The results of this study showed no significant differences in the distribution of causative isolates between the two groups. The PS group had a higher proportion of patients with previous OSD and poorer initial clinical characteristics, including a poorer initial BCVA and larger EDS, than the NPS group. The PS group experienced worse disease progression, including delayed epithelial healing and more surgical treatments, than the NPS group. In the PS group, the long-term group had a significantly higher proportion of patients with a previous OSD and larger epithelial defects than the short-term group, and the strong steroid group showed a relatively delayed hospital visit and epithelial healing time compared with the weak steroid group.
In this study, 17.5% of patients used topical steroids before their first visit to our hospital. In several reported studies, the proportion of prior steroid exposure in infectious keratitis has been reported to vary, with 3.1% of bacterial keratitis cases in Paris, France25; 7% of infectious keratitis cases in Japan5; and 41.3% of contact lens keratitis cases in Singapore26. In addition, a study of keratitis including acanthamoeba conducted in Nottingham, UK, reported that chronic steroid use before diagnosis was observed in 40% of cases27. Fluorometholone (55.9%) was the most commonly used topical steroid in the present study. In a Japanese study, betamethasone was the most commonly used drug (64.2%)5. The steroid use may vary depending on the country, region, and personal preference of the hospital medical staff in which the study was conducted.
Pseudomonas spp. were the most common in the PS group. Other studies have reported varying results regarding the frequently isolated microorganisms in steroid-treated groups. In one study of contact lens bacterial keratitis, P. aeruginosa was reported to be the most common pathogen in the steroid and non-steroid groups26. A Japanese study reported that S. epidermidis is the most common bacterium in steroid-exposed eyes5. Previous experimental studies have reported that certain strains proliferate better with steroid use4,20,28,29. One animal experimental study reported that steroids enhance the stromal growth of some bacteria, such as Pseudomonas spp., but not Staphylococcus spp. and Streptococcus spp29.
In this study, previous OSD was significantly more common in the PS group (41.2%) than in the NPS group (23.8%). In addition, the proportion of patients with previous OSD was 70.6% in the long-term group, which was significantly higher than that in the short-term (23.5%) and NPS (23.8%) groups. These findings may be related to the prior long-term use of steroids to treat underlying ocular surface conditions. Other studies have also reported a more frequent history of OSD in the PS group19,25,27. In the PS group, herpetic keratitis was the most common, accounting for 33.3% (4/14) of previous OSD. Similarly, Otri et al. reported that herpetic keratitis (42.6%) was the most common OSD in patients with sight-threatening infectious keratitis27. In bullous keratopathy, topical steroid use has been reported as a significant risk factor for the development of ulcerative keratitis (odds ratio 2.63, 95% confidence interval = 1.41–4.91)30 Notably, corticosteroids increase the risk of infection in eyes with OSD by suppressing innate corneal epithelial defense30,31,32,33.
Poor initial clinical characteristics, including poor initial BCVA (< 0.1, Snellen) and EDS of ≥ 5 mm2, were more common in the PS group than in the NPS group, which is consistent with the similar studies. Wong et al. also reported that the median diameter of epithelial defect in steroid-treated eyes was significantly larger (3.5 mm vs. 2.0 mm, p = 0.031)34. Wang et al. reported that patients with prior steroid use had a 7.7-fold increased risk of developing larger ulcers (p = 0.033)26. In general, topical steroid use may reduce the initial inflammatory response in infectious keratitis. However, if antibiotic treatment is not sufficiently administered simultaneously in early keratitis or if topical steroids are stopped after the infection is recognized, it can lead to rapid inflammation and disease progression. These mechanisms are thought to contribute to worse initial clinical characteristics in the PS group. In addition, a higher proportion of patients with previous OSD in the PS group than in the NPS group may have contributed to the higher proportion of patients with poor initial clinical characteristics in the PS group.
