Protocol optimization in the MRI evaluation of the rotator cuff focused on the supraspinatus tendon.

Transcription

1 Protocol optimization in the MRI evaluation of the rotator cuff focused on the supraspinatus tendon. Poster No.: C-1211 Congress: ECR 2015 Type: Scientific Exhibit Authors: M. Etancelin-Jamet, L. Bouilleau; Tours/FR Keywords: Musculoskeletal joint, Musculoskeletal system, MR, Diagnostic procedure, Tissue characterisation DOI: /ecr2015/C-1211 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. Page 1 of 38

2 Aims and objectives Shoulder pain is the third most common musculoskeletal complaint which patients consult their doctor for, after low back pain and knee pain (1) and has a prevalence in the general population between 16% and 26%. The most prevalent cause of shoulder pain, occurring in approximately 65%-70% of patients is rotator cuff disease (2). This prevalence increases with age, and it is estimated that, by the age of 70 years, more than 50% of the population will have a full or partial thickness rotator cuff tear. It is a real public health problem because of its functional impact. There are some different imaging techniques, which can be used to detect rotator cuff abnormalities including ultrasonography (US), and magnetic resonance (MR) imaging (3). US evaluation of the rotator cuff is performing to diagnose injuries with a standard study of each tendon in its two orthogonal planes following the curve of the distal part. We all know that US gives a really good view of the fibrillar structure of tendons (fig.1). Advantages of US include portability, low cost, lack of contraindications and the spatial resolution of images obtained with US is sometimes higher than images obtained with routine MR imaging (3), whereas both image acquisition and image interpretation seem to be very dependent on the experience of the physician. That's why, now, MR imaging tends to be the main technique used to make the assessment of the rotator cuff because radiologists consider this technique less operator dependent than with US and high accuracies in the diagnosis of rotator cuff disease can be achieved (4,5,6). The other huge advantage of the use of MR imaging is the global assessment of all shoulder structures. Nevertheless, there are some important difficulties, which have to be known by the radiologists. Indeed, there are difficulties like claustrophobia and contraindications owing to metallic implants and electronic devices, cost, and accessibility, which are the common lot of MR imaging. But for shoulder, some studies have shown that MR imaging is not without any serious technical limitations like patients movements due to shoulder pain because of the position, the magic-angle effect (fig.2) and possible partial volume effect (which are two important difficulties because of the curved part of tendons (7,8,9)), but also some problems inherent to the structure of the rotator cuff. The cuff has a complex anatomy (10,11) with a lot of layers and interconnections as we can see on the figures n 3 and n 4, for example supraspinatus and infraspinatus tendons have common crossed fibers (12). All these difficulties create an important intra and inter-observer variability (13,14), especially for non-specialized radiologists in musculoskeletal system. Page 2 of 38

3 Defined imaging techniques and protocols are important to minimize such operator dependence. Otherwise, the first tendon of the rotator cuff injured is supraspinatus tendon because of its position; its complex curve and its mobility depend of the humeral position (fig. 5). There is also a lack of vascularisation in the deepest part of the tendon with difficulties for regeneration. Most of the rotator cuff injuries found by imaging techniques concern the supraspinatus. Currently, MR protocol is based on gleno-humeral joint study composed of strict axial, oblique coronal and sagittal (based on the axis of the humerus) planes. We saw in previous paragraphs that US has the advantage of imaging the tendons in their two orthogonal planes with a good spatial resolution, and that supraspinatus tendon seems to be the first tendon concerned by injuries. We attempted to apply a similar methodology to MRI of the most movable and weakest tendon, the supraspinatus tendon with a new slice placement by reducing technical limitations like magic angle and partial volume effect. The main objective of our study is to compare the diagnostic performance of a new specific supraspinatus angled sagittal T2-weighted sequence labelled «of cuff» to that of our standard MRI protocol in detection of rotator cuff injuries. Then, we will assess if this new sequence is easier than the classical protocol for the identification of each tendon path and for the characterization of the tears. Images for this section: Page 3 of 38

4 Fig. 1: Fibrillar structure of supraspinatus tendon studying by ultrasonography. Fig. 2: Coronal MR images in T2-weighted sequence focused on the supraspinatus tendon illustrating the magic-angle effect: increased signal on the curved part of the tendon disappearing after different shoulder position. Page 4 of 38

