STRENGTHS AND LIMITATIONS OF THIS STUDY
-
This study examined the epidemiological spectrum of patients with uncommon infectious keratitis over a span of approximately 10 years at a single-referral hospital situated on the largest island in Korea.
-
Comprehensive data on all culture-positive infectious keratitis cases were analysed for antibiotic resistance and treatment success.
-
Cases involving relocation to different regions for treatment were excluded, and the interpretation of results should take into account the geographical characteristics of island regions.
-
The study’s retrospective design and reliance on chart reviews may limit the generalisability of the findings.
Introduction
Infectious keratitis is a sight-threatening condition characterised by corneal infiltration with microorganisms such as bacteria, fungi and protozoa.1 Early identification of causative organisms and prompt treatment are crucial for a favourable outcome. Empirical broad-spectrum antibiotics, including fluoroquinolones and fortified antimicrobials, are commonly used as initial therapy until culture results become available. Fluoroquinolones, effective broad-spectrum monotherapy antibiotics, provide broad coverage against both Gram-positive and Gram-negative pathogens and are known for their good ocular penetration, safety and stability.2 The increasing resistance to fluoroquinolones poses a significant challenge in the management of infectious keratitis worldwide. The dynamic nature of microbial susceptibility gradually diminishes the efficacy of this class of antibiotics.3 In 2020, studies indicated that Staphylococcus aureus and coagulase-negative staphylococci, the predominant Gram-positive pathogens, exhibited resistance rates exceeding 30% for fluoroquinolones and methicillin, with similarly high rates of multidrug resistance.4
Therefore, accurate antibiotic susceptibility testing is essential for selecting the appropriate fluoroquinolone, highlighting the importance of a proactive approach to addressing evolving microbial resistance. Additionally, data on epidemiological and demographic prevalence in a particular region can provide valuable information to assess the initial diagnosis and treatment options for infectious keratitis.
Jeju Island, the largest island in South Korea with a population of approximately 680 000, boasts unique geographical and climatic features. Located in the southernmost part of the country, Jeju Island has the highest annual temperature, precipitation, and relative humidity in Korea, as well as the lowest annual sunshine hours due to frequent rainfall.5 These distinctive environmental factors may contribute to the proliferation of specific causative microorganisms in infectious keratitis cases on the island. Additionally, Jeju Island is known for its extensive mandarin agriculture industry, potentially increasing the risk of ocular trauma among farm workers, particularly during the winter harvesting season when branches and thorns can inadvertently puncture the cornea.
Despite its sizeable population, there have been no epidemiological reports of infectious keratitis on Jeju Island thus far. The aim of this study is to investigate how the unique climate of Jeju, distinct from the mainland of South Korea, influences the strains and clinical characteristics of infectious keratitis. As an isolated island region, we seek to explore whether antibiotic susceptibility differs in Jeju compared with the mainland or other countries with similar climates. The goal is to provide reference data on strains and resistance that should be prioritised when treating infectious keratitis patients in Jeju, contributing valuable insights for clinical considerations.
Methods
This retrospective chart review study included patients clinically diagnosed with infectious keratitis who presented to the Cornea Clinic at Jeju National University Hospital’s Cornea Clinic between January 2011 and December 2022. Patients were excluded if they were transferred to another hospital, lost follow-up or resided outside of Jeju Island. Data collection involved a comprehensive review of medical records, including demographic information (age, sex and occupation), clinical characteristics (underlying diseases, previous ocular surgery, seasonal variation and triggering factors) and treatment outcomes (initial and final best-corrected visual acuity (BCVA) and duration of treatment).
Slit-lamp biomicroscopic examinations were performed to assess the presence of hypopyons and the size, location and depth of the corneal lesions. The lesion’s diameter at its longest point and location were assessed using ImageJ V.1.4 (National Institutes of Health, Bethesda, Maryland, USA) from anterior segment photography. The lesion location was divided into central (a 3-mm diameter zone of the cornea) and peripheral (five regions: superior, inferior, nasal, temporal and scleral). Corneal scrapings were obtained under topical anaesthesia by a single corneal specialist using a sterile No. 15 Bard-Parker blade. Microbiological investigations, including Gram staining, potassium hydroxide mount and cultures on blood, chocolate, MacConkey and Sabouraud dextrose agar plates, were performed for all patients at the initial presentation. An antibiotic susceptibility test was conducted for the identified pathogens.
