Faculty Bibliography

Purpose: Superior glenoid labral tears remain one of the most common causes of shoulder pain and loss of shoulder function. The clinical presentation for this group of patients can be non-specific. Thus imaging, in particular, MRI plays an important role in diagnosis. Currently, there is no consensus in terms of the most appropriate type of MRI to be used in this setting. The objective of this study was to evaluate the diagnostic accuracy of non-enhanced magnetic resonance imaging (MRI) and magnetic resonance arthrography (MRA) for diagnosis of superior labral tears in the shoulder. Materials and Methods: We performed a literature search (until August 2016) using PubMed (MEDLINE), Embase, ISIWeb of Science, Scopus, and national/international conference databases. The inclusion criteria consisted of original research studies that assessed the diagnostic accuracy of MRI, direct (d) MRA, and/or indirect (i) MRA for the detection of superior labral (SLAP) tears, while using arthroscopic findings as the reference standard. The methodological quality of each study was assessed with the use of the QUADAS 2 (Quality Assessment of Diagnostic Accuracy Studies) tool. Pooled sensitivities and specificities, and summary receiver-operating characteristic curves were calculated for each imaging strategy. Additional subgroup analyses compared 3 T and 1.5 T examinations of dMRI and MRA studies as well as low bias MRI and MRA studies. Study homogeneity was assessed visually on the basis of observed differences between study characteristics and methodologies, examination of the forest plots, and Cohran's Q tests of heterogeneity. Results: There were a total of 32 studies that met our inclusion criteria, including a total of 3525 total imaging examinations (1970 d-MRA, 1393 MRI, 162 iMRA). The sensitivities of dMRA, MRI, and iMRA for diagnosis of SLAP tear were 80.3% (CI 74.6-85.5), 62.9% (46.1-78.4), and 74.2 (66.1-81.6), respectively. The specificities of dMRA, MRI, and iMRA for diagnosis of SLAP tear were 90.7% (85.3-94.9), 90.8% (84.8-95.4), and 66.4 (51.0-79.6), respectively. The sensitivities of 3 T dMRA and 3 T MRI were 81.7% (71.6-89.9) and 79.4 (70.6-86.6). The specificities of 3 T dMRA and 3 T MRI were 94.3% (89.3-97.7) and 98.8% (95.8-99.8). The sensitivities of 1.5 T dMRA and 1.5 TMRI were 79.1% (68.9-87.7) and 81.2% (62.9-94.3). The specificities of 1.5 T dMRA and 1.5 T MRI were 83.6% (74.6-90.9) and 83.4% (78.9-87.3). The sensitivities of the low bias dMRA and low bias MRI were 83.2% (75.2-89.9) and 61.7 (40.7-80.7). The specificities of the low bias dMRA and low bias MRI were 90.6% (81.9-96.6) and 95.0% (85.1-99.7). No study heterogeneity was noted on the basis of observed differences between study characteristics, but it was noted upon examination of the forest plot for the pooled MRI sensitivities. Significance in departure from study homogeneity (p < .05) was seen for the dMRA and MRI pooled estimates. Conclusion: Overall, direct MR arthrography of the shoulder may be a better imaging option than non-contrast MRI for the diagnosis of a superior labral tear. 3 T MRI and 3 T dMRA may be better imaging options than their 1.5 T counterparts. These findings should be considered in the context of the degree of heterogeneity found in the available literature