5 Fig. 3: Rotator cuff anatomy: anterior side. Page 5 of 38

6 Fig. 4: Rotator cuff anatomy: superior side. Page 6 of 38

7 Fig. 5: Figures: drawing showing the mobility of the supraspinatus tendon based on the humeral position. On the left: Internal rotation: supraspinatus trajectory facing downwards. On the right: External rotation and abduction: the supraspinatus tendon seems to be more straight. Page 7 of 38

8 Methods and materials Conducted at a teaching hospital in Tours France, our study included all patients referred for MRI examination for possible rotator cuff injury since February Our standard MRI protocol (1.5 or 3 Tesla, Avanto, Siemens) after locator in the three planes included: - Fat-Sat T2-weighted strict axial sequence, based on the scapular orientation (fig.6), - After greater tubercle location, an oblique coronal Fat-Sat T2-weighted sequence in the plane of the suprapinatus tendon (fig. 7,8,9), - A strict sagittal Fat-Sat T2-weighted sequence based on the axis of the humerus and the coil (fig.10 and 11) - A strict sagittal T1-weighted sequence (fig.12). We added a sagittal T2-weighted sequence with Fat Sat labelled "of cuff", positioned orthogonally to the convex part of the supraspinatus tendon on the humeral head located on the coronal sequence (fig.13 and 14). A dedicated shoulder phased-array coil was used for signal detection. The humerus was positioned in slight external rotation and abduction (near 20 ). The parameters are noted in table n 1. We included all major patients without standard contraindications after oral information and oral consent. We obtained agreements of the ethics committee and the CNIL (Commission nationale de l'informatique et des libertés, France). Page 8 of 38

9 Since February 2014, we've included 66 patients. Their mean age is 50.8 years old (between 17 and 76 years old). There are 34 men and 32 women. We then compared the standard protocol to the additional sagittal T2 sequence for diagnostic performance. Four radiologists in blinded analysis currently review each shoulder examination: two senior radiologists specialized in musculoskeletal system, one resident in radiology and one general radiologist. They have to choose for each tendon (supraspinatus, infraspinatus, subscapularis and biceps) in each protocol (standard protocol or «of cuff» sequence) between the four following answers: no tear, tendinopathy, partial-thickness tear or full-thickness tear (definitions summarised in table n 2). Answers are then collected in an Excel table. There is at least one month between the reading of the standard protocol and the reading of the new sequence. Then, they have to say if the location of tendons and characterization of tears are easier with the new sequence or not. We can compare the performance of the both protocols by considering that the standard protocol is our reference. After this data collection, we try to track the future of patients (surgical reports, other additional examination, and arthroscopy reports) to detect possible differences with MRI imaging. This study is still in progress. Page 9 of 38

10 Images for this section: Fig. 6: Axial sequence in T2-weighted fat-suppressed sequence. Page 10 of 38

11 Fig. 7: Anatomical landmarks to place accurately coronal sequence. Page 11 of 38

12 Fig. 8: Slice placement for coronal sequence. Page 12 of 38

13 Fig. 9: Coronal T2-weighted fat-suppressed sequence with location of supraspinatus distal part. Page 13 of 38

14 Fig. 10: Slice placement for standard sagittal sequence in humeral long axis. Page 14 of 38

15 Fig. 11: Standard sagittal T2-weighted fat-suppressed sequence. Page 15 of 38

16 Fig. 12: Sagittal T1-weighted sequence. Page 16 of 38

17 Fig. 13: Slice placement for sagittal sequence labelled "of cuff": slices are disposed orthogonally to the supraspinatus distal part. Page 17 of 38

18 Fig. 14: Sagittal T2-weighted fat-suppressed sequence labelled of cuff. The difference between the anterior and the posterior surfaces of the greater tubercle (GT) is easy on this slice. Page 18 of 38

19 Table 1: Sequence parameters used for our study. Table 2: Definitions of the different possible appearances of tendons. Page 19 of 38

20 Fig. 15: Tendinopathy affecting the distal part of the supraspinatus tendon in coronal T2weighted fat-suppressed sequence. Page 20 of 38