In terms of treatment outcomes, EHT > 14 days was more common in the PS group (55.9% vs. 37.3%, p = 0.045). Delayed epithelial healing has been similarly reported in previous studies evaluating the effectiveness of a combination of antibiotics and topical steroids in infectious keratitis9,35,36. In one prospective study, randomized administration of topical steroids in culture-positive bacterial keratitis significantly delayed re-epithelialization in the steroid-treated group compared with that in the placebo group9. Chung et al. reported that topical steroid treatments had a negative effect on keratitis by exacerbating infection due to severe inflammatory side effects and also had an effect on microbial replication and delayed epithelial regeneration35. In an experimental study using rabbits, the rate of corneal epithelial regeneration was significantly slower in the steroid-administered group than in the control group36. In addition to delayed epithelial healing, the PS group showed relatively poor clinical features and treatment outcomes, which may be related to the high prevalence of previous OSD in the PS group. This is because in the case of previous OSD, poor visual acuity and poor treatment results may be more common due to pre-existing corneal lesions and opacities.
Our study showed that prior topical steroid use (Z = 2.46) was a significant risk factor for surgical intervention. This may be related to the relatively high incidence of delayed epithelial healing and impending corneal perforation in the prior topical steroid use group. Many studies have reported an increased risk of subsequent complications in bacterial keratitis associated with previous topical steroid use, including indolent ulceration17, perforation14,18, endophthalmitis37, and treatment failure.15,20 Coster et al. reported a higher proportion of successful treatment outcomes in patients without steroid use (78%) than in those who received steroids (69%)38. In addition, other significant risk factors for surgical intervention in the total study population were deep infiltration, diabetes, old age, and large EDS. These findings are similar to those of many other studies of bacterial keratitis that reported that the severity of the initial clinical findings and an older age were important risk factors for poor treatment outcomes27,39.
The risks of chronic topical corticosteroid use in ophthalmology, such as glaucoma and cataract development, are well known40. However, studies on the risk of long-term steroid use compared with short-term use in bacterial keratitis remain lacking. This study found that long-term steroid use was a significant risk factor for surgical intervention in the total group. Further, no significant differences in the treatment results between the long- and short-term groups, except for OSD and epithelial defect size, were observed. Interestingly, the short-term group accounted for half of the PS group. In the short–term group, most patients were referred to our hospital because steroids were used after the onset of keratitis symptoms, and the infection worsened afterwards. Therefore, physicians should be cautious in the early use of steroids for keratitis.
To our knowledge, few studies have reported clinical differences in infectious keratitis according to the potency of the topical steroid used. Of total groups, the strong steroid group was one of the significant risk factors for surgical intervention. No significant difference between the strong and weak steroid groups, except for delayed hospital visit and delayed epithelial healing time, was observed. These slight differences may be related to the stronger anti-inflammatory actions of steroids with greater potency. However, there are some limitations in interpreting the results according to the potency of steroids because the daily doses of all the steroids used in this study could not be investigated.
Topical steroids reduce inflammation in many ocular diseases. In the case of infectious keratitis, the use of topical steroids is contraindicated in fungal and acanthamoeba keratitis5,19,41. For bacterial keratitis, there have been discussions and literature reports on the therapeutic effect of topical steroids as adjuvants7,42. In clinical practice, determining the cause of infection until culture results are obtained is not always possible, and topical steroids can cause and exacerbate infections. Therefore, the perspective of this study, which focused on the disadvantages rather than the advantages of topical steroids, remains important for physicians.
This study had several limitations. First, this was a retrospective study of patients at a single tertiary referral hospital in South Korea. In addition, the difference in the ratio of the PS and NPS groups was not adjusted through a case–control match because the purpose of study was to present all consecutive data during the study period. Second, as this study analyzed only inpatients, it included patients with relatively more severe symptoms and signs. Therefore, the results of this study cannot be generalized to other population groups. Third, a detailed analysis according to various previous OSD etiologies could not be performed because of the smaller number of patients in each group and individual differences. Despite these limitations, this study highlights the risk of corneal infection exacerbation and the side effects of prior topical steroid use in clinical practice. In the future, a multicenter study can be considered to obtain more reliable results using a case–control matching method in larger dataset.
In this study, no significant difference in the distribution of causative isolates between the PS and NPS groups was observed. The PS group had a higher proportion of patients with previous OSD, worse initial clinical characteristics, a longer epithelial healing time, and more surgical treatments than the NPS group. Considering that topical corticosteroid used can be controlled by a physician, close monitoring with frequent regular follow-ups is necessary when administering topical steroid therapy for ocular surface diseases.