logMAR and “unfavorable outcome” when final CDVA >0.3 logMAR.

Statistical analysis

Data were collected and stored in a spreadsheet using Microsoft Excel software. Data management and coding were done in Excel. Data were analyzed using Version SPSS 16.0 (IBM Inc., Chicago, IL, USA). Descriptive analysis was primarily carried out, where categorical variables were presented in the form of frequencies and percentages and continuous variables in the form of mean ± standard deviation. Data were analyzed using the Student t-test and Chi-square test. Univariate analysis was performed to identify prognostic variables. A multivariate analysis was done for factors that were significant in univariate analysis. Factors analyzed for significance included age > 40 years, gender, time between onset and presentation, CDVA at presentation, mean dose of azathioprine, and late-stage complications. P <0.05 was considered statistically significant.

RESULTS

During the study period, 17 patients with new onset of acute VKH disease met the inclusion criteria. The mean age at presentation was 34.94 ± 11.04 (range: 18–57) years. There were 3 (17.64%) male and 14 (82.35%) female patients. Ethnically, all patients were of Asian origin. The mean interval between onset of symptoms and presentation was 14.4 ± 11.08 (range: 1-30) days. All patients had bilateral involvement and presented with sudden onset decreased vision. Headache and tinnitus were observed in 8 (47.05%) and 3 (17.64%) patients, respectively. None of the patients had vitiligo, alopecia, or poliosis. The mean CDVA at presentation was 0.89 ± 0.8 (range: 0.1–2.4) logMAR. Anterior segment findings included anterior chamber cells in 27 (79.41%) eyes, granulomatous keratic precipitates in 10 (29.41%) eyes, and nongranulomatous keratic precipitates in 4 (11.76%) eyes [Table 1]. Shallow anterior chamber was present in 8 (23.52%) eyes. Ultrasound biomicroscopy done in these eight eyes did not reveal any findings suggestive of rotation of ciliary processes. Furthermore, none of these eyes had angle-closure glaucoma at presentation. Posterior segment manifestations included exudative retinal detachment in 34 (100%) eyes and optic disc hyperemia in 23 (67.64%) eyes. Four eyes had shallow anterior chamber with choroidal folds and 360° shallow peripheral prominent choroidal detachment. Gonioscopy in these four eyes revealed narrowed but open angles. Ocular ultrasonography was performed in 22 (64.7%) eyes. The findings on ultrasonography included retinal detachment in 22 (100%) eyes, choroidal thickening of more than 2 mm in 19 (86.36%) eyes, and choroidal detachment in 4 (18.18%) eyes. OCT was performed in 16 (47.05%) eyes. OCT analysis revealed neurosensory detachment at the macula in 12 (75%) eyes, subretinal fibrin in 7 (43.75%) eyes, and retinal pigment epithelium undulations in 4 (25%) eyes. Probable and incomplete VKH disease was diagnosed in 14 (82.35%) and 3 (17.64%) patients, respectively, using the revised diagnostic criteria.

Table 1.

Clinical characteristics of the study population at presentation

Signs	Eyes (n=34), n (%)
Anterior segment findings
Anterior chamber cells	27 (79.41)
Granulomatous keratic precipitates	10 (29.41)
Shallow anterior chamber	8 (23.52)
Posterior synechiae	6 (17.64)
Cataract	4 (11.76)
Peripheral anterior synechiae	1 (2.94)
Posterior segment findings
Serous retinal detachment	34 (100)
Vitritis	29 (85.29)
Macular detachment	24 (70.58)
Optic disc hyperemia	23 (67.64)
Subretinal fibrin (on OCT)	7 (20.58)
Choroidal detachment (on ocular USG)	4 (11.76)

Ocular USG and OCT were performed in 22 and 16 eyes, respectively. USG: Ultrasonography, OCT: Optical coherence tomography

Sixteen (94.11%) patients received intravenous pulse methylprednisolone, followed by oral corticosteroids, and 1 (5.88%) patient was directly started on oral steroids. The mean initial dose of oral steroids was 1.29 ± 0.21 mg/kg/day. The mean dose of azathioprine was 2.15 ± 0.25 mg/kg/day. Nine (52.94%) patients were started on oral azathioprine simultaneously with oral steroids. In the remaining 8 (47.05%) patients, azathioprine therapy was initiated within 2 weeks of starting oral steroids. The mean follow-up period was 27.53 (range: 24–48) months. In two (11.76%) patients, azathioprine was discontinued after 3 and 5 months of therapy, respectively, due to a rise in the level of liver enzymes. One of these patients whose azathioprine was discontinued at 3 months showed recurrence of anterior segment inflammation 2 months after discontinuation. He was on 30 mg oral steroids when his azathioprine was discontinued. He was subsequently treated with cyclosporine A. Another patient whose azathioprine was discontinued due to raised liver enzymes presented with bilateral optic disc edema at 4 months after discontinuation of azathioprine. Repeat ICGA and FFA were suggestive of recurrence of choroidal inflammation. He was subsequently treated with cyclosporine A. Both of the above patients developed sunset glow fundus on follow-up [Table 2]. While on azathioprine, none of the eyes progressed to chronic recurrent granulomatous anterior uveitis. Ocular complications included cataract in 6 (17.64%) eyes, ocular hypertension in 2 (5.88%) eyes, and epiretinal membrane in 1 (2.94%) eye [Table 2].