Purpose: Several studies have demonstrated that shoulder ultrasound is an accurate, commonly utilized tool for the evaluation of rotator cuff tears, but the ordering of a follow up MRI has been incompletely investigated. The purpose of our study is to gain a better understanding of how often MRI is ordered after ultrasound for patients suspected of a rotator cuff tear, and the reasons for the ordering of this additional imaging examination. Materials and Methods: We performed a retrospective review of the electronic medical records of consecutive adult patients who underwent a shoulder ultrasound (US) examination for the evaluation of suspected rotator cuff pathology between January 1st 2014 and January 1st 2016. Exclusion criteria included patients who had prior shoulder rotator cuff surgery, and those with contraindications to MRI. Recorded data variables included age, sex, date of ultrasound, ultrasonography report findings, date of subsequent MRI (if obtained), MRI report findings, reason for performance of subsequent MRI, time interval between US and MRI (if obtained), whether the patients went on to rotator cuff surgery, ultrasound ordering physician specialty, and the length of time patients received follow up at our institution. Results: A total of 348 patients (176 males, 172 females); mean age 55.2 (range 18 -84) were included in this study based on our inclusion/ exclusion criteria. A shoulder MRI was obtained after shoulder ultrasound in 9.2% (n = 32) of patients. The mean time between shoulder US and MR was 88 days (range 0 - 262). The reasons for obtaining a subsequent shoulder MRI included, in descending order, persistent pain (n = 14), preoperative evaluation (n = 2), patient desire for anMRI (n = 2), too large body habitus (n = 1), concern for labral tear (n = 1), and possible brachial plexus lesion (1). In 11 cases, the reason for obtaining an MRI was not listed. There were 28 patients who underwent rotator cuff repair surgery. Of these patients who went to surgery, 82% (n = 23) went to surgery based on shoulder US only, and 18% (n = 5) obtained an MRI prior to surgery. Of the shoulder ultrasound examinations performed, the ordering clinician specialty was, in descending order, 90.2% (n = 314) orthopedics, 5.7% (n = 20) rheumatology, 2.6% (n = 9) primary care, and 1.4% (n = 5) physical medicine and rehabilitation. The mean follow up time for each patient was 21.1 months (range 9 - 32.4 months). Conclusion: The vast majority of patients who receive shoulder ultrasound for the evaluation of rotator cuff pathology at our institution do not end up getting a shoulder MRI afterwards. In the patients that did, the most common reason was for persistent pain. The majority of patients who went on to receive rotator cuff repair surgery did so based solely on shoulder ultrasound findings

Purpose: The optimal diagnostic imaging strategy for a suspected full-thickness rotator cuff tendon tear is controversial, in large part due to the significant up-front cost difference between magnetic resonance imaging (MRI) and ultrasound. Our objective was to compare the cost-effectiveness within the United States health care system of MRI versus ultrasound for initial imaging of patients with a suspected full-thickness rotator cuff tendon tear, and to determine the drivers of cost-effectiveness using sensitivity analysis. Materials and Methods: An expected-value decision analysis with rollback was utilized to compare the costs and outcomes of patients with a suspected full-thickness rotator cuff tendon tear who underwent either MRI, ultrasound, or ultrasound followed by MRI. A comprehensive literature search and expert opinion provided input data on probability estimates, event rates, costs, and health utility states for our model. All costs and benefits were considered from a societal perspective, and discounted at a 3% rate, to reflect the lower value of a delayed expense and the higher value of an earlier benefit. One-way sensitivity analyses were performed to test model robustness, followed by threshold analysis of all variables found to be sensitive. The primary outcomes assessed were costs in 2016 US dollars, effectiveness in quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs). Results: Expected costs per patient were lowest for ultrasound ($1,896), followed by ultrasound-MRI combined ($1,954), and MRI alone ($2,090). MRI was the most effective, providing 1.3410 QALYs, compared to ultrasound (1.3308 QALYs) and MRI after ultrasound (1.3317 QALYs). Ultrasound was the most cost-effective imaging strategy, as reflected by its lower cost/effectiveness ratio ($1425 per QALY) when compared to MRI after ultrasound ($1,467/QALY) and MRI ($1,559/ QALY). The incremental cost-effectiveness ratio for MRI was $18,951 per QALY gained, below the conventional willingness-to-pay threshold of $50,000 (WHO recommendation). The combination of MRI after ultrasound was not cost-effective (ICER $61,075 per QALY gained). Sensitivity analysis revealed that our model was sensitive only to the costs of MRI and ultrasound. Subsequent threshold analysis indicated that MRI remains cost-effective when the average cost for an MRI is less than $660. Moreover, if the cost for MRI falls below $148, or the cost for ultrasound rises above $379, MRI becomes both less costly and more effective than ultrasound, and thus the most cost-effective strategy. Conclusion: Our model showed that ultrasound had an expected cost that was $194 lower than MRI, but MRI was more effective than ultrasound by 0.0102 QALYs. Overall, ultrasound was the most cost-effective imaging strategy. For the base case, there was an ICER of $18,951 per QALY gained for MRI, well below the accepted threshold of $50,000 per QALY gained. Taken together, our results show that both ultrasound and MRI can be cost-effective imaging options to evaluate the patient suspected of a full thickness rotator cuff tendon tear

