Sensitivity and specificity of bell-hammer tear as an indirect sign of partial anterior cruciate ligament rupture on magnetic resonance imaging

Abstract

Purpose: The main purpose of this study was to evaluate the usefulness of the bell-hammer sign in the diagnosis of partial tears of the anterior cruciate ligament (ACL) of the knee on MRI.

Methods: A retrospective study was performed including all patients who underwent ACL reconstruction for partial or complete tears from 2008 to 2009. The diagnosis of partial or complete ACL tears was based on the appearance of the ligament bundles and the signal quality on MRI. On arthroscopy, which is considered the gold standard, each bundle was classified as normal, partially or completely torn depending on the extent of the rupture and the quality of the remaining fibres. The study included 312 patients, 83 women and 229 men (mean age 33.3 ± 19.6 years). A diagnosis of a tear was made in all patients on preoperative MRI. Arthroscopy did not show any normal ACL, 247/312 (79.2 %) complete tears and 65/312 (20.8%) partial tears, 50/65 (76.9%) on the anteromedial bundle (AM) and 15/65 (23.1%) the posterolateral bundle.

Results: The bell-hammer sign was found on MRI in 13/312 patients (4.5%). It involved 9/65 (13.8%) partial tears, all in the AM bundle, and 4/247 (1.6%) complete tears, significantly more frequent in cases of partial rupture (p < 0.0001). MRI diagnosed a partial tear in 15/65 cases without the bell-hammer sign (sensitivity CI 95% = 23.1 ± 10%, specificity CI 95% = 95.9 ± 2.5%) and with the bell-hammer sign in 23/65 cases (sensitivity CI 95% = 35.4 ± 11%, specificity CI 95% = 93.9 ± 3%). The association of the bell-hammer sign with conventional radiological diagnostic criteria has improved diagnosis performance of MRI for partial tears but not significantly (ns).

Conclusion: The most important interest of the bell-hammer sign in the day-to-day clinical work is to suggest partial tears on MRI. It aids making a diagnosis, but its absence does not exclude partial ACL rupture.

Level of evidence: Diagnostic study, Level II.