For the treatment of bacterial keratitis, fortified topical antibiotics consisting of 2.5% vancomycin and 5% ceftazidime were administered hourly. As clinical improvement was observed, the dosing frequency was gradually tapered while continuing antibiotic therapy, irrespective of the bacterial susceptibility results. In cases where clinical symptoms were exacerbated, the antibiotic regimen was modified after 3 days. When culture results were positive, antibiotics were adjusted according to the antibiotic susceptibility testing. However, if cultures remained negative, the empirical treatment guided by clinical manifestations was continued. When the clinical condition of the patient’s cornea or the results of culture tests indicated fungal keratitis, we promptly reduced the frequency of fortified antibacterial eyedrops and conducted frequent slit lamp examinations to monitor the cornea at short intervals. The use of antifungals was based on the principle of gradually reducing the dosage according to the clinical course, with a guideline of using them for more than 8 weeks. The isolation techniques were mostly performed using automated methods, including PCR techniques.
The treatment was considered a medical treatment failure when one of the following happened: (1) the final BCVA deteriorated more than 1 by logMAR (logarithm of the minimum angle of resolution) due to a corneal scar, thinning or astigmatism and (2) surgical interventions, including therapeutic penetrating keratoplasty or amniotic membrane transplantation, were required due to uncontrolled infection and inflammation manifesting as increased epithelial defects, anterior chamber hypopyon or perforation requiring surgical intervention. No treatment other than surgical intervention, such as the application of cyanoacrylate glue, was carried out.
For statistical analysis, continuous variables were compared using Fisher’s exact test, χ2 test and one-way analysis of variance. Variables with a p value of less than 0.05 in the univariate logistic regression analysis were used to perform the multivariate logistic regression analysis, and a p value of less than 0.05 was considered statistically significant.
Patient and public involvement
Neither patients nor the public were involved in the development and design of this study, the selection of outcome measures, or the conduct, analysis and data dissemination of the study. This decision was guided by the need for impartiality, ethical considerations and the nature of the research, ensuring the study’s scientific integrity.
Results
Among the 110 patients with culture-positive results, 69 (62.7%) had bacterial keratitis, 32 (29.1%) had fungal keratitis, and 9 (8.2%) had coinfection with both bacteria and fungi. The demographics of culture-proven cases are given in table 1. The mean age of patients with fungal keratitis (71.56±7.88 years) was higher than that of patients with bacterial keratitis (59.05±22.48 years), although this difference was not statistically significant (p=0.113).
Demographic characteristics of patients with culture-proven infectious keratitis
Clinical characteristics, including location and size of the corneal lesion, the presence of hypopyon, previous ocular surgery and treatment duration for infectious keratitis, did not significantly differ between the groups (table 2). In 45 (40.9%) of the 110 patients, a history of recent corneal injury was documented. The predominant cause of trauma was vegetable matter (n=25, 22.7%). Additional traumatic factors included contact lens-related injuries (n=6), chemical material exposure (n=4), stone-related injuries (n=3), finger-related injuries (n=3), metal-related injuries (n=2), plastic-related injury (n=1) and pencil-related injury (n=1). A history of previous ocular surgeries predisposing to infectious keratitis was noted in 25 patients (22.0%) out of the 110. These surgeries included cataract surgery (n=9), pterygium removal (n=4), refractive surgery (n=4), pars plana vitrectomy (n=2), amniotic membrane transplantation (n=2), penetrating keratoplasty (n=1), trabeculectomy (n=1) and corneal laceration repair (n=1). Six patients (5.5%) presented with pre-existing ocular conditions predisposing to infectious keratitis, comprising herpes keratitis (n=2), bullous keratopathy (n=2), scleritis (n=1) and recurrent corneal erosion (n=1).
Clinical characteristics of patients with culture-proven infectious keratitis
The most commonly isolated bacterial pathogen was Pseudomonas species (12 cases), followed by Staphylococcus epidermidis (eight cases), S. aureus (seven cases), Moraxella species (six cases) and other Streptococcus species (six cases). The total treatment success rate for bacterial keratitis was 67.5%, while only 56.7% of fungal keratitis cases were successfully treated. Fusarium species had the highest incidence of eight cases of fungal keratitis, while Candida species were seen in six cases and Aspergillus species were found in only three cases. In cases of treatment failure, 27 patients with bacterial keratitis and 21 patients with fungal keratitis were included (table 3).
Prevalence of organisms in infectious keratitis and rate of treatment success
Microorganisms that were frequently isolated during the early (2011–2016) and late periods (2017–2022) are presented in table 4.
Most frequently isolated microorganisms from culture-proven infectious keratitis
S. aureus was resistant to oxacillin, ciprofloxacin and vancomycin at 40%, 60%, and 0%, respectively, in the early periods (2011–2016) and 50%, 60% and 0%, in the late periods (2017–2022). S. epidermidis was resistant to oxacillin, ciprofloxacin and vancomycin at 75%, 100% and 0%, respectively, in the early periods and at 75%, 50% and 0%, respectively, in the late periods.
