Evaluation of Fungitell (1,3)-β-D-glucan assay in tear samples for rapid diagnosis of fungal keratitis.

Abstract

Fungal keratitis (FK) is a serious suppurative and ulcerative corneal infection leading to blindness and vision loss. Rapid diagnosis of FK can contribute to prompt clinical management with early recovery. The study aimed to standardize the detection of (1,3)-β-D-glucan (BDG) and establish the diagnostic cut-off concentration in tears of suspected FK patients along with non-infected controls. This prospective multicentric study was conducted between the period of August 2022 and April 2024. Samples were collected from three tertiary eye-care facilities across India. All suspected FK patients were enrolled in the study. Prior to tear collection, the eye and eyelid were gently cleansed using sterile normal saline (NS) and lint-free tissue paper. Subsequently, 50 µL of sterile NS was instilled into the eye, followed by tear sample collection after 60 s using sterile fine microtips. Tear samples were collected from the contralateral eyes of the FK patients, and those from healthy volunteers served as controls. The concentration of BDG in tears in varying dilutions was quantitatively measured using the Fungitell Assay Kit (Associates of Cape Cod, East Falmouth, Massachusetts). A total of 53 tear samples were analyzed at 1:10 and 1:20 dilutions. The receiver operating curve revealed an area under the curve (AUC) of 0.919 for the 1:10 dilution, with a cut-off value of 123 pg/mL, yielding a sensitivity of 100% and specificity of 84.85%. The corresponding Youden Index was 0.798. At the 1:20 dilution, the AUC was 0.898 with a cut-off of 84 pg/mL, achieving a sensitivity of 70% and specificity of 96.88% with a Youden Index of 0.670. Given the higher specificity at the 1:20 dilution, it was further validated in 145 tear samples. The validation cohort demonstrated a sensitivity of 95.56%, specificity of 83%, positive predictive value (PPV) of 71.67%, negative predictive value (NPV) of 97.65%, and diagnostic odds ratio of 112.4. Notably, a significantly higher BDG concentration (P < 0.0001) was observed in infected tear samples compared to controls (270.6 ± 128.5 vs 37.31 ± 31.32). Detection of BDG in tear samples is a new, non-invasive, and rapid technique showing excellent performance and can be effectively implemented for diagnosing FK in laboratories.