- Email: [email protected]
Shoulder MRI after surgical treatment of instability
European Journal of Radiology 30 (1999) 2 – 4
Shoulder MRI after surgical treatment of instability Martin Vahlensieck a,*, Philipp Lang b, Ulli Wagner c, Frank Mo¨ller c, Urs van Deimling c, H.K. Genant b, Hans H. Schild a b
a Uni6ersity of Bonn, Department of Radiology, Sigmund – Freud-Strasse 25, 53105 Bonn, Germany Uni6ersity of California San Francisco, Department of Radiology, 505 Pamassus A6enue, San Francisco, CA 94143, USA c Uni6ersity of Bonn, Department of Orthopedic Surgery, Sigmund – Freud-Strasse 25, 53105 Bonn, Germany
Received 17 June 1998; received in revised form 7 October 1998; accepted 8 October 1998
Abstract Objecti6e: To analyze magnetic resonance imaging (MRI) findings of the shoulder after an instability operation. Materials and methods: Physical examinations, radiographs and MRI of 10 patients after anterior glenoid bone block insertion for ventral instability were compared. MRI included T1-weighted spin-echo (TR = 600, TE = 20 ms) and T2*-weighted gradient-echo sequences (TE = 600, TE = 18, Flip =30°) in the axial, oblique-coronal and oblique-sagittal planes. Results: No patient suffered from recurrent subluxation. We found fusion of the bone block with the anterior glenoid in seven cases, dislocation of the bone block without contact to the glenoid in one case, and no visible bone block in two cases. On MRI, the bone block showed either signal intensity equivalent to fatty bone marrow (n = 4) or was devoid of signal consistent with cortical bone or bone sclerosis (n= 4). In all patients, a low signal intensity mass, 2 – 4 cm in diameter, was visible next to the glenoid insertion site. Conclusion: Insertion of a bone block onto the anterior glenoid induces formation of scar tissue, increasing the stability of the shoulder joint. This scar is well visible on MRI and forms independently of the behavior of the bone block itself. MRI is ideally suited for evaluating postoperative shoulder joints after bone-grafting procedures. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Shoulder MRI; Surgery; Postoperative MRI; Instability
1. Introduction
2. Material and methods
One possible surgical procedure in the management of shoulder instability is the insertion of a bone block onto the anterior glenoid (Eden – Hybinette) [1,2] (Table 1). Bone block resorption or dislocation, seen clearly on radiographs, surprisingly does not impair functional long-term results. We therefore wanted to study the long-term soft-tissue changes of the shoulder after this operation by means of magnetic resonance imaging (MRI) in order to elucidate the therapeutic principle and to establish MRI criteria for follow-up.
We examined 10 patients with prior history of an Eden–Hybinette operation. The operated patients consisted of seven males and three females, 30–53 years old (mean 45 years). The operation had been performed 15–17 years before MRI. History and physical examination were compared with conventional radiographs. MRI of the shoulder was performed using a standard surface loop coil and T1-weighted spin-echo (TR =600 ms, TE= 20 ms) and T2*-weighted gradient-echo sequences (TR= 600, TE=18 ms, Flip=30°) in the coronal, sagittal, and axial plane. We used a 1.5 Tesla superconducting system (ACS, Philips Medical), a fieldof-view of 180 mm2, a 205× 256 matrix, and 4 mm slice thickness. Images were particularly assessed for contour and signal abnormalities of the bony glenoid, the ante-
* Corresponding author. Tel.: +49 228 2875875; fax: + 49 228 2876793.
0720-048X/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S0720-048X(98)00146-6
M. Vahlensieck et al. / European Journal of Radiology 30 (1999) 2–4
3
rior glenoid labrum, the visibility, form, and signal intensity of the bone block, and the configuration and signal intensity of the surrounding soft tissues. Analysis was performed by two radiologists by consensus reading.
