
Background
A false-positive diagnosis is more common for the medial than the lateral meniscus. In the authors’ experience, some of the longitudinal tears were not confirmed at arthroscopy.
Objective
A study was conducted to determine if false-positive MRI diagnoses of a medial meniscal tear are associated with specific clinical variables on MRI findings.
A false-positive diagnosis of longitudinal tears is more common after acute trauma and when there is an anterior cruciate ligament (ACL) tear.
Design
Retrospective review.
Methods
MRIs of 559 patients were reviewed and correlated with clinical and arthroscopy findings. There were 50 patients with longitudinal tears included in this study group.
Results
The original MR sensitivity for diagnosing medial meniscal tears was 97% (343 of 355) and specificity was 84% (172 of 204). The positive predictive value (PPV) was highest for bucket handle and complex tears (100%) followed by horizontal tears (97%), flap tears (91%), radial tears (85%), and root tears (83%).
Longitudinal tears only had a PPV of 64%.
The PPV reached statistical significance between longitudinal and other tear patterns, except between longitudinal and radial/root tears.
Twenty-three of the false-positive tears were peripherally located. False-positive diagnoses of medial meniscal tears were more common in patients who had a prior episode of acute trauma (P =0.004) or an anterior cruciate ligament (ACL) tear (P <0.0001). At arthroscopy, 37 longitudinal tears were found, all of them in the posterior horn; 32 (86%) were diagnosed on the original MRI.
There were 18 false-positive diagnoses, with arthroscopy showing a completely normal meniscus in 14 patients and a completely healed tear in 4 patients.
All but 1 of the 50 patients with MRI diagnosis of longitudinal tear had an acute injury; 1 had chronic knee pain. The time interval between injury and MRI and between MRI and arthroscopy did not affect the PPV of longitudinal tears significantly.
The PPV was significantly lower when the tear contacted only the superior surface (P =0.016) or the surface at the meniscocapsular junction (P =0.0004).
Conclusions
False-positive MR diagnoses of medial meniscal tears are more common for longitudinal tears than other types of meniscal tears.
They are also more common if the tear contacts the superior surface or involves the meniscocapsular junction. Spontaneous healing of longitudinal tears accounts for some false-positive MR diagnoses.
A false-positive diagnosis of longitudinal tears is more common after acute trauma and when there is an anterior cruciate ligament (ACL) tear.
Reviewer’s Comments
This study explains that the decreased PPV in longitudinal meniscal tears is at least in part secondary to rapid healing.
Author: Cornelia Wenokor, MD
Reference:
De Smet AA, Nathan DH, et al. Clinical and MRI Findings Associated With False-Positive Knee MR Diagnoses of Medial Meniscal Tears. AJR; 2008; 191 (July): 93-99
Permalink: http://www.radiologydaily.com/?p=1917
Tags: acl, acl tear, ALL, anterior cruciate ligament, Arthroscopy, CT, EFE, knee arthroscopy, Longitudinal Meniscal Tears, medial meniscal tears, meniscal tear, meniscal tears, MI, MR, mri, Musculoskeletal Radiology, PE, Positive Predictive Value, ppv, rad, RFA, SPECT, TTE
Related
Free Special Reports on leading Radiology topics for you to download now. Plus, get free email newsletters.