We read with interest the recent article by den Uyl et al (1), in which they reported that prednisolone (30 mg/day or 60 mg/day) did not affect glucose tolerance in patients with active rheumatoid arthritis (RA). We agree that this is a very important finding; the fast action of prednisone therapy improves clinical manifestations of RA, a result that was also recently found by Bakker et al in the Computer Assisted Management in Early Rheumatoid Arthritis trial II (2).
High doses of intravenous methylprednisolone (IVMP) pulse therapy (1,000 mg/day for 3–5 days) are commonly administered to treat severe manifestations of either systemic lupus erythematosus or systemic vasculitis (3). The principal rationale for this treatment regimen is to achieve a higher therapeutic effect than that achieved with conventional oral steroid doses, while minimizing toxicity (4).
We performed an observational study of patients without diabetes who received IVMP at our hospital. Glucose measurements were obtained at baseline and 3 times per day in the fasting state (capillary glycemia) for the 3 days of treatment with IVMP. Parametric one-way analysis of variance (ANOVA) was performed using GraphPad Prism 4 software.
Initially we assessed 84 patients; 18 patients were excluded because they had diabetes or required treatment with insulin at admission. Thus, 66 patients were included in the study (44 women and 22 men; mean ± SD age 35.0 ± 14.2 years). Glucose levels after treatment were statistically significant compared to baseline levels as determined by one-way ANOVA followed by Tukey's post hoc test. The mean ± SD glucose level at baseline was 110.7 ± 19.9 mg/dl. All patients received IVMP at baseline; after 8 hours, the glucose level was 151.9 ± 53.3 (mean change 41.2; P < 0.001), and after 16 hours of IVMP therapy, the glucose level was 171.3 ± 46.4 mg/dl (mean change from baseline 60.7 ± 46.4 mg/dl; P < 0.001) (Figure 1). Twelve patients received insulin treatment after the first serum glucose measurement (8 hours). After 16 hours of IVMP, when glucose levels were higher, an additional 12 patients received treatment with insulin. In summary, after 16 hours of receiving IVMP, insulin treatment was required in >35% of our patients. Since the glycemia was modified after treatment with insulin, our analysis included only the first 24 hours when the majority of patients had not received insulin treatment (Figure 1).
We recognize that impaired glucose metabolism is probably related to inflammation in patients with RA and that the high doses of prednisone administered to patients in den Uyl and colleagues' study did not change glucose tolerance. However, we suggest that it is necessary to monitor glucose levels even in patients without diabetes who receive IVMP (5). Tamez Perez et al previously described the effects of IVMP on glucose profiles (6); however, our study demonstrated that serum glucose levels increased and insulin therapy was required after IVMP treatment. Even though the potential benefit of high-dose prednisolone in the treatment of patients with inflammatory diseases is evident, we suggest that glucose monitoring is necessary when high-dose glucocorticoids, particularly IVMP, are administered.