16S
/
18S
metabarcoding versus culture for aetiological assessment in infectious keratitis: A prospective paired‐sample study
Sandra M. Skovdal, Christen Rune Stensvold, Henrik Vedel Nielsen, Stine Elkjær Nielsen, Jesper Hjortdal, Lise Kristensen, Anders Ivarsen Abstract
Purpose
Infectious keratitis (IK) is a vision‐threatening condition with diverse microbial aetiologies, yet conventional culture‐based diagnostics often have limited sensitivity. We investigated whether 16S/18S rRNA amplicon‐based next‐generation sequencing (NGS) can enhance aetiological assessment when used alongside standard microbiological methods.
Methods
In this prospective paired‐sample study, 57 patients with IK underwent corneal sampling with conventional culture, PCR (HSV/VZV and Acanthamoeba ) and 16S/18S metabarcoding. Microbiological findings were interpreted in a structured workflow incorporating clinical correlation and comparison with samples from the contralateral healthy eye.
Results
Excluding entirely viral cases, NGS did not significantly increase the overall diagnostic outcome compared with culture and PCR alone ( n = 53, 42% vs. 51%, p = 0.27) but improved the combined diagnostic yield (n = 53) to 58%. NGS identified clinically relevant pathogens in 39% of cases, including four cases not detected by culture, whereas culture identified nine cases missed by NGS. Concordance between methods was observed in 55% of positive cases. NGS on samples from the contralateral eye aided interpretation of NGS data in over 25% of cases by distinguishing commensal background from likely pathogens. Although microbiological findings led to treatment modification in 16% of cases, these were all in cases diagnosed by conventional methods.
Conclusion
Based on our findings, 16S/18S metabarcoding may be useful as a targeted supplementary test in atypical, refractory or culture‐negative IK cases where conventional sampling is compromised. Findings should be interpreted alongside conventional microbiology, clinical findings and preferably contralateral‐eye sampling.