This retrospective study included more cases than those in previous studies and investigated risk factors in a diverse group of pediatric steroid responders. Young age, male sex, primary disease, increase amount of WBC, and high steroid doses were discovered to be independently associated with SIOH.

Krag et al. hypothesized that young age is a risk factor for pediatric steroid responders [9]. Some studies have suggested a link between age and high IOP, but the relationship is still debated. According to Gupta et al., patients younger than 6 years old are more likely to develop SIOH [10]. Sugiyama et al. found that SIOH is more common in children under the age of 10 [6].

In this study, steroid responders were younger, with a cutoff age of 11.0 years, which is consistent with some previous findings. Previous studies have shown that glucocorticoids can alter the morphological and biological structure of the trabecular meshwork (TM) and inhibit aqueous humor outflow, resulting in increased IOP [1, 11]. It has also been suggested that the TM continues to grow until around the age of 8 [12], implying that steroids may have a greater impact on the immature TM in younger patients.

Rodrigo et al. investigated gender differences in SIOH using a SIG rat model, discovering that male rats had significantly higher IOP than female rats after a 24-week follow-up period [13]. A clinical retrospective analysis of 1783 cases of SIOH following photorefractive keratectomy identified male sex as a risk factor [14].

Research on TM length may shed some light on the link between the male sex and SIOH. Cho et al. measured the TM length in 26 responders and 78 non-responders and discovered that patients with SIOH had significantly shorter TM than non-responders [15]. Kasuga et al. also found that males had a shorter TM height than females, though the differences were not significant [16]. These findings suggest that males with shorter TMs may be more vulnerable to steroid exposure.

Primary disease was also extracted as one of the risk factors in relation to SIOH. Because there are too many explanatory variables, it is difficult to analyze each disease are included in GEE analysis. We checked for average steroid dose in each category. The amount of steroid use was 59.0, 66.8, 173.0, and 100.6 mg/kg for hematologic, renal, central nervous system, connective tissue disease, respectively, whereas it was 39.6, 72.3, 76.8, 33.7 mg/kg in the non-responder group. In the steroid responder group, steroid dosage is particularly high in central nervous system and connective tissue disease. Therefore, we speculate that these difference by category is probably due to steroid dosage.

The current study’s cutoff steroid dose was 40.7 mg/kg/month, or approximately 1.3 mg/kg/day, indicating a high dose. There is a discussion about the relationship between steroid dose and SIOH. Some studies indicate a dose-dependent relationship [1, 6], whereas others have found no causal link between high steroid doses and SIOH [17,18,19].

Previous research has focused on children with specific diseases [4, 6, 18]. To identify risk factors for steroid responders, we classified primary diseases requiring steroid treatment. However, our current study did not reveal a clear relationship between primary disease and responders. Only one previous study proposed a link between connective tissue disease and elevated IOP; thus, the role of primary disease in steroid responsiveness is assumed to be minor.

This study examined the difference in blood cell counts before and after steroid administration. The findings indicate that WBC counts may contribute to IOP elevations in children with ALL [5]. Increased hematocrit and blood viscosity are also risk factors for high IOP [20, 21]. In our study, increase the amount of WBC was extracted the risk factor of SIOH as other previous research. However, the cutoff value is unreliable due to low AUC, that is expected to be less important than other risk factors. We hypothesize that in hematologic diseases, blood cell counts frequently fluctuate rapidly during treatment, making it difficult to accurately assess count changes. Although this is just for reference, 18 (60.0%) responders and 15 (31.2%) non-responders had a WBC change greater than the cutoff value of 3.4 × 10²/µl.

There are several limitations to this study. First, we used rebound tonometry to check the IOP in each examination, although applanation tonometry is gold standard device to measure IOP. Because children sometimes refuse applanation tonometry, because of its invasiveness. Moreover, some children undergoing anticancer agent treatment are isolated in private rooms. In such cases, rebound tonometry can be very useful. Therefore, in this study, we decided to standardize the results of IOP measurements to rebound tonometry.

We limited the follow-up period to one month after starting steroid treatment. This is because some patients are discharged from the hospital months after receiving steroids. Subsequently, their ophthalmologic care is managed at their own clinic, making it difficult for us to follow up directly. Consequently, we could not determine whether the duration of steroid administration is a risk factor for steroid responders. To address this issue, a prospective multicenter study is recommended.

For our analysis, we converted all administered steroids to prednisone equivalents to account for differences between the various steroid types used. As a result, we did not investigate the association between the types of steroids used and steroid responders. Given that high-titer dexamethasone has been shown to significantly increase IOP [18, 22], it is also important to consider the type of steroids used.

Furthermore, systemic steroid administration may necessitate large infusion volumes, causing an increase in IOP due to fluid overload [23]. Therefore, the infusion loading volume should be considered for a more thorough analysis.

The definition of SIOH is not a consensus. Several studies divided SIOH into three categories, non-responder, intermediate- responder, high-responder. Because in this study there are few high-responder, and the main point in this study is to search for the risk factor of SIOH, so we decided to follow the papers that divide SIOH into two categories, responder or non-responder. If we have more participants, we will analyze using three categories.

In conclusion, in children receiving systemic steroid administration, age of < 11 years, male gender, primary disease, high increase in WBC during treatment, and high steroid doses may contribute to steroid-induced IOP elevation. Although our findings are reasonably consistent with clinical findings and previous reports, a more precise analysis will require prospective studies involving a large cohort of patients.