Efficacy of 0.5% Levofloxacin and 5.0% Povidone-Iodine Eyedrops in Reducing Conjunctival Bacterial Flora: Metagenomic Analysis

Abstract

Bacteria associated with postoperative endophthalmitis mostly originate from the normal bacterial flora of the patient's conjunctiva and eyelids, so the incidence of endophthalmitis may be reduced by eliminating the ocular and adnexal flora before surgery. We assessed the effectiveness of eyedrops of 0.5% levofloxacin and 5.0% povidone-iodine (PVI) in reducing conjunctival bacterial flora by metagenomic analysis. A total of 2.4 × 10⁶ high-quality sequencing reads were generated from 93 conjunctival samples obtained from 31 eyes scheduled for cataract surgery before prophylactic therapy (group 1), after administration of 0.5% levofloxacin eyedrops into the conjunctival sac 8 times before surgery (group 2), and at 3 minutes after instillation of 5.0% PVI solution in the conjunctival sac (group 3) followed by surgery irrigation. The alpha diversity and beta diversity results demonstrated that group 3 had the least richness and biodiversity. Corynebacterium, Pseudomonas, Staphylococcus, Acinetobacter, and Streptococcus were predominant in all samples. The relative abundance of these bacterial species was 30.94%, 27.48%, 5.26%, 4.55%, and 2.61% in group 1, 16.32%, 44.10%, 2.19%, 5.39%, and 0.97% in group 2, and 5.90%, 65.55%, 0.39%, 5.36%, and 0.10% in group 3, respectively. The most easily and difficultly eliminated were Corynebacterium and Pseudomonas, respectively. In conclusion, the metagenomic analysis using high-throughput sequencing provides a scientific way for evaluating the effectiveness of a disinfection method from the perspective of analyzing the composition and diversity of the conjunctival microbiome. Despite the use of preoperative antisepsis regimens, the ocular surface of patients receiving cataract surgery could not be rendered completely aseptic, indicating that more strict disinfection methods need to be adopted to reduce the risk for anterior chamber contamination and endophthalmitis after cataract surgery.

Figure 1

The differences in the indices of observed species, Chao1, Shannon, and Simpson were significant between groups, especially when group 3 was compared with group 1 or group 2. Group 1 was measurably more diverse than group 3. (a) Observed species, treatment. (b) Chao1, treatment. (c) Shannon, treatment. (d) Simpson, treatment.

Figure 2

The principal coordinate analysis was derived from weighted UniFrac distances of 16S rRNA gene between microbial communities of the three groups. Values in parentheses on the axis labels indicate that the percentage variation accounts for each axis. (a) The samples were clustered based on different drug interventions. (b) Groups 1 and 2 were a continuum, and there was little distinction between these two groups. (c, d) There was significant difference between group 3 and the other two groups.

Figure 3

The change tendency of relative abundances of bacterial genera in different groups.

Figure 4

Relative abundances of the top 10 genera in the three groups in a heat map. Values in color key indicated the relative abundance of each genus.

Figure 5

Composition and relative abundances of the conjunctival microbiota in the three groups.