Underestimation of labral pathology in adolescents with anterior shoulder instability

Shoulder instability is not uncommon in the adolescent athlete, and yet the ability for either clinical examination or magnetic resonance imaging (MRI) with arthrogram to accurately detect pathology in this younger population has not been elucidated yet. This study was performed to characterize the ability of physical examination and MRI to identify intra-articular pathology in those adolescents suspected of having anterior shoulder instability.

A retrospective review of patients treated over a year between 2008 and 2009 was undertaken. Included were patients with detailed physical examinations, preoperative MRI, and shoulder arthroscopy. Patients with previous shoulder surgery for instability, those who underwent surgery for brachial plexopathy, and those without an MR arthrogram were excluded. Demographics, age, sex, sports participation, and physical examination findings were recorded. Diagnostic arthroscopy findings were then compared with the clinical suspicion (based on history and physical examination) and the MRI findings. Imaging and arthroscopic results were categorized as anterior tear, anterior + tear (anterior tear with extension superior or posterior), other intra-articular pathology, or normal examination. An anterior instability cohort was then created by applying an exclusion criterion against patients without a clinical suspicion of anterior instability.

Forty-three patients were included after application of all inclusion and exclusion criteria. The clinical suspicion of anterior labral tear was 59% accurate (positive predictive value of 79%) and the MRI was 86% accurate (positive predictive value of 95%). Among all included patients, 23 adolescents (24 shoulders) were identified with a preoperative clinical suspicion of anterior labral tear based on history, physical examination, and plain radiographs (8 girls/15 boys). Mean age at surgery was 15.9 years (13.3 to 18.8). In this suspected anterior labral tear cohort, 79% had arthroscopic confirmation of the clinical suspicion, but 58% had extension of the labral tear either superior or posterior. MRI was 100% sensitive, 55% specific for an isolated anterior tear; yet, the MRI was 46% sensitive, 100% specific at identifying the larger anterior + labral tears. A statistically significant difference existed between the extent of the labral tear found on MRI and that found at the time of surgery (P=0.006), with tears of the glenoid labrum often extending beyond what was predicted by MRI.

Clinical suspicion and preoperative MR arthrogram of anterior instability in adolescents seem to detect evidence of labral pathology but have limited success in identifying the true extent of the labral pathology as proven by arthroscopy.

Level III Diagnostic Studies.