# Clinical and OCT Predictors of Refractory Vogt–Koyanagi–Harada Disease

## TL;DR

This study identifies factors like poor vision and delayed treatment that predict difficult-to-treat Vogt–Koyanagi–Harada disease.

## Contribution

The study identifies baseline clinical and OCT predictors of refractory VKH disease using a large cohort.

## Key findings

- Poor baseline visual acuity was the strongest predictor of refractory VKH.
- Greater central retinal thickness and delayed pulse therapy initiation were also associated with refractoriness.
- Female sex was linked to lower odds of refractory disease.

## Abstract

To identify baseline clinical and OCT predictors of refractoriness in Vogt–Koyanagi–Harada (VKH) disease.

A retrospective cohort study.

Two hundred thirty-seven patients (472 eyes) with VKH disease seen at Yokohama City University Hospital between November 2009 and March 2024.

Patients were diagnosed according to the revised international diagnostic criteria and all received systemic corticosteroid pulse therapy. Baseline demographic, clinical, and OCT findings were collected, and logistic regression analyses were performed to evaluate associations with refractory disease, defined as ≥2 recurrences or the need for additional immunosuppressive therapy. The study was approved by the institutional review board and conducted in accordance with the Declaration of Helsinki.

Associations between baseline clinical and OCT findings, including visual acuity (VA), central retinal thickness (CRT), bacillary layer detachment (BALAD), and interval from onset to pulse therapy, and the development of refractory VKH.

Twenty-six patients (11.0%; 52 eyes) developed refractory VKH, whereas 211 patients (89.0%; 420 eyes) were nonrefractory. In the primary multivariable analysis, worse baseline VA (odds ratio [OR], 1.53; 95% confidence interval [CI], 1.19–1.96; P < 0.001), greater CRT (OR, 1.66; 95% CI, 1.19–2.32; P = 0.003), and a longer interval to initiation of pulse therapy (OR, 1.27; 95% CI, 1.02–1.57; P = 0.031) were significantly associated with refractoriness, whereas female sex was associated with lower odds of refractory VKH (OR, 0.50; 95% CI, 0.26–0.96; P = 0.039) and age at onset was not. Bacillary layer detachment, although frequent in the acute stage, was not associated with refractoriness. In a sensitivity analysis additionally adjusting for the number of pulse therapy courses, only baseline VA and the number of pulse courses remained significantly associated with refractoriness.

Poor baseline VA was the most consistent predictor of refractory VKH. Greater CRT, delayed initiation of pulse therapy, and female sex showed associations in the primary model but should be regarded as candidate predictors that require confirmation, whereas BALAD was not associated with refractoriness.

The authors have no proprietary or commercial interest in any materials discussed in this article.