2015 ISAKOS Biennial Congress Paper #0

Effect of Time And Contrast Use for Magnetic Resonance Imaging in Acute Anterior Shoulder Instability: Determining the Accuracy Of Labrum Tear Size

Joshua C Setliff, BA, Pittsburgh, PA UNITED STATES
Justin James Greiner, MD, Papillion, NE UNITED STATES
Nicholas P Drain, MD, Pittsburgh, PA UNITED STATES
Asher B Mirvish, BA, Pittsburgh UNITED STATES
Michael Andrew Fox, MD, Pittsburgh, PA UNITED STATES
Noel Carlos, MS, Pittsburgh, PA UNITED STATES
Lena L Vodovotz, BS, Pittsburgh UNITED STATES
Ajinkya Rai, BS, Pittsburgh UNITED STATES
Samantha Ashok, BS, Pittsburgh UNITED STATES
Guang-Ting Cong, MD, Holmdel, NJ UNITED STATES
Bryson P. Lesniak, MD, Presto, Pennsylvania UNITED STATES
Albert Lin, MD, Pittsburgh, PA UNITED STATES
University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, UNITED STATES

FDA Status Not Applicable

Summary: Extent of labral tear differed between MRI and intraoperative findings after acute shoulder instability, with no statistical differences based on time of imaging or the addition of intra-articular contrast; thus, need for arthrogram MR in the setting of anterior shoulder instability should be carefully considered.

Abstract:

Introduction

Controversy exists regarding the inclusion of intra-articular contrast to enhance labral tear detection following acute anterior shoulder instability. The purpose of this study was to compare the accuracy of MR arthrogram to non-arthrogram MR in both the acute and delayed setting following anterior shoulder instability.

Methods

Patients sustaining an acute anterior shoulder dislocation or subluxation between 2012 and 2021 were included if a specific date of injury and subsequent arthrogram or non-arthrogram MRI was available. Musculoskeletal trained radiologists determined the size of labral tear based on a clock face from MRI. The MRI interpretation was compared to the gold standard of intraoperative clock face labral tear size at the time of arthroscopic stabilization. The difference in extent of labrum tear size between MRI and surgical findings at every half hour was recorded. Days from injury to MR was further categorized as acute (0 to 7 days) or delayed (>7 days).

Results

Thirty-nine patients (mean age 24.5 years) met inclusion criteria. Median time from injury to MRI and surgery was 9 and 45 days, respectively. 16 patients underwent non-arthrogram MR (8 acute, 8 delayed) with 23 patients underwent arthrogram MR (6 acute, 17 delayed). The entire cohort demonstrated a mean of 4.7 half-hour labral tear size differences between MR and intraoperative findings. No statistical difference was identified for labral tear size differences between non-arthrogram MR (3.3) and arthrogram MR (5.7, p=0.83). ANOVA testing demonstrated no difference in labral tear size detected (p=0.2116) based on number of days from injury to MR.

Discussion

Extent of labral tear differed between MRI and intraoperative findings after acute shoulder instability, with no statistical differences based on time of imaging or the addition of intra-articular contrast. The additional cost, time, and morbidity of arthrogram MR should be weighed in the setting of anterior shoulder instability.