Infectious keratitis is a potentially blinding condition, therefore proper and rapid identification of the causative organism are imperative to guide the management to pursue better visual outcomes. Although early treated patients may successfully recover (Fig. 1), some infections are real challenges and may result in severe complications. (Fig. 2) The main pathogens found in these infections are gram-positive and gram-negative bacteria, fungi and Acanthamoeba. (Fig. 3)
The present study investigated the microbiological patterns of the infectious keratitis of patients admitted to our hospital. The positive rate found in these 8 years was 55.6%, which is in agreement with the rates reported by other authors in Brazil12, 13, Paraguay14, Mexico15, United States16, 17, Canada18, United Kingdom19, Spain4, Portugal20, Sweden21, Israel22, India23, Nepal24, Korea25, China26, Taiwan9 and Australia27 (32.6–89.6%).
The profile of etiologic pathogens worldwide depends on different risk factors related to specific characteristics of population and environment. Usual predisposing factors related to the individual are occupation, contact lens use, eyelid/ocular or systemic illness, previous ocular traumas or surgeries and use of topical steroids. Environmental-associated risk factors comprises climate, socioeconomic and virulence of the microbial1, 7, 28. Reports from North America, Europe, Australia and Singapore reveal the highest rates of bacterial etiology, regions in which the prevalence of contact lens wear is remarkable28. Mycotic keratitis is more frequently reported in developing and tropical countries, particularly in poor rural populations, in which agriculture-related trauma with soil or vegetable matter consists in predisposing factor of fungal ulcer30. Reviews from Paraguay14, India23, China18 and Nepal24 found 44.5–64.5% of keratitis with fungal etiology. Other studies from Canada18, USA16, 17, Mexico15, Brazil12, 13, United Kingdom19, Spain4, Sweden21, Portugal20, Israel22, Korea25 and Taiwan8 reported rates from 0.2–23.3% of mycotic keratits.
Bacterial isolates were most of the cases in this review, 86.8% of the positive isolates, which was similar to studies from Brazil12, 13, Korea25, Taiwan8, Israel22, Australia26 and countries from North America and Europe15–21 (76.7–100.0%). Polymicrobial infection represented 19.0% of the isolates, comparable to other series (0.7–26.8%)4, 8, 12, 17–19, 21, 22, 23–26.
The study revealed a predominance of Gram-positive in bacterial keratitis isolates. Cutibacterium acnes (12.2%) Staphylococcus epidermidis (6.6%), Streptococcus pneumoniae (6.6%) and Corynebacterium (6.2%) were the most common gram-positive bacteria. This high rate of Gram-positive pathogens was similar to other series4, 8, 12–14, 17–21, 24–27, however coagulase negative Staphylococcus4, 9, 13, 17–20, 22, 23, 28, 29, Staphylococcus epidermidis5, 13 and Streptococcus spp.24, 25 were the most common gram-positive isolated in the literature.
Sagerfors reported anaerobic diphtheroid rods (predominantly C.acnes) as the second main agents in a study performed in Sweden21. Tena found C. acnes and Corynebacterium as the second and third agent, respectively, in a study from Guadalajara4. A New Zealander study also reported C.acnes as the second most common gram-positive bacteria in infectious keratitis, after coagulase negative Staphylococcus29. In a Portuguese tertiary centre, the most common bacteria was Corynebacterium macginleyi20. Although C.acnes30 and Corynebacterium may be considered as contaminants, since they are part of eyelid and ocular surface microbiota, its role as pathogenic agents in corneal affections has been described31.
In our patients, Pseudomonas aeruginosa was the predominant gram-negative specie isolated, accounting for 9.9% of the positive samples, which was a similar result to rates in other centers (ranging 2.2–24.4%) 4, 8, 12–20, 22, 23–28.
Fusarium was the most common fungal isolate (6.2% of all isolates), also similar to other studies (ranging 2.6–39.5%)12–15, 18, 19, 23, 24, 26. Farrell reported Aspergillus and Candida as the fungi most found in a series of cases in Ireland32 and Oliveira-Ferreira20, Sagerfors21 and Tena4 found Candida as the first fungal etiology in Portugal, Sweden and Spain, respectively.
All gram-positive tested were sensitive to vancomycin; studies from Korea, Taiwan, United Kingdom, Spain, USA and Canada also described high rates of susceptibility to this antibiotic (99.4%-100%)4, 8, 17, 18, 19, 24. There was a high rate of gram-positive resistance to moxifloxacin (29.2%), higher than the range in other series (0–6.2%)24, 25.
The resistance to ciprofloxacin was found in 24.1% and 1.3% of the gram-positive and gram-negative tested, respectively. Other reports also showed higher ciprofloxacin resistance rates in gram-positive bacteria (10.3–45.0%)4, 17, 25 than in gram-negative (1.0–8.8%)4, 8, 18, 19, 26. The resistance to gentamicin was 16.7% and 2.5% in gram-positive and gram-negative, respectively, which was in accordance with other series findings, that revealed rates ranging from 12.1–19.0% for gram-positive4, 17, 19, 25 and rates ranging from 0–15.3% for gram-negative bacteria4, 8, 17, 18, 19, 25, 26.
The analysis of gram-negative resistance to amikacin and ceftazidime revealed rates of 1.3% for both antibiotics; other authors reported amikacin resistance rates ranging from 0–11.7%8, 17, 19, 25 and ceftazidime resistance rates ranging from 0–16.7%4, 8, 17, 18, 19, 25.
The method of obtaining the corneal scrapes has been described with many instruments, such as sterile cotton swab, sterile calcium alginate swab, heat-sterilized platinum (Kimura) spatula, 26 or 23-gauge needle, blade, cotton-tipped applicator2, 33, 34. In our study, we used sterile steel spatulas that were changed for each smear.
In 2020, due to the COVID-19 pandemic, the visits of patients at the Ophthalmology Emergency Room decreased compared to prior years. The number of corneal scrapes was significantly lower than in the other years studied; it happened because the total number of patients presenting infectious keratitis dropped and many patients that were referred to our hospital came with more severe cases of infections, in which the corneal scrape was not safe because of perforation or risk of perforation. Agarwal also reported a reduction in the eye-specific emergency department visits in India35.
Our study had some limitations: it was a retrospective review of microbiological data; we did not correlate the laboratory results with the clinical outcomes. Moreover, we could not exclude the possibility that some of our positive isolates could be a result of a contaminant germ or non-pathogenic agent, a part of the ocular surface microbiome of the patients.