21 Fig. 16 Page 21 of 38

22 Fig. 17: Figures 16 and 17:Partial-thickness tear: partial disruption affecting the distal part of the infraspinatus tendon on its inner face in coronal and sagittal "of cuff" T2-weighted fat-suppressed sequences. Page 22 of 38

23 Fig. 18: Full-thickness tear affecting the supraspinatus tendon with retraction (*) of this one in coronal T2-weighted fat-suppressed sequence. Page 23 of 38

24 Results The intermediate results, noted in June 2014 showed some important points. First of all, radiology technicians need a quite long period to learn the new protocol because the slice placement has been really different than the standard use. That's why, quality of the exams during about a month, was poor. They had to break away from their own practices and radiologist had to teach them the good slice placement. 36 MRI have been done. On these exams, only 31 respected the protocol, 19 of them showed rotator cuff tears. Radiologists found that the interpretation of 13 of them was easier with «cuff» protocol specially to identify each tendon (supraspinatus versus infraspinatus). New protocol seems to increase sensivity to detect little tears (Fig. 19 to 28). Otherwise, there isn't any loss of information compared to the standard protocol. Images for this section: Page 24 of 38

25 Fig. 19 Page 25 of 38

26 Fig. 20 Page 26 of 38

27 Fig. 21: Figures 19,20,21: Coronal T2-weighted fat-suppressed images: there is possibly a tear (*) at the distal part of supraspinatus or infraspinatus tendon. It is quite difficult to specify the localization on this sequence alone. Page 27 of 38

28 Fig. 22 Page 28 of 38

29 Fig. 23 Page 29 of 38

30 Fig. 24: Figures 22,23,24: Standard sagittal T2-weighted fat-suppressed images: we can see that there is at least a partial-thickness tear of the supraspinatus distal part but we cannot exclude a possible thin tear of the infraspinatus distal part. The fluid-like signal seems to reach the infraspinatus anterior fibers. Page 30 of 38

31 Fig. 25 Page 31 of 38

32 Fig. 26 Page 32 of 38

33 Fig. 27 Page 33 of 38

34 Fig. 28: Figures 25,26,27,28: Sagittal "of cuff" T2-weighted fat suppressed images: we can more easily assert that the tear is a full-thickness tear injurying supraspinatus distal part but the infraspinatus tendon seems to be spared. The limit between both tendons is more evident by visualizing the two sides of the greater tubercle. Supraspinatus tendon is viewed in a perpendicular way what makes the characterization of tears clear. Page 34 of 38

35 Conclusion Our purpose was to be closer than US methodology with MRI technique to improve performance in rotator cuff study especially for the supraspinatus tendon by reducing technical limitations like magic angle and partial volume effect. This approach has already been considered by a workgroup of radiology of University of Wisconsin Hospital in 2001 and By a Korean team in early 2014 with inconsistent result. Tuite and al. (15), showed in 2001 that there was a slight improvement in the diagnostic accuracy for partial-thickness tear with the angled oblique sagittal sequence but only for one reader; in 2014 Kim and al. (16) showed there was no significantly different diagnostic performance in detection of partial and full thickness supraspinatus tendon tears compared with standard MRI protocol. But there are some differences between these studies and our work: indeed, we study the entire cuff and not only the supraspinatus tendon even if it remains central in developing this work. Most of our MRIs are done on a 3-T MRI. Tuite's study (15) is becoming to be quite old and we know that MRI performances are in perpetual improvement. Our reference is the standard protocol and not the arthroscopy considered in the other studies like the gold standard. Moreover, if the main objective is to compare diagnostic performance, there are some important different points like the ease of reading the exam even for a non-specialised radiologist in musculoskeletal system. We showed that it was complicated to break away from the standard use but after a learning time, we can hope a good quality of this exam. Based on our first results, we can say that interpretation of abnormalities seems to be easier with the «cuff» protocol because of a better location of each tendon and a better facility to identify little injuries. These first impressions have to be confirmed at the end of our study. It seems to give a quite important hope to facilitate shoulder study for exigent radiologists, who believe that MRI imaging is operator dependent as the US study is usually considered. It could even help those who aren't specialised in musculoskeletal system. We can expect a decrease of inter and intra-observer variability and a better quality of this exam. Personal information Page 35 of 38

