PROJECT SUMMARY
Timely and accurate diagnosis of ocular infections is critical to guide early effective therapy and save vision.
However, because current methods for diagnosis of ocular infections are time-consuming and not sensitive,
empirical treatment with broad-spectrum therapies selected on the basis of clinical judgment continues to be the
mainstream approach. Since ocular infections are caused by a wide range of organisms, which may present with
overlapping features, many patients are treated unnecessarily with inappropriate antimicrobials. These broad-
spectrum scattershot approaches lead to increased toxicity, select for resistant bacteria, and are increasingly
compromised by the global emergence of antimicrobial resistance (AMR). AMR is a global pandemic that poses
a major threat to vision health, as ocular pathogens are becoming increasingly resistant to first-line antimicrobials
used in ophthalmology. Unlike infections at many other sites, delays in diagnosis and in initiation of targeted
effective treatment of ocular infections can lead to serious consequences. Treatment failures or reduced efficacy
due to AMR can result in delayed infection resolution and consequences ranging from incremental visual loss to
loss of the eye. Metagenomic sequencing-based approaches are increasingly being explored to detect
pathogens and predict their antimicrobial resistances. Recent technological advancements now place the
development of rapid near point-of-care diagnostic tests that employ third-generation sequencers (e.g., Oxford
Nanopore) within reach. Our goal is to capitalize on these advances, and combine them with novel bioinformatic
tools for real time pathogen identification and methods for pathogen enrichment to create a completely
cultureless precision diagnostic tool for improved diagnosis of sight-threatening ocular infections. First, we will
develop and validate a practical metagenomic sequencing approach for rapid, sensitive and comprehensive
detection of bacteria, fungi and acanthamoeba from ocular samples using the portable Nanopore sequencer.
This will be accomplished using specimens derived from ex-vivo models of keratitis and endophthalmitis with
known microbial loads. Clinical validation will be performed using our existing large biorepositories of clinical
specimens collected from patients presenting with sight-threatening keratitis and endophthalmitis and additional
samples to be collected prospectively. Additionally, we will also adopt and validate direct association approaches
for antibiotic resistance prediction using the minimum of new sequence information necessary. At the
conclusion, we will have developed a rapid and culture-independent precision diagnostic approach that would
transform ocular infection diagnosis and treatment from an empirical one-size-fits-all approach to a precise and
more effective one tailored to the patient’s needs. This has the potential to significantly advance clinical
management and help improve the care of patients suffering from sight-threatening infections.