Table 2.

Long-term complications in Vogt-Koyanagi-Harada disease

Complications	Eyes (%)
Cataract	6 (17.64)
Foveal atrophy	6 (17.64)
Granular pigmentary changes at foveal area	6 (17.64)
Sunset glow fundus	4 (11.76)
Nummular scars	4 (11.76)
Subretinal fibrosis	2 (5.88)
Ocular hypertension	2 (5.88)

Corticosteroid-sparing effect was achieved in 14 (82.35%) patients. The median time to achieve the steroid-sparing effect was 3 months (range: 1.5–5 months). The mean cumulative dose of corticosteroids was 4067.82 ± 791.89 mg. Slow tapering of oral steroids as per mentioned protocol was followed in all except one patient. This patient was a 26-year-old female (35 kg) started on initial dose of 50 mg oral steroids and azathioprine 25 mg. Since she was intolerant to oral steroids, it was tapered over a period of 6 weeks to 10 mg once a day. At 6 weeks while on 10 mg oral steroids and 75 mg azathioprine, repeat ICGA and FFA did not showed any disc dye leakage or active choroidal lesion. Steroid dose of 5 mg was continued for next 2 weeks and stopped while oral azathioprine 75 mg was continued.

The mean CDVA at presentation was 0.89 ± 0.8 logMAR, with 14 (41.17%) eyes having vision of >0.8 logMAR. Following treatment with oral steroids and azathioprine, CDVA at 1, 3, and 6 months improved to 0.21 ± 0.2, 0.13 ± 0.2, and 0.14 ± 0.27 logMAR, respectively. The mean CDVA at 2 years was 0.14 ± 0.24 logMAR, with 25 (73.52%) eyes having vision better than or equal to 0.1 logMAR and 28 (82.35%) eyes having vision better than or equal to 0.3 logMAR. Thus, favorable outcome was seen in 28 eyes. The mean CDVA at 2 years was lower than 3 months due to progression of cataract in four eyes and two eyes developed cataract while on treatment. The difference between visual acuity at presentation and 2 years following treatment was statistically significant (P < 0.01). Two (5.88%) eyes had CDVA of >0.8 logMAR at 2 years. Based on univariate analysis, risk factors associated with final CDVA worse than 0.3 were age more than 40 years (P < 0.0001), presenting CDVA >1.0 logMAR (P = 0.024), cataract (P < 0.0001), and foveal atrophy (P < 0.0001) [Table 3]. Multivariate analysis failed to identify any independent risk factor (s).

Table 3.

Odds ratio of various variables on outcome by univariate analysis

Factors	Favorable outcomes (CDVA ≤0.3) (n=28 eyes)	Unfavorable outcomes (CDVA >0.3) (n=6 eyes)	OR	95% CI for OR	P
Age (>40 years) (patients)	5	5	2.00	1.076–3.717	<0.0001
Time interval between onset and presentation (>1 week)	14	4	4.286	0.426–43.141	0.189
Mean azathioprine dose (>1.5 mg/kg/day)	20	2	0.30	0.042–1.174	0.211
CDVA at presentation (>1.0)	8	4	10.5	1.014–108.755	0.024
Subretinal fibrin	11	0	0.783	0.631–0.971	0.094
Cataract	1	5	6.00	1.003–35.908	<0.0001
Nummular scars	4	0	0.821	0.691–0.976	0.262
Sunset glow fundus	4	0	0.821	0.691–0.976	0.262
Foveal atrophy	2	4	54.00	3.93–741.790	<0.0001
Subretinal fibrosis	2	0	0.844	0.727–0.979	0.545

Visual acuity was measured in logMAR. CDVA: Corrected distance visual acuity, CI: Confidence interval, OR: Odds ratio

DISCUSSION

This study highlights the efficacy of azathioprine and systemic corticosteroids as a first-line therapy in new onset of acute VKH disease in an Indian population. All patients included in this study received uniform treatment. We reported good corticosteroid-sparing effect in 82.35% of the cases. Similarly, at the last follow-up, 82.35% of the cases achieved CDVA of 0.3 logMAR or better.

