Anterior-segment spectral domain optical coherence tomography in epidermolysis bullosa

Abstract

Purpose

Assess epidermolysis bullosa (EB)-related corneal pathology using anterior segment optical coherence tomography (AS-OCT) and correlate imaging with clinical metrics in EB patients versus age-matched controls.

Methods

EB patients and controls were recruited during an EB conference (July 2018) and at Tufts Medical Center (June–August 2019). Subjects completed a questionnaire, had best corrected visual acuity (BCVA) tested, and underwent AS-OCT scanning. Stromal and epithelial thickness were measured. Depth, length, and type of the three largest lesions were assessed by a masked examiner using a novel pathology grading system. Multivariate analysis of AS-OCT findings and clinical metrics was performed.

Results

62 EB patients and 60 age-matched controls were enrolled. Mean BCVA was 1.8 lines worse in patients (p < 0.001). Vision loss was associated with increased stromal thickness. Discrete lesions were seen in 60.2% of patient eyes, averaging 1.71 ± 1.75 lesions in patients and 0.14 ± 0.42 in controls (p < 0.001). Mean primary lesion depth was 151.88 ± 97.49 μm in patients. Patients showed significant stromal thickening versus controls and lesions were most common in the periphery and inferiorly. Differences in frequency and duration of abrasions and severity of pain were all statistically and clinically significant in patients versus controls (p < 0.001).

Conclusions

AS-OCT can visualize and quantify differences in the corneas of EB patients compared with age-matched controls. Novel findings include quantification of average depth, length, and severity of discrete lesions, and sparing of the superior quadrant from stromal thickening in EB patients. These results support use of AS-OCT and a questionnaire in clinical trials for new EB therapies.

Introduction

Epidermolysis bullosa (EB) is a devastating connective tissue disorder that can lead to disfigurement, immobility, and blindness [1,2]. It is caused by genetic mutations affecting adhesion proteins that are critical for anchoring stratified squamous epithelium to underlying tissue [[3], [4], [5]]. Phenotypic expression involves blistering and poor wound healing, typically of the skin, ocular surface—including the cornea, conjunctiva, and lids— and the gastrointestinal and genitourinary tracts [6]. EB is classified as an orphan disease affecting an estimated 19.6 people per million in the United States [7].

Ophthalmic manifestations are most common in two EB subtypes known as recessive dystrophic EB (RDEB), caused by impaired type VII collagen function, and junctional EB (JEB), caused by absence of laminin 5 [8]. The precise incidences of complications in EB are not well established. Previous studies have reported a wide range: in RDEB, corneal blisters and erosions are estimated to occur in 35–74% of patients, scarring in 24–41%, and vision loss in 3–64% [[9], [10], [11], [12], [13], [14]]. In JEB, estimated incidences range from 24 to 80%, 14–80% and 0–33%, respectively. The third subtype, known as simplex (EBS), causes abrasions in fewer patients (up to 6%), yet vision loss still occurs in approximately 15% [1]. Standard management is lubrication and light avoidance [6]. Advanced therapies for those severely affected include bandage contact lenses to stabilize the corneal surface and autologous limbal epithelial cell transplant to promote epithelialization [15].

At present, numerous phase 1–2 clinical trials of disease-specific therapies are in progress. Standard endpoints for skin trials include changes in the surface area of discrete and active EB lesions, and biopsy-proven increases in anchoring fibrils [16,17]. To our knowledge, no non-surgical therapies are being developed for ophthalmic use. One critical hurdle for drug development is the inability to objectively show clinically meaningful disease modification. Without the ability to non-invasively “biopsy” the cornea, especially in children, subjective assessments such as visual acuity testing, interpretation of photographs, and slit-lamp examination remain the primary tool for assessment. Thus, there is a need for objective and reliable methods of quantifying and monitoring ocular surface pathology.

In the last three decades, optical coherence tomography (OCT) has revolutionized the practice of ophthalmology. Although OCT is most routinely used to image the posterior segment, anterior segment (AS) protocols allow high-resolution, depth-resolved imaging of the cornea [18,28,29,40]. As such, AS-OCT is a valuable tool for non-invasive assessment of corneal pathology[33,[37], [38], [39]]. To our knowledge, large studies of AS-OCT in EB patients have not been published.

Prior studies using animal models have shown that 89% (17/19) of EB-affected murine corneas are histologically different than wild-types [19]. Specifically, EB-affected mice show epithelial hyperplasia and stromal fibrosis. Thus, we hypothesized that EB patients would show thickening of both epithelium and stroma on AS-OCT when compared with age-matched controls. Furthermore, such increases in epithelial and stromal thickness would correlate with vision loss.

Herein, we describe the first use of AS-OCT to visualize EB-related corneal lesions. We generated a novel pathology grading system for EB lesions, based upon grading systems commonly used for retinal disease [20,21]. We also tested best corrected visual acuity (BCVA) and applied a short questionnaire tool to query corneal abrasion frequency, duration, pain severity, and onset and etiology of scarring and vision loss. Finally, we explored associations between AS-OCT findings and clinical metrics, including ocular symptoms and vision.