Purpose
To evaluate the structural and functional changes of the meibomian gland and correlate with subjective and other objective dry eye parameters in Stevens-Johnson syndrome (SJS) patients.
Design
Prospective cross-sectional study.
Methods
This study recruited 60 patients (120 eyes) with SJS and chronic ocular sequelae. All patients underwent evaluation with a Standard Patient Evaluation of Eye Dryness (SPEED) questionnaire, tear osmolarity, Keratograph 5M, LipiView, Schirmer’s 1, corneal staining, Meibum Expression score (MES, 0-3), and Meibum Quality score (MQS, 0-3). The meibomian gland dropout area (meiboscore) was calculated for both lids and graded on a 4-point scale from 0 to 3.
Results
The mean age of patients (n = 60) was 31.08 ± 12.94 years; 25 were males. The commonest cause for SJS, in 51 patients (85%), was drug reaction. The mean tear osmolarity, lipid layer thickness (LLT), Schirmer’s test-1, and tear break-up time was 322.70 ± 17.82 mOsm/L, 53.07 ± 27.0 nm, 6.62 ± 5.74 mm, 3.38 ± 1.90 seconds, respectively. Out of 240 eyelids (both upper and lower), 160 (65%) eyelids revealed severe meibomian gland loss. The mean upper and lower lid gland loss was 77.36 ± 28.82% and 76.65 ± 29.33%, respectively, Seventy-four eyelids (61%) had no expressible glands. Meiboscore showed positive correlation with SPEED ( P < .001), corneal staining scores ( P < .001), MES ( P < .001), and MQS ( P < .001). The LLT negatively correlation with SPEED ( P < .01), meiboscore ( P < .001), MES ( P < .01), and MQS ( P < .001).
Conclusion
Significant alterations in anatomical and functional aspects of the meibomian gland are seen in SJS. The high meiboscore, MES, MQS, and decreased LLT contributed to the worsening dry eye state, as seen by their correlation with other dry eye parameters. This study highlights the need to evaluate meibomian gland structure and function in patients with chronic ocular sequelae of SJS.
INTRODUCTION
S tevens-Johnson syndrome (SJS) and its more severe form–toxic epidermal necrolysis–are rare, life-threatening diseases affecting the skin and at least 2 mucous membrane sites. The chronic ocular sequelae of SJS cause significant ocular morbidity, the most common being dry eye, and all the tear film components are adversely affected. , Aqueous tear deficiency (ATD) primarily occurs due to conjunctival scarring, which obstructs the ductules that transport tears from the lacrimal gland to the conjunctival cul-de-sac. Damage to the conjunctival epithelium and goblet cells in the acute phase of the disease results in mucin deficiency. Sotozono and associates and Lekhanont and associates have reported altered meibum secretion in the chronic stage of the disease. , In addition, lid margin keratinization and associated mechanical blink-related micro-trauma further damage the ocular surface, resulting in persisting ocular inflammation, conjunctivalization, and keratinization of the ocular surface. , , The pathogenesis of ATD is well documented in SJS; , , however, the etiopathogenesis and severity of meibomian gland dysfunction (MGD) are still unclear. It has been reported that around 73% to 87.5% of SJS patients have meibomian gland involvement. , , The meibomian glands are holocrine lipid-excreting glands with approximately 20 to 25 glands along the lower eyelid and 30 to 40 glands along the upper eyelid. Meibum from these glands reduces the surface tension and stabilizes the tear film, and reduces tear evaporation. Alterations in the quality and/or quantity of meibum destabilize the tear film.
Meibomian gland dysfunction is one of the commonest cause of dry eye. The pathophysiology of MGD involves a hyposecretory or hypersecretory etiology. There are terminal duct obstruction and/or qualitative/quantitative changes in glandular secretion secondary to inflammation or the presence of altered bacterial commensals at the lid margin. There is limited literature on morphological and functional aspects of meibomian gland disease in SJS. Although some studies have shown a high degree of meibomian gland dropout in patients with SJS, the correlation of meibomian gland dropout with the functional changes (quantitative lipid layer thicknesses, MQS) and their relationship with other dry eye parameters in SJS have not been clearly elucidated. ,
This study aimed to evaluate the severity of MGD and its correlation with other dry eye parameters in SJS. It was conducted to understand and emphasize the need to evaluate meibomian gland structure and function in eyes with chronic ocular sequelae of SJS.