36 Mathilde ETANCELIN-JAMET Resident in radiology 4th year CHU TOURS, FRANCE Dr. Loïc BOUILLEAU Department of radiology CHU TOURS, France References 1. Mitchell C., Adebajo A., Hay E., Carr A. "Shoulder pain: diagnosis and management in primary care. " BMJ ; (2005): Shanahan EM., Sladek R. "Shoulder pain at the workplace." Best Pract Res Clin Rheumatol ;25-1(2011): Nazarian L.N., Jacobson J.A., Benson C.B., Bancroft L.W., Bedi A., McShane J.M., Miller T.T, et al. "Imaging Algorithms for Evaluating Suspected Rotator Cuff Disease: Society of Radiologists in Ultrasound Consensus Conference Statement." Radiology 267, no. 2 (May 2013): Magee T. and Williams D. "3.0-T MRI of the Supraspinatus Tendon." AJR. American Journal of Roentgenology 187, no. 4 (October 2006): Krausé D., Lambert A., Loffroy R., Guiu B., Ben Salem D., Mezzetta L., Méjean N., and Cercueil J.P. "Les ruptures de la coiffe des rotateurs en IRM 3T." Journal de Radiologie 89, no. 10 (October 2008): Page 36 of 38

37 6. Rutten M.J.C.M., Spaargaren G.J., van Loon T., de Waal Malefijt M.C., Kiemeney L.A.L.M., and Jager G.J. "Detection of Rotator Cuff Tears: The Value of MRI Following Ultrasound." European Radiology 20, no. 2 (February 2010). 7. Madden M.E. "The Magic-Angle Effect of the Supraspinatus Tendon." Radiologic Technology 77, no. 5 (June 2006): Wright T., Yoon C., and Schmit B.P. "Shoulder MRI Refinements: Differentiation of Rotator Cuff Tear from Artifacts and Tendonosis, and Reassessment of Normal Findings." Seminars in Ultrasound, CT, and MR 22, no. 4 (August 2001): Timins M.E., Erickson S.J., Estkowski L.D., Carrera G.F., and Komorowski R.A. "Increased Signal in the Normal Supraspinatus Tendon on MR Imaging: Diagnostic Pitfall caused by the magic-angle effect." AJR. American journal of roentgenology, no. 164 (1995): Nové-Josserand L., Godenèche A., Noël E., Liotard J.-P., Walch G. "Pathologie de la coiffe des rotateurs" EMC (Elsevier Masson SAS, Paris), Appareil Locomoteur, A-10 (2008) 11. Rutten, Matthieu J. C. M., Bas J. Maresch, Gerrit J. Jager, Johan G. Blickman, and Marnix T. van Holsbeeck. "Ultrasound of the Rotator Cuff with MRI and Anatomic Correlation." European Journal of Radiology 62, no. 3 (June 2007) 12. Mochizuki T. "Humeral Insertion of the Supraspinatus and Infraspinatus:New Anatomical Findings Regarding the Footprint of the Rotator Cuff." The Journal of Bone and Joint Surgery (American) 90, no. 5 (May 1, 2008): 962. Page 37 of 38

38 13. Theodoropoulos, John S., Gustav Andreisek, Edward J. Harvey, and Preston Wolin. "Magnetic Resonance Imaging and Magnetic Resonance Arthrography of the Shoulder: Dependence on the Level of Training of the Performing Radiologist for Diagnostic Accuracy." Skeletal Radiology 39, no. 7 (July 2010): Halma, J.J., Eshuis R., Krebbers Y.M.J., Weits T., and de Gast A. "Interdisciplinary Inter-Observer Agreement and Accuracy of MR Imaging of the Shoulder with Arthroscopic Correlation." Archives of Orthopaedic and Trauma Surgery 132, no. 3 (March 2012): Tuite, M. J., D. Asinger, and J. F. Orwin. "Angled Oblique Sagittal MR Imaging of Rotator Cuff Tears: Comparison with Standard Oblique Sagittal Images." Skeletal Radiology 30, no. 5 (May 2001): Kim, Jihee. "The Difference of Diagnostic Performance for Detecting Rotator Cuff Tears by Adding Angled Oblique Sagittal Images to the Routine Shoulder MRI," (2014) Page 38 of 38