3. Results All patients had no recurrent luxation after the operation, and physical examination yielded no abnormality. Radiographs of these patients showed solid ingrowth of the bone into the anterior glenoid in seven cases, bone block dislocation without contact to the glenoid in one case, and no visible bone block in two cases. Two patients showed osteoarthritis of the shoulder joint, and in one case, a Hill – Sachs defect was visible. The MRI demonstrated high-signal intensity of the bone block in four cases, consistent with fatty bone marrow, and low signal intensity in four cases, consistent with cortical or sclerotic bone. In each case, a low signal intensity mass ventral to the anterior glenoid was visible. This mass was 2 – 4 cm in diameter. In two cases, we found multifocal signal voids within the mass which were emphasized on gradient-echo images. The mass itself and its relation to the capsular mechanism and bony glenoid was best visible on axial images (Figs. 1 and 2). However, it was also appreciable on coronal and sagittal views (Figs. 3 and 4). The ventral capsular mechanism was visible as a low signal intensity band in all patients. The anterior labrum was not visible in seven cases, was blunted in one case, and was triangular in two cases.
Fig. 1. Axial T1-weighted spin-echo MRI image (TR = 600, TE = 20 ms) of a patient 15 years after anterior glenoid bone block insertion. Adjacent to the anterior glenoid, a large low signal intensity mass is visible (curved arrow). This mass also showed low signal intensity on T*-weighted images (not shown). The bone block itself was resorbed 2 and was not visible on radiographs and MRI images. Within the anterior glenoid, some high signal intensity suggests fatty bone marrow at the operation site (arrow)
4. Discussion Knowledge of normal MRI findings in postoperative shoulder joints is important in order to delineate complications, and in order to detect persistent or recurrent disease from an uncomplicated course. Only a few articles describe such findings [3,4], particularly after surgery for rotator cuff pathology. In the present study, we analyzed shoulder MRI after surgery for anterior instability using bone block insertion.
Table 1 Some types of surgical procedures used for recurrent anterior Glenohumeral joint dislocations (from Ref. [6]) Procedure
Technique
Bankart [7]
Repair of anterior capsular mechanism using drill holes and sutures Putti–Platt [8] Shortening of the anterior capsule and subscapularis muscle Magnuson–Stack [9] Transfer of the subscapularis tendon from lesser tuberosity to greater tuberosity Eden–Hybinette [1,2] Bone graft to anterior glenoid region Oudard [10] Bone graft to coracoid process Trillat [11] Osteotomy with displacement of coracoid process Bristow–Helfet [12] Transfer of coracoid process with the first attached tendons to the neck of the scapula
Fig. 2. Schematic drawings of axial shoulder MRI sections in a healthy subject (a) and after anterior gienoid bone block insertion (b). A fatty marrow equivalent, high signal intensity bone block is surrounded by a large, low signal intensity scar.
4
M. Vahlensieck et al. / European Journal of Radiology 30 (1999) 2–4
Fig. 3. Oblique-sagittal T1-weighted spin-echo MRI image (TR = 600, TE=20 ms) of a patient 16 years after anterior glenoid bone block insertion. Anteriorly to the glenoid fossa, lies the bone block with signal intensity equivalent to fatty marrow (curved arrow) surrounded by low signal intensity scar tissue (arrows). This tissue was also of low signal on T*-weighted images (not shown). C, coracoid process. 2
The insertion of a bone block onto the anterior glenoid seems to induce the development of a soft-tissue mass, which, given its low signal intensity on T1- and T*-weighted images, appears to consist of fibrous scar 2 tissue. Since this mass develops also in the cases of bone block resorption and dislocation, it is considered to be a regular postoperative finding, and most suggestive of importance for the therapeutic principle in the Eden– Hybinette operation. The fibrous mass obviously acts as mechanical stabilization of the shoulder and prevents recurrent luxation. However, since we did not encounter patients in whom the operation failed, the last consideration cannot be proven confidently. The radiographical appearance of the bone block itself cannot be used as an indicator for therapeutic success, since persistent, resorbed and dislocated bone block were encountered independently from the clinical outcome. One limitation of our study is the lack of histologic confirmation of the ventral mass to represent fibrous scar tissue. However, due to the signal behavior of low intensity on both T1- and T2-weighted MRI images, as well as the location next to the bone block insertion site, no other conclusion seems to be likely. Shrinking of the ventral capsular soft tissues after shoulder instability operations, due to development of scar tissue, was also reported in a computed tomography-arthrographic study by Kreitner et al. [5]. Since fibrous scar tissue is not directly visible on radiographs, MRI is superior for
Fig. 4. Oblique-coronal T1-weighted spin-echo MRI image (TR = 600, TE=20 ms) of a patient 16 years after anterior glenoid bone block insertion. The bone block images as a mostly low signal intensity structure, consistent with predominantly cortical bone and sclerosis (arrow), surrounded by low signal intensity scar tissue, with thickening of the inferior joint capsule at the axilar recess (curved arrow). The signal intensity of the surrounding mass was also low on T2*-weighted images (not shown).
follow-up of patients with insufficient clinical results after such operations.
