We conducted a multicenter retrospective interventional case series combined with laboratory investigation of patients with MRSA keratitis who were treated at our institutions between 2014 and 2022 (Massachusetts Eye and Ear) and 2005-2021 (University of British Columbia, Canada). Prospective approval for chart review for this study was obtained by the institutional review boards at both institutions prior to the start of this work. Approval for collection of discarded bacterial isolates and laboratory characterization of these isolates was approved by the institutional review board at Massachusetts Eye and Ear (Mass General Brigham IRB). The study was completed in compliance with the US Health Insurance Portability and Accountability Act and the Canadian Personal Information Protection and Electronic Document Act as well as the ethical standards established by the Declaration of Helsinki. Patients were included in this study if they had culture-proven MRSA keratitis treated at either of our institutions during the study period. Patients were excluded from analysis at MEE if no mec gene was identified on genomic analysis. Clinical data and in vitro susceptibilities of the isolated MRSA strains were collected from the electronic health record. Sex was recorded as documented in the sex category in the medical record and analyzed as a biological variable.

Isolates collected from patients treated at MEE (N = 41) underwent additional susceptibility testing to obtain minimum inhibitory concentration (MIC) data for antimicrobials (all acquired from Sigma-Aldrich, St. Louis, MO), that were not originally tested by the clinical laboratory, including tobramycin, ofloxacin, levofloxacin, ciprofloxacin, and moxifloxacin. Minimum inhibitory concentration analysis for vancomycin was also redone for MEE isolates. Before testing, isolates were grown for two consecutive days on 5% blood agar. In vitro testing was completed using the broth microdilution method as established by the Clinical Laboratory Standards Institute (CLSI) and results were assessed using CLSI breakpoint data in the M100 Performance Standards for Antimicrobial Susceptibility Testing, 30th Edition. This edition included interpretive criteria for moxifloxacin susceptibility in S. aureus , which were used for all analyses in this study.

Isolates obtained at MEE underwent further microbiological analysis to assess (1) the presence of heterogeneous vancomycin-intermediate S. aureus (hVISA) phenotype, (2) vancomycin tolerance, and (3) delta hemolytic capability. Vancomycin-intermediate S. aureus strain Mu50, hVISA strain Mu3, and pan-susceptible S. aureus ATCC 29213 were used as reference control strains. We screened for hVISA phenotype using the BHI agar screening method previously described by Satola. Isolates were determined to be potentially heteroresistant to vancomycin if ≥1 droplet had ≥2 colonies. These colonies were propagated onto tryptic soy agar (TSA) plates and hVISA phenotype confirmed using the population analysis profile method. Vancomycin tolerance testing was performed using the broth microdilution method. ^, A broth microdilution assay with log ₂ dilutions of vancomycin (ranging from 0 μg/mL to 32 μg/mL) was prepared and inoculated according to standard procedures. Plates were incubated for 24 h at 37 °C. Following incubation, aliquots (100 µL) from wells showing no visible growth were plated onto TSA plates to determine the minimum bactericidal concentration (MBC) of vancomycin against each strain. Plates were incubated at 37 °C and colony counts were recorded after 24 and 48 h. Colony-forming units per milliliter (CFU/mL) were calculated based on colony counts and plated volume, and percent survival was determined by comparison to the starting inoculum The MBC was defined as the lowest concentration of vancomycin resulting in a ≥ 99.99% reduction in viable counts relative to the initial inoculum. Vancomycin tolerance was defined as an MBC/MIC ratio ≥32. Accessory gene regulator ( agr) dysfunction has been associated with increased survival of MRSA under vancomycin selective pressures. Delta hemolytic capability screening can be used as a proxy for agr function. ^, Delta hemolysin production was assessed using modified methods described by Cafiso with isolates streaked perpendicularly to the reference strain RN4220 on blood agar plates.

MRSA isolates from patients treated at MEE were obtained from the clinical laboratory and stored in glycerol and tryptic soy broth at −80 °C. Isolates were streaked on blood agar plates for two consecutive days before use to guarantee purity. DNA was extracted using the Qiagen DNeasy Blood and Tissue Kit (Qiagen, Germantown, MD). DNA was quantified using a Qubit fluorometer and the dsDNA high-sensitivity assay kit (ThermoFisher, Waltham, MA). We used the Nextera XT DNA library kit (Illumina, San Diego, CA) to create a DNA library for each isolate. Genomes were sequenced as 2 × 150 bp or 250 bp reads on an Illumina HiSeq or MiSeq sequencer according to the manufacturer’s specifications, with a minimum depth of coverage of 30 ×. Sequence reads were assembled de novo using CLC Genomics Workbench (Qiagen, Germantown, MD). The Center for Genomic Epidemiology pipeline ( https://www.genomicepidemiology.org/ ) was used to obtain confirmation of species identification and sequence type. Whole genome sequence data were additionally screened for antimicrobial resistance genes using ResFinder within the Center for Genomic Epidemiology pipeline.

Core genome-multilocus sequence typing (cg-MLST) phylogenetic relationships among S. aureus isolates was performed using Pathogenwatch v12.0.4 ( https://pathogen.watch ). The phylogeny was visualized and annotated with iTol v4. Clonal complexes (CCs) were determined using the MLST data by the goeBURST algorithm ( http://www.phyloviz.net/goeburst/ ). For isolates with a new/unknown sequence type, the CC was determined based on their clustering in the phylogenetic tree.

