Quantification of 16S Bacteria and Torque Teno Virus in Dry Eye Syndrome and Clinical Correlations

Abstract

Purpose: Dry eye syndrome (DES) causes significant ocular morbidity world-wide. The role of the ocular surface microbiome in DES has not been established. This study investigated the 16S-bacterial and viral load of torque teno virus (TTV) in patients with DES and their clinical correlations.

Methods: Patients with clinical diagnosis of DES and controls were recruited from the University of Miami and Miami Veterans Affairs Healthcare. All underwent full exam including corneal fluorescein staining, Schirmer, tear break-up time (TBUT), and ocular surface disease index (OSDI) questionnaire. Conjunctival swabs were analyzed by qualitative and quantitative PCR. Analyses were performed with Wilcoxon signed rank and Spearman correlation tests. Multivariate regression and logistic regression models were used to analyze the contribution of each variable to the total 16S load and the status of DES.

Results: A total of 71 eyes of 37 patients (22 with DES and 15 controls) were included. Mean age was 63.6 in the DES group and 47.7 in the control group (p-value 0.0005). There were 20 males (91%) in the DES group and 12 (80%) in the control group. Between eye correlation was measured using the intraclass correlation coefficient (ICC=0.520). Median level of 16S bacteria (normalized to human actin copies) were 0.034 in the DES patients and 0.120 in the controls (p-value 0.005). There was a negative correlation between the total 16S bacteria and the OSDI score (rho -0.526, p-value<0.0001). Similar negative correlation was observed between the total 16S bacterial load and the staining score (rho -0.373, p-value 0.003). Both Schirmer score and TBUT positively correlated with 16S load (rho 0.464, p-value 0.0002 and rho 0.402, p-value 0.002, respectively). The multivariate regression model that included age, OSDI, TBUT, corneal stain, and Schirmer was significant (p-value 0.001, adjusted R²=0.28) in predicting total 16S. The most significant variable was the OSDI. The logistic regression model that included age, gender, TTV, and 16S showed that both TTV and 16 loads negatively correlated with DES status (OR: 0.025, 0.002, respectively).

Conclusions: This study highlights the disruption of the normal ocular surface microbiome and potentially infectious role in DES. Decreased level of 16S bacterial load and TTV correlated significantly with DES. Whether this relationship is correlative or causative remains to be determined.