To the Editor:

We appreciate the interest and comments of Dr. Arashvand and Dr. Bertoli and their colleagues.

We agree with Dr. Arashvand that our study population was small and that further studies are needed to confirm that initial pulses with high-dose methylprednisolone are beneficial in the treatment of GCA. The study by Chevalet et al (1) has the strength of a larger patient population; however, the design was different from ours and the differences in outcomes are not surprising. Chevalet and colleagues' patients received a single dose of 250 mg methylprednisolone, while we gave 3 doses of ∼1,000 mg on consecutive days. Our earlier studies in a human mouse chimera model showed that such high doses are needed to suppress the T cell response in the artery (2).

We think it justified not to include the pulses in calculation of the cumulative dose because the biologic effects of a pulse are completely different from those of more long-term prednisone treatment. Pulse steroids have their own specific adverse effects, and our study population may have been too small to enable detection of them. Again, further studies are needed to assess the risk/benefit ratio of pulse treatment.

Dr. Bertoli and colleagues raise the question of why no differences in acute-phase response parameters were seen between the 2 study groups if there were more disease flares in the control group. The flares mostly occurred after 4 months of treatment, but were then relatively equally distributed over the subsequent observation period; i.e., the number of flares at each time was low, and they were immediately controlled by treatment adjustment. Our study was designed to detect treatment required to control disease while being insensitive to differences in clinical or laboratory disease activity markers.

In designing the study we also had carefully considered the safety concerns raised by Bertoli et al. The notion that patients were treated with low-dose steroids for a certain initial interval is an apparent misunderstanding of our statement that patients receiving low-dose prednisone (e.g., for treatment of preexisting polymyalgia rheumatica) were not excluded. All patients were immediately started on 40 mg of prednisone daily when diagnosed with GCA. It is correct that GCA can have devastating complications if left untreated; however, complications are rare after daily prednisone treatment has been started (alternate-day treatment may have higher complication rates and should not be used) (3, 4). The recommendation to start prednisone treatment at 1 mg/kg per day is not evidence based, and lower doses are sometimes used in clinical practice in patients who are less likely to tolerate higher doses. Given that all patients were very closely monitored, we believe that treatment with 40 mg of prednisone daily was safe. In retrospect, we were surprised by how well the control group did within the first 6 months of treatment and that more intensive treatment was necessary only to prevent the ongoing smoldering of the disease on followup—not during the initial control of clinical symptoms.

Jörg J. Goronzy MD*, Cornelia M. Weyand MD*, Gene G. Hunder MD†, Mehrdad Mazlumzadeh MD‡, * Emory University School of Medicine, Atlanta, GA, † Mayo Clinic Rochester, MN, ‡ Mayo Clinic, Scottsdale, AZ.