Follow-up of occult bone lesions detected at MR imaging: Systematic review

Purpose: To perform a systematic review of the literature regarding the natural course of posttraumatic occult bone lesions (often referred to as bone bruises) detected at magnetic resonance (MR) imaging. Materials and A systematic review of the literature was performed by imaging. Methods: A systematic review of the literature was performed by searching the MEDLINE database (from January 1966 to February 2003) with the keywords bone bruise. trauma, follow-up, and MRI. Keywords were linked by using the Boolean operator AND. Studies were included if all of the following criteria were fulfilled: patients sustained trauma, MR imaging was used as a diagnostic method, results of clinical or MR imaging follow-up were available, and study Was Written in English, Dutch.. German, French, Spanish, Italian, Swedish, Danish, or Norwegian. The quality of each study was assessed by using a standardized criteria set, and kappa statistics were estimated to rate the level of agreement between the two reviewers. Results were compared with regard to study design and quality scores. Results: The MEDLINE search identified 266 articles, 13 of which met the inclusion criteria. The quality of the included studies was moderate. The two reviewers initially agreed on 179 quality items (kappa = 0.84). The study population was generally small, and follow-up periods ranged from 1 to 73 months. Four different classification systems were used, and in two studies bone bruise was not specified. Study results suggest a generally good clinical prognosis of bone bruises. Normalization of MR imaging appearance is possible and is most often encountered after the occurrence of reticular lesions. Cartilage loss at follow-up is often found in cases of initial cartilage damage (impaction or osteochondral fracture). Conclusion: In general, a healing response was often encountered after sustained posttraumatic occult bone lesions. The initial MR imaging appearance appears to have prognostic value. (c) RSNA, 2006.