Sixty-nine patients with traumatic knee hemarthrosis were evaluated an average of
3 days after trauma by high field (1.5T) magnetic resonance imaging (MRI) using sagittal
T1, T2-weighted and coronal 3D-gradient echo images. All knees were arthroscopically
examined shortly afterwards. The diagnostic validity of MRI for intraarticular pathology
was determined using arthroscopy as golden standard. All patients had pathological
findings on arthroscopy. The injuries were sports-related in 77 % of the cases. MRI
was highly sensitive (86 %) and specific (92 %) for diagnosis of anterior cruciate
ligament tears. Diagnosis of medial meniscal tears showed a 74 % sensitivity and 66
% specificity. MRI detected lateral meniscal tears in 50 % with an 84 % specificity.
As such, MRI missed 10 significant meniscus ruptures requiring surgical treatment.
The sensitivity for partial or total medial collateral ligament tears was 56 %, the
specificity 93 %. Rupture of the medial retinaculum in cases with patellar dislocation
or significant damage of articular cartilage were only detected by MRI in a few cases
(27 % and 20 % sensitivity, respectively). MRIs low diagnostic validity for intraarticular
pathology with hemarthrosis may be attributed to the shifting paramagnetic properties
of the blood remains and catabolic processes in meniscal and chondral tissues during
the hemoglobin degradation process. Accordingly, MRI, with the technique used, could
neither replace arthroscopy in the diagnosis and screening of acute knee injuries,
nor select patients with need for immediate arthroscopic meniscal surgery.
Key words
Acute knee injuries - hemarthrosis - magnetic resonance imaging - arthroscopy - diagnostic
validity
