In their recent article, Constantinou and associates reported clinical outcomes in patients who experienced infectious keratitis after penetrating keratoplasty. Their principal conclusion was that “prolonged use of corticosteroid eye drops is a major risk factor for the occurrence of postkeratoplasty infectious keratitis in failed and clear grafts.” Although there are many reasons to believe that topical corticosteroid use may predispose a patient to infection, the data presented in their article do not support such a conclusion. To evaluate the risk of corticosteroid use on the development of infectious keratitis in this setting, data regarding steroid use would be required not only from eyes that demonstrated infection, but also from those that did not. This would allow a comparison of infection rates in eyes exposed or not exposed to steroids. This case series only evaluated patients who experienced infection.
Topical steroid use was common in their study patients with clear grafts (76%) and was universal in those with failed grafts. However, low-dose topical corticosteroid use often is continued indefinitely after penetrating keratoplasty in an effort to prevent graft rejection. To evaluate whether the use of steroids may be a risk factor for infection, the corresponding proportion of patients who used steroids but did not have infection must be known. Even if this information were available, a retrospective study such as this can be prone to various confounding factors or biases that make it difficult to know the explanation for an observed association. To obtain a definitive answer to this question would require a prospective study design capable of identifying a causal relationship.
Although the authors have provided a useful description of the clinical outcome and microbiological profiles of infections occurring after penetrating keratoplasty, further work is needed to understand the potential role of corticosteroid use in contributing to these infections.