We read with interest the article “Rituximab in the Treatment of Refractory Noninfectious Scleritis” by Cao and associates. In conjunction with our recently published experience with rituximab (RTX) on scleritis secondary to granulomatosis with polyangiitis (GPA), in which we observed similar results in 8 patients, we think it valuably adds to RTX use in such conditions, especially as per their primary outcome measure. Nonetheless, it is common practice to observe glucocorticoid (GC) dependence, with short-term independence of these drugs. Therefore, their goal seems ambitious and we would inquire if consideration was given to maintaining low (≤10 mg) prednisone dose, as in other studies, instead of requiring maintenance with other immunosuppressants, as was the case in 13 of 15 patients, some prescribed in combination, with the use of RTX and cyclophosphamide in 2 patients, who, interestingly, were the only lost to follow-up. Linked to this, the authors did not state the mean or median doses of prednisone the patients received at time of RTX treatment.

A very important issue relates to RTX side effects. Though we agree that this biologic is also effective as maintenance treatment, serious consideration must be given to its potential in fostering systemic infections, cancer development, hematologic events, and, most importantly, hypogammaglobulinemia, an issue not explicitly mentioned but probably present in 5 of 15 patients who were treated with intravenous immunoglobulin (IVIG). This also relates to the continued cost of not only RTX but also IVIG, which contradicts their important statement regarding access to these medications, especially if RTX is considered as maintenance treatment—the doses which with that aim are unspecified in their article—and especially if not uniform and including high doses. Indeed, there are patients who have achieved remission and maintained response with lower than the “rheumatologic” protocol doses, an issue that would deserve inclusion not only of high but also of low RTX dosages in appropriately prospective longitudinal trials, as they pertinently stated. Such studies would unravel which dose has the best safety and cost/effectiveness profiles, and also provide a chance to explore definite GC tapering schemes plus adequate tools for evaluation of activity, severity, and search for early indicators of a refractory course or, on the other hand, of adequate treatment response in this disease.

It is interesting to mention that few studies have used the McCluskey activity scale for scleritis, certainly not those in patients with rheumatoid arthritis (RA) or GPA, and its validation could be also an aim to achieve. If such a study is ever to be done, it might provide support for RTX approval in noninfectious scleritis, just as it has occurred in RA and anti-neutrophil cytoplasmic antibody–associated vasculitis.

Finally, we emphasize the value of this article in supporting the use of RTX in refractory, noninfectious scleritis, either idiopathic or secondary to other autoimmune diseases.