OBJECTIVE: To determine whether there is an association between T2/T2* mapping and supraspinatus tendon mechanical properties as assessed by shear-wave ultrasound elastography (SWE). MATERIALS AND METHODS: This HIPAA-compliant prospective pilot study received approval from our hospital's institutional review board. Eight patients (3 males/5 females; age range 44-72 years) and nine shoulders underwent conventional shoulder MRI, T2/T2* mapping on a 3-T scanner, and SWE. Two musculoskeletal radiologists reviewed the MRI examinations in consensus for evidence of supraspinatus tendon pathology, with tear size measured for full-thickness tears. T2/T2* values and ultrasound shear-wave velocities (SWV) were calculated in three corresponding equidistant regions of interest (ROIs) within the insertional 1-2 cm of the supraspinatus tendon (medial, middle, lateral). Pearson correlation coefficients between T2/T2* values and SWV, as well as among T2, T2*, SWV and tear size, were calculated. RESULTS: There was a significant negative correlation between T2* and SWV in the lateral ROI (r = -0.86, p = 0.013) and overall mean ROI (r = -0.90, p = 0.006). There was significant positive correlation between T2 and measures of tear size in the lateral and mean ROIs (r range 0.71-0.77, p range 0.016-0.034). There was significant negative correlation between SWV and tear size in the middle and mean ROIs (r range -0.79--0.68, p range 0.011-0.046). CONCLUSION: Our pilot study demonstrated a potential relationship between T2* values and shear wave velocity values in the supraspinatus tendon, a finding that could lead to an improved, more quantitative evaluation of the rotator cuff tendons.

OBJECTIVES: (1) Assess the association between the B-mode morphologic appearance and elasticity in the rotator cuff tendon using shear wave elastography (SWE). (2) Assess the association between SWE and symptoms. METHODS: Institutional Review Board approval and informed consent were obtained. A retrospective review identified 21 studies in 19 eligible patients for whom SWE was performed during routine sonographic evaluations for shoulder pain. Evaluations were compared with 55 studies from 16 asymptomatic volunteers and 6 patients with asymptomatic contralateral shoulders. Repeated studies were accounted for by resampling. Proximal and distal tendon morphologic characteristics were graded from 1 to 4 (normal to full-thickness tear), and average shear wave velocity (SWV) measurements were obtained at both locations. In 68 examinations, deltoid muscle SWV measurements were available for post hoc analysis. RESULTS: The morphologic grade and SWV showed weak-to-moderate negative correlations in the proximal (P < .001) and distal (P = .002) rotator cuff tendon. A weakly significant SWV decrease was found in the proximal tendon in symptomatic patients (P = .049); no significant difference was seen in the distal tendon. The deltoid muscle SWV showed weak-to-moderate negative correlations with the morphologic grade in the proximal (P = .004) and distal (P = .007) tendon; the deltoid SWV was also significantly lower in symptomatic shoulders (P = .001). CONCLUSIONS: Shear wave elastography shows tendon softening in rotator cuff disease. It captures information not obtained by a morphologic evaluation alone; however, a poor correlation with symptoms suggests that SWE will be less useful in workups for shoulder pain than for preoperative assessments of tendon quality. Deltoid muscle softening seen in morphologically abnormal and symptomatic patients requires further exploration.