Data were analyzed using IBM SPSS Version 29 (IBM Corp., Armonk, NY). Patient characteristics, clinical presentation, and outcomes were summarized using descriptive statistics. Comparisons between groups were performed using appropriate tests based on variable type and sample size. Continuous variables (eg, age and visual acuity) were analyzed using Mann-Whitney U test. Categorical variables were compared using Fisher’s exact test. All tests were two-sided, and a P value <.05 was considered statistically significant.

Patients were predominantly older individuals with ocular surface diseases and multiple systemic comorbidities. A total of 70 episodes of MRSA keratitis in 67 patients (3 patients had two episodes separated by more than 1 year) were included in the final analysis ( Table 1 and Table S1). Forty-one patients (58.6%) were treated at MEE and 29 (41.4%) were treated at UBC. Overall, forty-three patients (61.4%) were female. The MEE cohort skewed more female than the UBC cohort (MEE N = 28, 68.3%; UBC N = 15, 51.7%). The median age of all patients was 63.5 years (range, 13-96 years). The average age of the patient population at UBC (70.8 ± 18.8 years) was significantly older than the average age at MEE (56.8 ± 25.3 years; P =.01, 95% CI _difference : 2.9-25.1).

Ocular surface disease was common in this population, with 44 patients (67.7%) having at least 1 ocular surface disease reported. Forty-five (65.2%) patients had some history of ocular surgery, with 16.1% of patients presenting with MRSA keratitis in the 2 weeks following an intraocular surgery. Most patients who reported contact lens use were using bandage contact lenses (N = 18, 26.9%). Less common in this population was eyelid or lacrimal apparatus pathology (22.9%), a history of ocular trauma (17.1%), and soft refractive contact lens use (4.5%). Systemic comorbidities were prevalent, with 61 cases (87.1%) reporting at least one systemic comorbidity (median 2.5, range 0-8). Common comorbidities included hypertension (N = 24, 34.8%) and diabetes mellitus (N = 16, 23.5%). Seven cases (10.1%) had concurrent immunosuppression, defined as taking an immunosuppressive medication or having a condition that results in immunodeficiency (not including diabetes mellitus). Four cases (5.7%) were in two patients with Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis. Five cases (7.2%) were in patients with a history of reported intravenous drug use.

Topical medication use prior to presentation was relatively common in the study population. A small fraction of patients was on topical steroids (N = 24, 38.1%) or antibiotics at presentation (N = 22, 34.9%). Only 17 cases (25.4%) had extensive healthcare exposures in the year preceding the episode, defined as residence in a long-term care facility or inpatient hospitalization for 2 or more consecutive days. Fourteen cases (20.9%) had a history of prior MRSA infection (MEE 14.6%, UBC 27.6%). Table S1 lists detailed demographics and clinical presentation data for the study population.

Average time to clinical evaluation from symptom initiation was 14.3 days ± 20.3. The average time to positive culture was 2.4 days ± 1.8. Upon presentation to our centers, most patients (n = 68, 97.1%) were empirically treated with topical antibiotics ( Table 2 ). The most primary regimen was a combination of fortified vancomycin and tobramycin (65.7%), followed by vancomycin combined with other antibiotics (15.7%) and vancomycin monotherapy (7.1%). Nonvancomycin-based regimens were employed in only 8.6% of the cohort. Procedural or surgical interventions were required in 17 (24.3%) cases. Among these, penetrating keratoplasty was the most common surgical procedure (10.0%), while cyanoacrylate glue application with bandage contact lens and tarsorrhaphy were each required in 4.3% of patients. More complex interventions, including combined gluing/tarsorrhaphy (2.9%), corneal patch grafting (1.4%), and keratoprosthesis replacement (1.4%), were utilized in a minority of the population.

Pre- and post-treatment visual acuity in logarithm of minimal angle of resolution (logMAR) for all patients with documented visual acuity (VA) data demonstrated that most patients presented with poor VAs that did not significantly improve after treatment ( Figure 1 , P =.4284 pre- vs post-treatment). The overall median presenting LogMAR best corrected visual acuity was 2.60 (IQR: 1.00-3.00), which improved to 2.30 (IQR: 0.56-3.00) at the final visit ( Table 3 ). While presenting vision was similar between institutions (MEE: 2.60 [IQR: 0.74-3.00] vs UBC: 2.60 [IQR: 2.30-2.60]; P =.559), patients treated at MEE showed a trend toward better final outcomes with a logMAR median of 1.25 (IQR: 0.49-2.60) compared to 2.60 (IQR: 0.83-4.00) at UBC ( P =.060). Fourteen cases (20.0%) were lost to follow up defined as lack of attendance at consecutive scheduled follow-up appointments within 30 days of the initial presentation. Of the patients who achieved clinical resolution (n = 37 out of 47 with follow up information) the median time to clinical resolution was 41 days (range 4-171 days). Ten cases (14.2%) did not achieve any clinical resolution in the 12 months following the initial presentation (Table S1).

Only gold members can continue reading. Log In or Register to continue

Share this:

• X
• Facebook

Related

Related posts:

1. Microperimetry as an Outcome Measure Improves Patient Eligibility and Efficacy Detection in RPGR Gene Therapy Trials
2. Posterior Segment Risk Factors for Penetrating Keratoplasty Failure
3. Clinical Utility of Ultra-Widefield Swept-Source OCT for Intraocular Tumors: Comparison With Ultrasonography, SD-OCT, and MRI
4. Comment on: Visual Field Progression in Patients With Glaucoma With Delayed Follow-Up

Stay updated, free articles. Join our Telegram channel

Join