The overall rate of methicillin-resistant Staphylococcus species (MRS) was 60%, and the rate of MRS increased from 55% (2011–2016) to 66% (2017–2022) during the study period. There was no vancomycin-resistant Staphylococcus.
Treatment failure by age, sex, diabetes mellitus (DM) status, previous history of ocular trauma or surgery, size, location and number of corneal lesions, initial BCVA, the presence of hypopyon and causative organism were investigated for prognostic factors. Age, initial BCVA, size and location of the corneal lesion and hypopyon were significant factors based on univariate analyses (p<0.05). Multivariate analyses were also performed for these factors, and the size of the corneal lesion (OR=4.03 for >3 mm; p=0.018) was found to be a significant factor for treatment failure (table 5).
Prognostic factors for treatment failures in infectious keratitis (multivariate analysis)
Discussion
This study provides valuable insights into the epidemiological characteristics and clinical outcomes of infectious keratitis on Jeju Island, South Korea, over a 12-year period. The mean age of patients with culture-proven keratitis in this study was higher than that reported in previous studies. Hahn et al
6 reported that 33.8% of patients with infectious keratitis in Korea were aged over 60 years, and Bharathi et al
7 found that 33% of patients were aged over 50 years. In particular, older patients and female patients had higher rates of fungal keratitis than bacterial keratitis. According to a report by Wong et al
8 that analysed the risk factors for infectious keratitis, trauma was significantly associated with fungal keratitis. This finding may be attributed to the unique age structure of Jeju Island, where older females actively participate in agricultural activities, potentially increasing their risk of exposure to causative microorganisms.
The mean initial BCVA we found was worse in bacterial keratitis (0.88±0.61 logMAR) than in fungal keratitis (0.56±0.37 logMAR). However, the final BCVA was worse in fungal keratitis (0.45±0.35 logMAR) than in bacterial keratitis (0.44±0.43 logMAR). The poor visual prognosis associated with fungal keratitis in this study may be related to the advanced age of the affected patients and the delayed initiation of targeted antifungal therapy. Prompt recognition of fungal aetiology and early administration of appropriate antifungal agents are crucial for improving treatment outcomes in these cases.
In this study, the positive culture rate of bacterial keratitis was approximately 45%, similar to previous reports. However, the composition of the most commonly identified bacteria showed slight differences compared with previous studies.9–11 The most commonly isolated bacterial keratitis pathogen was Pseudomonas species (12 cases). Park et al
1 reported that S. epidermidis was the most common microorganism isolated from bacterial keratitis, though other studies have reported that the Pseudomonas species was the most common Gram-negative isolate, comprising 9.3%–22.4% of total culture-proven pathogens.12–15 In the present study, Pseudomonas species consist of 10.9%, squarely within the range of percentages reported by others. The Moraxella species, which was the second most common pathogen in Gram-negative bacteria found in our study, is known to grow well in a moist and warm atmosphere.16 17 Kim et al
18 reported that Moraxella species were identified as the causative pathogen in only 0.3% of 689 cases of infectious keratitis in the Korean peninsula area of temperate climate, which is much lower than the rates reported in our study and is potentially reflective of the subtropical climate of Jeju Island.
While Stenotrophomonas maltophilia was detected in two cases in our study, this pathogen is very rare in ocular infections. It is also difficult to treat. A significant risk factor for infection with S. maltophilia is previous ocular surgery, such as cataract surgery and penetrating keratoplasty.19 There are reports that S. maltophilia may also be resistant to ciprofloxacin and ceftazidime.20 The patient infected by S. maltophilia in the present study was treated with topical ciprofloxacin and ceftazidime, which failed, and he also had a history of a previous cataract operation. Besides, it was remarkable that there was one Burkholderia cepacia keratitis case, which is very rare and previously known as a subclass of Pseudomonas species.21 22 The patient was successfully treated with fortified 5% ceftazidime eyedrops.