METHODS
A cross-sectional observational study was conducted on all SJS patients seen at the ocular surface clinic of a tertiary ophthalmic centre in South India between July 2019 and January 2020. The study was approved by the Institutional Review Board and Ethics committee, and adhered to the tenets of the Declaration of Helsinki. Informed consent was obtained from every study participant. The diagnosis of SJS was confirmed with a past history of an acute hypersensitivity reaction that manifested as vesico-bullous lesions on the skin along with mucosal involvement in at least 2 sites in the acute phase. Patients were excluded with a history/clinical features of: Sjogren’s syndrome, ocular cicatricial pemphigoid, chronic graft versus host disease, chemical and thermal burns, trachoma, allergic conjunctivitis, contact lens wear, ocular trauma, and ocular surgeries like punctual cautery, mucous membrane grafting, and cataract surgery. Patients on anti-glaucoma medication, hormone therapy, pregnant and lactating women were also excluded.
DRY EYE PROTOCOL
Patients underwent a detailed dry eye evaluation according to a set protocol. The test sequence involved the least invasive to the most invasive test, and the right eye was tested first. The following tests were carried out sequentially: best-corrected visual acuity (BCVA), Standard Patients Evaluation of Eye Dryness (SPEED) questionnaire, tear osmolarity (TearLab, Tear Science), Keratograph 5M (Oculus; used to assess Tear meniscus height (TMH) and meibomian gland imaging), LipiView (Tear Science INC, and Johnson and Johnson; to assess lipid layer thickness), Schirmer’s 1, slit-lamp examination including fluorescein tear break-up time (FTBUT), corneal staining scores, gland expressibility (MES) and secretion quality (MQS) ( Figure 1 ). For standardization, all the dry eye tests were performed 2 hours after the application of lubricants. All clinical tests were performed at the first visit and under an air-conditioned room with a temperature around 20 to 24 °C to maintain the same environmental conditions.
DRY EYE EVALUATION
All subjects completed the SPEED questionnaire. The SPEED questionnaire symptom score varied from 0 (no symptoms) to 28 (severe symptoms). Tear osmolarity was measured using the TearLab osmolarity system as per the manufacturer’s protocol; the machine reports an osmolarity range between 270 to 400 mOsm/L and values below or above the range cannot be measured. A diagnostic cutoff of 308 mOsm/L was used to confirm an unstable tear film.
The Keratograph 5M, a non-contact Placido disc-based topographer and an infrared-based imaging system, was used to assess the TMH, non-invasive tear break-up time (TBUT), and meibomian gland imaging. TMH was recorded at the 6′ o clock position as per the standard guidelines. The non-invasive TBUT for most patients was unobtainable because of highly distorted mires caused by poor ocular surface, short palpebral aperture opening, and scarring; therefore, FTBUT was included for the final analysis. Meibomian gland imaging of the upper and lower lids was performed with the infrared imaging system. The machine illumination and contrast were adjusted to get the best possible image in the eyes with thickened lids, lid edema, and eyelid keratinization.
MEIBOMIAN GLAND IMAGE ANALYSIS
Meibography images taken by the Keratograph 5M were analyzed using ImageJ software ( https://imagej.nih.gov/ij/download.html ) to identify the presence of meibomian gland dropouts. The mucocutaneous junction was identified, and the boundaries of the total tarsal area were marked using the freehand selection tool. , The percentage of gland dropouts in both the upper and the lower lids were calculated by dividing the dropout area with the total area multiplying by 100. A 4-point scale, as described by Arita and associates, was used to grade the images (Grade 0: no gland loss; Grade 1: <33% gland loss; Grade 2: 33%-66% gland loss; Grade 3: >66% gland loss). The dropout grades of all the upper lids and lower lids were measured separately (0-3). For correlation analysis between meibomian gland dropout and other dry eye parameters, the meibogrades of both lids were added to give a meiboscore for that eye (0-6).