Various studies reported a recurrence rate varying from 22.5% to 79% with corticosteroids alone.[11,29,30,31,32,33] Hence, it is now an undisputed fact that IMTs play a significant role in the management of VKH disease.[15] Various nonsteroidal immunomodulatory drugs widely used in the care of patients with VKH disease include azathioprine, cyclosporine, mycophenolate mofetil, and methotrexate.[16,17,18,19,20] In a randomized controlled trial comparing efficacy between methotrexate and mycophenolate mofetil, 74% of patients taking methotrexate achieved corticosteroid-sparing effect as compared to 53% of patients taking mycophenolate mofetil.[19] In a cohort of Chilean patients with chronic VKH disease, 50% of patients randomized to prednisolone and azathioprine achieved control over inflammation compared to 78% of patients receiving prednisolone and cyclosporine.[20] The specific IMT that is most effective for treatment of VKH disease still remains questionable considering the challenging results given the heterogenicity in the disease severity and duration, small sample sizes, nonstandardized regimens, and outcome criteria.

Chronic recurrent VKH disease in comparison to the new-onset acute VKH disease is significantly associated with poor visual acuity and mean retinal sensitivity at the last follow-up.[5,34] The chronic granulomatous uveitis phase of VKH disease is more refractory to treatment and might require more than one IMT for control of inflammation.[35] Yang et al. evaluated a reduced dose of corticosteroids (0.4–0.8 mg/kg/day) combined with varied IMT (cyclosporine A, cyclophosphamide, methotrexate, and azathioprine) in three groups of patients of VKH disease differing at the time to initiate treatment (within 2 weeks, between 2 weeks and 2 months, and after 2 months of onset of uveitis). These authors reported a complete control of intraocular inflammation (absence of anterior uveitis and active choroiditis) in 100%, 96.7%, and 91.3% at 3 months after treatment.[36] Paredes I et al. proposed that patients receiving corticosteroids and early IMT had better visual outcomes compared to corticosteroids with or without delayed IMT.[17] In another prospective study involving 38 Asian patients, Abu El-Asrar et al. evaluated the effectiveness of mycophenolate mofetil as a first-line therapy combined with systemic corticosteroids in the initial onset of acute VKH disease.[16] They reported a corticosteroid-sparing effect in all patients, and a visual acuity of 20/20 was achieved in 93.4% of eyes while none of the eyes progressed to the chronic recurrent phase or sunset glow fundus. Contrary to this, Ei Ei Lin et al. documented poor control of uveitis in all patients of VKH disease receiving mycophenolate mofetil.[37] In the present study, we achieved a corticosteroid-sparing effect in 82.35% of VKH patients. We also had excellent visual acuity outcomes with 82.35% of eyes attaining a CDVA of 0.3 logMAR or better and 73.52% of eyes achieving a CDVA of 0.1 logMAR or better at 2-year follow-up. Our visual outcomes were better than those achieved by two large retrospective studies where it was observed that 70%–80% of patients with acute VKH disease achieved a final visual acuity of 0.3 logMAR or better.[5,38,39] Importantly, none of the eyes progressed to chronic recurrent anterior granulomatous uveitis, but sunset glow fundus developed in four eyes.

Adverse effects such as nausea, vomiting, mild liver toxicity, and leukopenia have been reported with azathioprine but are dose-dependent and reversible.[38] In the present study, azathioprine was well tolerated with minimal side effects in most patients. Two (11.76%) patients, however, required discontinuation of IMT owing to mild hepatic toxicity. None of the patients developed severe side effects like myelosuppression. Another important aspect associated with any IMT is the cost of treatment and the underlying financial burden it imposes on the patient.[21] This aspect becomes especially important in third-world countries. Even though various IMTs such as cyclosporine, mycophenolate mofetil, and cyclophosphamide have shown good clinical results, we preferred azathioprine over others due to its effective corticosteroid-sparing action, drug safety profile, and its cost-effectiveness.

The limitation and possible biases in the present study are that it was performed in a single institution, was retrospective in nature, and had a small sample size. The treatment modalities were greatly influenced by the availability and affordability of drugs as well as the personal preferences of the uveitis specialist involved. Nevertheless, given the fact that patients with VKH disease tend to be referred to a tertiary eye care institutes, these data may help some clinicians draw useful conclusions.

CONCLUSION

Azathioprine has an effective steroid-sparing effect in new-onset acute VKH disease. A treatment strategy involving azathioprine as a first-line therapy in new onset of acute VKH disease is associated with good visual outcomes and low rates of recurrence.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.