Session X - Pelvis
*A Prospective Randomized Double-Blind Placebo-Controlled Trial of the Efficacy of Indomethacin as Prophylaxis for Heterotopic Ossification after Operative Treatment of Acetabular Fractures
Madhav A. Karunakar, MD1; Michael J. Bosse, MD2; Janette M. Hall, MS1; Stephen H. Sims, MD2; Theodore T. Le, MD1; James F. Kellam, MD2; James A. Goulet, MD1; Mark A. Freeborn, MD1;
Purpose: We prospectively evaluated the efficacy of indomethacin as prophylaxis for heterotopic ossification (HO) after operatively treated acetabular fractures.
Methods: An institutional review board-approved prospective double-blind placebo-controlled clinical trial was performed at two level I trauma centers to evaluate the efficacy of indomethacin as a prophylaxis for HO after the operative treatment of acetabular fractures. Between January 1, 1999 and May 31, 2003, 232 patients with acetabular fractures were treated operatively through a posterior approach. Patients with the following conditions were excluded from study participation: age of less than 18 years, spinal cord injury, ankylosing spondylitis, burns, gastrointestinal bleeding, Glasgow coma scale score of less than 12, cerebrovascular accident, pregnancy, and use of other nonsteroidal antiinflammatory drugs. A total of 157 eligible patients were identified and 125 patients were enrolled in the clinical trial; 107 were observed long enough to be included in data analysis. All patients underwent operative stabilization of their acetabular fractures by either a combined anterior and posterior approach or an isolated posterior Kocher-Langenbeck approach. After fixation and prior to wound closure, any necrotic gluteus minimus muscle was debrided to viable muscle. Sixty-one patients were randomized to the placebo group and 46 patients to the indomethacin-treatment group. Indomethacin 75 mg SR and the placebo were administered to the patients by the investigational drug pharmacy in a blinded fashion. The medication was taken once daily for 6 weeks. Patient compliance was measured by obtaining indomethacin serum levels at the first postoperative visit (2 weeks). The extent of HO was evaluated on plain radiographs (AP and Judet) at 3 months postoperatively. The radiographs were scored for the presence of HO by using the Brooker classification as modified by Moed. The data were analyzed two ways: 1) by excluding patients with protocol deviations and 2) by using an intent-to-treat model, where all enrolled subjects with 3-month Brooker scores were included in the analysis, regardless of whether they withdrew or were dropped from the study for clinical reasons. The sample size was estimated to produce a statistical power of 80% to detect a difference of 15% between the two treatment groups with alpha = 0.05.
Results: There were no significant differences with regard to age, sex, body mass index, injury severity score, and complications between the two treatment groups. The overall incidence of HO (Brooker I to IV) was 52.8%, and the overall incidence of significant HO (Brooker III to IV) was 19.6%. There were four patients with Brooker IV HO. There was no significant difference between the treatment groups in the incidence of HO according to Brooker class (P = 0.23). Significant HO (Brooker grades III and IV) occurred in 8 patients (17%) in the indomethacin group and 13 patients (21%) in the placebo group. There was no significant difference in the presence of moderate-to-severe HO (Brooker III and IV) between the two treatment groups (Fisher's exact test'P = 0.81). Eighty-two of 107 patients completed the protocol. Twenty-five patients did not complete the treatment protocol for the following reasons: stopped medication due to side effects, did not receive medication at discharge, lost medication, or medication stopped by another physician who did not understand the purpose of the study. Nine patients (8.4%) did not receive the full medication course, 16 patients (15%) were dropped or withdrew from the study for adverse events or gastrointestinal symptoms. Twelve patients dropped or withdrew from the indomethacin group and three from the placebo group. Forty percent of patients in the indomethacin group had nondetectable serum levels at 2 weeks. Complications identified in the indomethacin treatment group included deep venous thrombosis (5), wound infection (2), nonunion (1), gastrointestinal bleeding (1), and perforated ulcer (1). Complications identified in the placebo group included deep venous thrombosis (6) and wound infection (2).
Conclusion: In this prospective randomized study, a placebo provided as effective prophylaxis against the development of HO as indomethacin. More patients withdrew from the indomethacin group for gastrointestinal side effects or adverse events than in the placebo group. Patient compliance with indomethacin was poor, with 40% of patients having no detectable indomethacin serum level. Serious gastrointestinal complications (gastrointestinal bleeding and perforated ulcer) occurred in two patients treated with indomethacin.