ShoulderMRArthrography: Comparative Evaluation of Three Different Contrast Injection Techniques Using an Anterior Approach

Background MR arthrography (MRA) is commonly used in the assessment of shoulder internal derangements. Correct intra-articular contrast injection is required for this modality. Anterior injections under fluoroscopic, ultrasound-guidance, or without image-guidance have been described in the literature. However, no simultaneous comparison has been performed between the three techniques. Purpose To compare the accuracy and performance of fluoroscopy (FL)-guided, ultrasound (US)-guided and non-image-guided intra-articular contrast injection via an anterior approach for performing shoulder MRA. Study Type Prospective. Subjects Two-hundred and ten patients (180 men and 30 women; mean age, 33 +/- 12 years; range 20-60 years) with clinically suspected shoulder pathology. Field Strength/Sequence 1.5T/fat-suppressed T-1-weighted, T-2-weighted, and 3D-gradient-echo images. Assessment Patients underwent shoulder MRA after anterior intra-articular contrast injection under FL- or US-guidance or without image-guidance. Patients were randomized among the three techniques with each group comprising 70. The techniques were compared according to the accuracy of intra-articular needle placement, attempts success rate, pain during and 24 hours after injection, procedure times, contrast extravasation rate, joint distension, and MRA diagnostic efficacy. Pain was assessed by the visual analog scale (VAS) pain-score. Statistical Tests Pearson's chi-squared and Kruskal-Wallis tests. Results FL- and US-guided injections (100% accuracy) were significantly more accurate than non-image-guided (85.7% accuracy) (P < 0.05). US-guidance was the least painful, with statistical differences between image-guided and non-image-guided techniques regarding the first attempt success rate (95.7% and 92.8% for FL- and US-guided vs. 78.6% for blinded), VAS-score 24 hours-post-procedure (1.7 +/- 1.7, and 1.5 +/- 1.4 vs. 2.2 +/- 1.4), procedure time (11.9 +/- 1.6, and 7.4 +/- 1.7 vs. 4.3 +/- 0.76 minutes), and contrast extravasation rate (5.7%, and 8.6% vs. 30%) (allP < 0.05). Procedure time was also significantly different between FL and US-guidance (P < 0.05). Data Conclusion Imaging-guided injections are more accurate and tolerable than non-image-guided and should be considered to confirm intra-articular needle position, hence adequate capsular distension and good diagnostic quality of shoulder MRA. US guidance is a less painful, rapid, and safe alternative to the FL approach. Evidence Level: 2 Technical Efficacy Stage: 5.