The lipid layer thickness (LLT ) was measured using the Lipiview ocular surface interferometer described by Blackie and associates. The lipid layer’s interference pattern was scored based on the interferometric color unit (ICU), where 1 ICU = 1 nm of LLT. The machine calculated the mean lipid layer thickness based on interferometric color images over a 20-second interval. The test was repeated more than once in case of an unreliable/erroneous report. An LLT value of > 75 nm was considered thicker. Schirmer’s 1 test was performed without topical anesthesia, and the wetting at 5 minutes was documented.
The FTBUT and corneal staining scores were evaluated after a single drop of fluorescein into the conjunctival sac under a cobalt blue filter. The patient was asked to blink 3 times; the interval between the last blink and the appearance of the first dry spot (break-up) on the corneal surface was recorded with a stopwatch. In eyes with conjunctivalization, the break up was recorded over the area of the cornea that still retained a corneal phenotype. The FTBUT was repeated 3 times, and an average value was taken for analysis. An FTBUT of <10 seconds was considered reduced. The corneal staining score was obtained by dividing it equally into 3 parts: superior, middle, and lower third. The staining pattern was graded as: 1, few separated spots; 2, many separated spots; 3, confluent spots for each parts. The scores of all 3 parts were added to obtain a total score from 0 to 9 for each eye.
MEIBUM EXPRESSION SCORE AND MEIBUM QUALITY SCORE
The Meibum Expression Score (MES) and Meibum Quality Score (MQS) were measured by applying digital pressure on the lower tarsal plate for 5 to 6 seconds to observe the number of expressible glands and secretion quality within the central 5 glands of the lower eyelids. The MES was graded from 0, indicating all expressible glands to 3, indicating no expressible glands as per established protocol. The MQS was also graded from 0 to 3, where 0 indicated clear meibum fluid, Grade 1 indicated cloudy liquid, Grade 2 indicated cloudy particulate fluid, and Grade 3 indicated inspissated toothpaste-like fluid.
EYELID MARGIN MORPHOLOGY
Eyelid margin abnormalities were recorded based on absent = 0 and present = 1 for the parameters like the irregularity of lid margin, presence of symblepharon, lid margin keratinization, and punctal stenosis.
STATISTICAL ANALYSIS
Descriptive and inferential statistical analysis of data was performed with SPSS Version 20.0. Frequency and percentage were used for categorical data. Data were presented as mean ± standard deviation. Paired t test was used to compare grade of meibomian gland dropout in the upper and lower eyelids. Spearman correlation was used to correlate between the degree of meibomian gland dropout and lipid layer thickness with other dry eye parameters. The statistical significance was set at 5%.
RESULTS
A total of 65 patients with SJS were screened during the study period, and 5 were excluded from the analysis because of the inability to perform dry eye tests due to the completely dermalized ocular surface. Therefore, a total of 120 eyes of 60 patients with a mean age of 31.08 ± 12.94 years were included in the final analysis. Of the 60 patients, 25 were male and 35 female. The mean duration from SJS to time to presentation was 55.70 ± 60.07 months (range 1-180). Drug reaction was the commonest cause of SJS (51 patients, 85%), with antibiotics being the most frequent medication, followed by infections (3, 5%) and unknown etiology (6, 10%). Visual acuity was classified according to the World Health Organization (International Classification of Disease 2018) visual impairment category. Severe visual impairment was observed in 64 eyes (BCVA 20/200-20/400), followed by 22 eyes with moderate visual impairment (BCVA 20/70-20/200), 16 eyes with mild visual impairment (BCVA 20/40-20/70), and normal BCVA (20/20-20/40) in 18 eyes.
The mean tear osmolarity (97 eyes), TMH, FTBUT, Schirmer’s test-1, and corneal staining scores were 322.70 ± 17.82 (mOsm/L), 0.17 ± 0.15 (mm), 3.38 ± 1.90 (seconds), 6.62 ± 5.74 (mm), and 4.20 ± 2.90, respectively. The mean meiboscore of the upper and lower lids was 2.54 ± 0.72 and 2.52 ± 0.72, respectively; MES and MQS were 2.49 ± 0.75 and 1.95 ± 0.69, as shown in Table 1 .
