ARTHRITIS & RHEUMATISM Vol. 48, No. 4, April 2003, pp 1102–1108 DOI 10.1002/art.10861 © 2003, American College of Rheumatology Changes in Bone Mineral Density Following Discontinuation or Continuation of Alendronate Therapy in Glucocorticoid-Treated Patients A Retrospective, Observational Study Ronald Emkey, 1 Pierre D. Delmas, 2 Stefan Goemaere, 3 Uri A. Liberman, 4 Patrice E. Poubelle, 5 Anastasia G. Daifotis, 6 Nadia Verbruggen, 6 Antonio Lombardi, 6 and Michelle Czachur 6 Objective. To evaluate the effects of discontinuing or continuing alendronate (ALN) therapy on bone min- eral density (BMD) after patients on a long-term regi- men of glucocorticoids (GCs) completed a 1-year treat- ment period with ALN. Methods. Eligible patients were individuals with GC-induced osteoporosis who had received ALN (5 or 10 mg) for 1 year in a prior clinical trial and, at the end of the year, were still taking GCs at an average daily dose of >7.5 mg of prednisone or equivalent. Patients were contacted 3–5 years after completion of the prior ALN trial for followup measurements of the lumbar spine BMD and hip BMD, and retrospective information was collected about serious or drug-related adverse experi- ences and concomitant medication use. Some patients remained on GCs, and some remained on ALN, either alone or in combination with other drugs. The primary response parameter was the percentage change in lum- bar spine BMD from the end of year 1 to the followup visit. Change in BMD at the hip was a secondary response parameter. Results. Ninety (49.2%) of the eligible 183 pa- tients participated in the retrospective study. The fol- lowup period, which began at the end of year 1 of the original clinical trial, ranged from 3.3 years to 4.6 years. The mean number of days of treatment with ALN was 507. Fifty patients were included in the analysis because they had received supraphysiologic doses of GCs (doses above the lowest tertile of GC use for the study popula- tion; that is, higher than 6 mg/day), and they had not taken (defined as <6 months of use) other bone- affecting agents except ALN. Eleven of the 50 patients discontinued taking ALN (duration of use <90 days), 8 took ALN between 90 days and 300 days, and 31 continued to take ALN for >300 days after year 1 of the clinical trial. GC users who discontinued treatment with ALN (<90 days of therapy) had numerically greater decreases in BMD at the lumbar spine, femoral neck, and total hip from the end of year 1 (mean change 5.1%, 9.2%, and 6.6%, respectively), compared with patients who continued to take ALN for >300 days (mean change 0.1%, 0.9%, and 1.8%, respectively). Conclusion. Substantial loss of BMD in the lum- bar spine and hip was seen in patients who discontinued treatment with ALN but who continued to take >6 mg/day of GCs. However, patients receiving GCs who remained on the ALN regimen appeared to benefit from continued ALN treatment, since BMD was maintained in this latter group. Glucocorticoid (GC)–induced osteoporosis (GIOP) is the most common form of iatrogenic osteo- porosis (1,2). It is the most debilitating side effect of GC treatment because of osteoporosis-associated fractures (especially those at the vertebrae, hip, and rib), which Supported by a grant from Merck & Co., Inc. 1 Ronald Emkey, MD: Radiant Research, Wyomissing, Penn- sylvania; 2 Pierre D. Delmas, MD, PhD: Hopital Edouard Herriot, Lyon, France; 3 Stefan Goemaere, MD: Ghent University Hospital, Ghent, Belgium; 4 Uri A. Liberman, MD, PhD: Tel Aviv University, Petach Tikva, Israel; 5 Patrice E. Poubelle, MD: Le Centre Hospitalier Universitaire de Que ´bec, Quebec, Canada; 6 Anastasia G. Daifotis, MD, Nadia Verbruggen, MSc, Antonio Lombardi, MD, Michelle Czachur, MPH: Merck, Rahway, New Jersey. Address correspondence and reprint requests to Ronald Emkey, MD, Radiant Research, 1235 Penn Avenue, Suite 200, Wyo- missing, PA 19610. E-mail: ronaldemkey@radiantresearch.com. Submitted for publication April 26, 2002; accepted in revised form December 16, 2002. 1102