References [1] Eden R. Zur operation der habituellen schulterluxation unter mitteilung eines neuen verfahrens bei abriß am inneren pfannenrand. Dtsch Z Chir 1918;144:269 – 75. [2] Hybinette S. De la transplantation d’un fragment osseux pour reme´dier aux luxations re´cidivantes de l’e´paule: constatations et re´sults ope´ratoires. Acta Chir Scand 1932;71:411–20. [3] Owen RS, lannotti JP, Kneeland JB, Dalinka MK, Deren JA, Olega L. Shoulder after surgery: MR imaging with surgical validation. Radiology 1993;186:443 – 7. [4] Evancho AM, Stiles RG, Fajman WA, Flower SP, Macha T, Brunner MC, Fleming L. MR imaging diagnosis of rotator cuff tears. Am J Radiol 1988;151:751 – 4. [5] Kreitner KF, Lehmann M, Mildenberger P, Kirschner P, Schild HH. CT-Arthrographie der schulter nach operativer versorgung ventraler Instabilita¨ten. Fortschr Ro¨ntgenstr 1991;155:299–304. [6] Resnick D, Goergen TG. In: Resnick D, editor. Diagnosis of bone and joint disorders. Philadelphia, USA: Saunders Company, 1995:2700. [7] Bankart A. Recurrent or habitual dislocation of the shoulder joint. Br Med J 1923;2:1132 – 43. [8] Osmond-Clarke H. Recurrent dislocation of the shoulder: the Putti – Platt operation. J Bone J Surg Br Vol 1948;30:19–31. [9] Magnuson PB, Stack JK. Bilateral habitual dislocation of the shoulders in twins, afamily tendency. JAMA 1940;144:2103–14. [10] Oudard P. La luxation re´cidivante de l’e´paule (varie´te´ antero-interne). Procede´ ope´ratoire. J Chir 1924;23:1924–7. [11] Trillat A. Traitment de la luxation re´cidivante de l’e´paule: Considerations techniques. Lyon Chir 1954;49:986 –90. [12] Helfet AJ. Coracoid transplantation for recurring dislocation of the shoulder. Bone J Surg Br Vol 1958;40:198 –203.
Shoulder MRI after surgical treatment of instability Martin Vahlensieck a,*, Philipp Lang b, Ulli Wagner c, Frank Mo¨ller c, Urs van Deimling c, H.K. Genant b, Hans H. Schild a b
a Uni6ersity of Bonn, Department of Radiology, Sigmund – Freud-Strasse 25, 53105 Bonn, Germany Uni6ersity of California San Francisco, Department of Radiology, 505 Pamassus A6enue, San Francisco, CA 94143, USA c Uni6ersity of Bonn, Department of Orthopedic Surgery, Sigmund – Freud-Strasse 25, 53105 Bonn, Germany
Received 17 June 1998; received in revised form 7 October 1998; accepted 8 October 1998
Abstract Objecti6e: To analyze magnetic resonance imaging (MRI) findings of the shoulder after an instability operation. Materials and methods: Physical examinations, radiographs and MRI of 10 patients after anterior glenoid bone block insertion for ventral instability were compared. MRI included T1-weighted spin-echo (TR = 600, TE = 20 ms) and T2*-weighted gradient-echo sequences (TE = 600, TE = 18, Flip =30°) in the axial, oblique-coronal and oblique-sagittal planes. Results: No patient suffered from recurrent subluxation. We found fusion of the bone block with the anterior glenoid in seven cases, dislocation of the bone block without contact to the glenoid in one case, and no visible bone block in two cases. On MRI, the bone block showed either signal intensity equivalent to fatty bone marrow (n = 4) or was devoid of signal consistent with cortical bone or bone sclerosis (n= 4). In all patients, a low signal intensity mass, 2 – 4 cm in diameter, was visible next to the glenoid insertion site. Conclusion: Insertion of a bone block onto the anterior glenoid induces formation of scar tissue, increasing the stability of the shoulder joint. This scar is well visible on MRI and forms independently of the behavior of the bone block itself. MRI is ideally suited for evaluating postoperative shoulder joints after bone-grafting procedures. © 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Shoulder MRI; Surgery; Postoperative MRI; Instability
1. Introduction
2. Material and methods
One possible surgical procedure in the management of shoulder instability is the insertion of a bone block onto the anterior glenoid (Eden – Hybinette) [1,2] (Table 1). Bone block resorption or dislocation, seen clearly on radiographs, surprisingly does not impair functional long-term results. We therefore wanted to study the long-term soft-tissue changes of the shoulder after this operation by means of magnetic resonance imaging (MRI) in order to elucidate the therapeutic principle and to establish MRI criteria for follow-up.
