Tear soluble MUC16 and inflammatory biomarker profiles in diabetic retinopathy-associated keratoconjunctivitis sicca: Associations with disease severity and combined discriminative performance

Background: Ocular surface dysfunction is common in patients with diabetic retinopathy (DR), but the profiles of tear mucins and inflammatory biomarkers associated with keratoconjunctivitis sicca (KCS) remain unclear. We analysed the relationships of tear soluble mucin 16 (sMUC16), interleukin-6 (IL-6), interleukin-8/C-X-C motif chemokine ligand 8 (IL-8/CXCL8), and matrix metalloproteinase-9 (MMP-9) with KCS and its severity in patients with DR. We evaluated the discriminative performance of combined biomarker testing. Methods: This single-centre retrospective study included 73 patients with DR, of whom 44 had KCS, and 29 did not. Tear sMUC16, IL-6, IL-8/CXCL8, and MMP-9 were measured, and the KCS phenotype and severity were assessed together with clinical ocular surface parameters. Correlation analysis and logistic regression were used to identify tear biomarkers independently associated with KCS; receiver operating characteristic (ROC) curves were used to compare the discriminative performance of individual biomarkers and the combined model; and bootstrap resampling was used for internal validation. Results: Tear levels of sMUC16, IL-6, IL-8/CXCL8, and MMP-9 were higher in patients with DR and KCS than in those without KCS, and all four biomarkers generally increased with increasing KCS severity (P<0.05). Multivariable logistic regression showed that sMUC16, IL-6, and MMP-9 were independently associated with KCS, with adjusted odds ratios (ORs) of 2.624, 2.518, and 2.721, respectively (P<0.05). The areas under the curve (AUCs) for the individual biomarkers ranged from 0.706–0.777; the combined model constructed from sMUC16, IL-6, and MMP-9 had an AUC of 0.879 (95% confidence interval [CI]: 0.794–0.963), which was higher than that of IL6, the best-performing individual biomarker (P=0.025), with a sensitivity of 81.82% and a specificity of 89.66%. After 1,000 bootstrap resamples, the optimism-corrected AUC was 0.860. Conclusion: In patients with DR, KCS was accompanied by abnormal changes in tear sMUC16 and inflammation-related biomarkers; sMUC16, IL-6, and MMP-9 provided relatively complementary information, and their combined measurement improved the ability to identify KCS in this study cohort.