MRS is increasingly prevalent with the widespread use of antibiotics worldwide. The incidence of MRS is approximately 60%, with 33.3%–53.5% of infections resulting in MRS epidermidis in Japan and Korea, respectively.23–25 In the present study, MRS rates were 54% from 2010 to 2017, while the rate of MRS increased from 55.5% (2011–2016) to 66% (2017–2022) during this same period, indicating that the antibiotic usage practices of patients with keratitis might be contributing to increasing resistance on Jeju Island. We believe that the isolation of this population on an island also had an impact on these results. It is also remarkable that there was no vancomycin-resistant Staphylococcus and that the resistance to ciprofloxacin was as high as 60%, demonstrating that the widespread use of fluoroquinolone for the treatment of keratitis resulted in high ciprofloxacin resistance rates. The fluoroquinolone class of antibiotics used as first-line monotherapy is widely used due to its low ocular toxicity, good corneal permeability, convenience of commercialisation and stability at room temperature, but resistance has recently become an issue due to overuse.26 27 Alexandrakis et al
28 reported increasing resistance of S. aureus to fluoroquinolone-class antibiotics. Soleimani et al
29 reported that the odds of ciprofloxacin and levofloxacin resistance increased by 1.25 and 1.15 for each 1-year increase in a 6-year infectious keratitis analysis study.
The Fusarium species was most prevalent (21.6%) among cases of fungal keratitis in the present study, a finding consistent with reports by Hahn et al and Kim et al.6 18 However, Bhartiya et al
30 found that the Candida species was most common, and Laspina et al
12 reported that the Acremonium species was most frequently isolated in these cases. This suggests that the prevalence of various pathogens differs according to the population epidemiology and geography of a region. For example, the Fusarium species, which is relatively common in tropical and subtropical areas, is known to be associated with trauma.31 The high frequency of the Fusarium species reported here appears to be affected by increased levels of agriculture-related trauma in our population and Jeju Island’s subtropical climate. The 50% treatment success rate in Fusarium species infections contributed to the overall low treatment success rate of 56.7% for fungal keratitis cases, reflecting the high frequency of Fusarium species infections in our study. Another study of the treatment of keratitis and endophthalmitis caused by Fusarium species31 also reported a low treatment success rate (23%) in three patients who had undergone enucleation and seven who had received keratoplasty out of a total of 13 patients. Considering the low treatment success rate of Fusarium infections, an accurate diagnosis of Fusarium is crucial. Fusarium shares many morphological similarities with Aspergillus, making it challenging to differentiate between the two. In such cases, assessing treatment outcomes and employing in vivo confocal microscopy can be valuable for distinguishing Fusarium from Aspergillus.32
Furthermore, a distinctive aspect of our study is the relatively high incidence of Colletotrichum and Acremonium fungal keratitis, considered relatively rare in fungal keratitis in Korea. In a prior study investigating the prevalence of fungal keratitis reported in Korea, Fusarium and Aspergillus fungi were most frequently identified, followed by Altanaria fungi.33 34 Colletotrichum fungal keratitis was rarely identified in Korea, according to previous studies.33 However, in our research, Colletotrichum fungal keratitis in the Jeju region, particularly identified between 2017 and 2022, corresponding to the latter half of our study, showed one of the highest frequencies among identified fungal keratitis cases. This demonstrates a notable increase in frequency in the Jeju region over the past 6 years, marking a difference from the distribution of fungal keratitis found in Korea.34 Colletotrichum fungi, belonging to the pigmented hyphae group among fungi, are relatively common in tropical and subtropical regions, suggesting an association with Jeju’s regional characteristics.35–37 Therefore, when encountering patients with fungal keratitis in Jeju, it is crucial to consider the possibility of diagnosing Colletotrichum as one of the causative agents.
This study has several limitations that should be considered when interpreting the results. First, the retrospective design and reliance on medical record review may have introduced potential biases and limited the availability of certain data points. Second, as this study was conducted at a single tertiary referral centre on Jeju Island, the findings may not be generalisable to other regions with different demographic and environmental characteristics.
Despite these limitations, this study provides valuable epidemiological data and highlights the unique features of infectious keratitis on Jeju Island. The findings emphasise the importance of considering regional variations in causative microorganisms and antibiotic resistance patterns when managing infectious keratitis. Further prospective, multicentre studies are warranted to validate these results and develop targeted prevention and treatment strategies for infectious keratitis in this distinct geographical setting.
Conclusion
The incidence of Moraxella and Colletotrichum species in our study was higher than those reported in other districts with different climates and environments. The results reported here reflect the unique environmental features of Jeju Island, characterised by high humidity and temperatures. Continuous patient education and tailored treatment protocols based on local epidemiological data may help enhance the therapeutic outcomes for infectious keratitis on Jeju Island and potentially beyond.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants, was approved by the Jeju National University Hospital Institutional Review Board (IRB No. 2017-06-033-001) and was performed in accordance with the Declaration of Helsinki. Informed consent was waived by the IRB due to the retrospective nature of the study.
This post was originally published on https://bmjopen.bmj.com