AIMS: To evaluate whether an ultra-low-dose CT protocol can diagnose selected limb fractures as well as conventional CT (C-CT). PATIENTS AND METHODS: We prospectively studied 40 consecutive patients with a limb fracture in whom a CT scan was indicated. These were scanned using an ultra-low-dose CT Reduced Effective Dose Using Computed Tomography In Orthopaedic Injury (REDUCTION) protocol. Studies from 16 selected cases were compared with 16 C-CT scans matched for age, gender and type of fracture. Studies were assessed for diagnosis and image quality. Descriptive and reliability statistics were calculated. The total effective radiation dose for each scanned site was compared. RESULTS: The mean estimated effective dose (ED) for the REDUCTION protocol was 0.03 milliSieverts (mSv) and 0.43 mSv (p < 0.005) for C-CT. The sensitivity (Sn), specificity (Sp), positive predictive value (PPV) and negative predictive value (NPV) of the REDUCTION protocol to detect fractures were 0.98, 0.89, 0.98 and 0.89 respectively when two occult fractures were excluded. Inter- and intra-observer reliability for diagnosis using the REDUCTION protocol (kappa = 0.75, kappa = 0.71) were similar to those of C-CT (kappa = 0.85, kappa = 0.82). Using the REDUCTION protocol, 3D CT reconstructions were equivalent in quality and diagnostic information to those generated by C-CT (kappa = 0.87, kappa = 0.94). CONCLUSION: With a near 14-fold reduction in estimated ED compared with C-CT, the REDUCTION protocol reduces the amount of CT radiation substantially without significant diagnostic decay. It produces images that appear to be comparable with those of C-CT for evaluating fractures of the limbs. Cite this article: Bone Joint J 2016;98-B:1668-73.

There has been a rapid increase in the number of shoulder arthroplasties, including partial or complete humeral head resurfacing, hemiarthroplasty, total shoulder arthroplasty, and reverse total shoulder arthroplasty, performed in the United States over the past two decades. Imaging can play an important role in diagnosing the complications that can occur in the setting of these shoulder arthroplasties. This review is divided into two parts. The first part provides a general discussion of various imaging modalities, comprising radiography, CT, MRI, ultrasound, and nuclear medicine, and their role in providing useful, treatment-guiding information. The second part focuses on the most appropriate imaging algorithms for shoulder arthroplasty complications such as aseptic loosening, infection, fracture, rotator cuff tendon tear, and nerve injury. The evidence-based ACR Appropriateness Criteria guidelines offered in this report were reached via an extensive analysis of current medical literature from peer-reviewed journals and the application of well-established methodologies (the RAND/UCLA Appropriateness Method and the Grading of Recommendations Assessment, Development, and Evaluation) for rating the appropriateness of imaging and treatment procedures for specific clinical scenarios. Further analysis and review of the guidelines were performed by a multidisciplinary expert panel. In those instances in which there was insufficient or equivocal data for recommending the appropriate imaging algorithm, expert opinion may have supplemented the available evidence.

Objectives: Proper surgical planning requires accurate and reliable pre-operative patient information. The more comprehensive the data, the more the surgeon can tailor a general surgical technique to an individual patient's unique anatomy. A previous retrospective study demonstrated that three-dimensional magnetic resonance imaging more accurately characterized rotator cuff tears compared to two-dimensional images when checked against intraoperative pictures. The purpose of this study was to determine if three-dimensional MRI imaging would continue to be more accurate than two-dimensional imaging in a prospective study. Methods: Patients were prospectively included if they had a full-thickness primary rotator cuff tear on pre-operative MRI. Intra-op videos were taken from the posterior and lateral portals, with a grasper fully mobilizing the torn tendon in each view. 7 surgeons then reviewed the videos and independently characterized the shape of the tears into crescent, U-shaped tears, L-shaped tears, or massive tears. This was considered the gold-standard. Two musculoskeletal radiologists reviewed the corresponding MRI studies independently and blind to the arthroscopic findings and characterized the shape on the basis of the tear's retraction and size 2D MRI. The 3D reconstructions of each cuff tear were reviewed by each radiologist to characterize the shape. Statistical analysis included 95% confidence intervals and fleiss's kappa. Results: 37 patients were enrolled in the study. Among the 7 surgeons, agreement on cuff tear was 93% ( =.87). The accuracy for differentiating between crescent-shaped, longitudinal, and massive tears using measurements on 2D MRI was 73.4% for reader 1 and 71.2% for reader 2. The accuracy for tear shape characterization into crescent and longitudinal U- or L-shaped using 3D MRI was 92% for reader 1 and 94% for reader 2. When further characterizing the longitudinal tears as massive or not using 3D MRI, both readers had an accuracy of 83%. The overall accuracy of 3D MRI was 91%, significantly different (P=.07) from 2D MRI accuracy (70.4). Conclusion: Our study has demonstrated that 3D MR reconstructions of the rotator cuff improve the accuracy of characterizing rotator cuff tear shapes compared with current 2D MRI-based techniques. (Figure Presented)