We examined 10 patients with prior history of an Eden–Hybinette operation. The operated patients consisted of seven males and three females, 30–53 years old (mean 45 years). The operation had been performed 15–17 years before MRI. History and physical examination were compared with conventional radiographs. MRI of the shoulder was performed using a standard surface loop coil and T1-weighted spin-echo (TR =600 ms, TE= 20 ms) and T2*-weighted gradient-echo sequences (TR= 600, TE=18 ms, Flip=30°) in the coronal, sagittal, and axial plane. We used a 1.5 Tesla superconducting system (ACS, Philips Medical), a fieldof-view of 180 mm2, a 205× 256 matrix, and 4 mm slice thickness. Images were particularly assessed for contour and signal abnormalities of the bony glenoid, the ante-
* Corresponding author. Tel.: +49 228 2875875; fax: + 49 228 2876793.
0720-048X/99/$ - see front matter © 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S0720-048X(98)00146-6
M. Vahlensieck et al. / European Journal of Radiology 30 (1999) 2–4
3
rior glenoid labrum, the visibility, form, and signal intensity of the bone block, and the configuration and signal intensity of the surrounding soft tissues. Analysis was performed by two radiologists by consensus reading.
3. Results All patients had no recurrent luxation after the operation, and physical examination yielded no abnormality. Radiographs of these patients showed solid ingrowth of the bone into the anterior glenoid in seven cases, bone block dislocation without contact to the glenoid in one case, and no visible bone block in two cases. Two patients showed osteoarthritis of the shoulder joint, and in one case, a Hill – Sachs defect was visible. The MRI demonstrated high-signal intensity of the bone block in four cases, consistent with fatty bone marrow, and low signal intensity in four cases, consistent with cortical or sclerotic bone. In each case, a low signal intensity mass ventral to the anterior glenoid was visible. This mass was 2 – 4 cm in diameter. In two cases, we found multifocal signal voids within the mass which were emphasized on gradient-echo images. The mass itself and its relation to the capsular mechanism and bony glenoid was best visible on axial images (Figs. 1 and 2). However, it was also appreciable on coronal and sagittal views (Figs. 3 and 4). The ventral capsular mechanism was visible as a low signal intensity band in all patients. The anterior labrum was not visible in seven cases, was blunted in one case, and was triangular in two cases.
Fig. 1. Axial T1-weighted spin-echo MRI image (TR = 600, TE = 20 ms) of a patient 15 years after anterior glenoid bone block insertion. Adjacent to the anterior glenoid, a large low signal intensity mass is visible (curved arrow). This mass also showed low signal intensity on T*-weighted images (not shown). The bone block itself was resorbed 2 and was not visible on radiographs and MRI images. Within the anterior glenoid, some high signal intensity suggests fatty bone marrow at the operation site (arrow)
4. Discussion Knowledge of normal MRI findings in postoperative shoulder joints is important in order to delineate complications, and in order to detect persistent or recurrent disease from an uncomplicated course. Only a few articles describe such findings [3,4], particularly after surgery for rotator cuff pathology. In the present study, we analyzed shoulder MRI after surgery for anterior instability using bone block insertion.
Table 1 Some types of surgical procedures used for recurrent anterior Glenohumeral joint dislocations (from Ref. [6]) Procedure
Technique
Bankart [7]
Repair of anterior capsular mechanism using drill holes and sutures Putti–Platt [8] Shortening of the anterior capsule and subscapularis muscle Magnuson–Stack [9] Transfer of the subscapularis tendon from lesser tuberosity to greater tuberosity Eden–Hybinette [1,2] Bone graft to anterior glenoid region Oudard [10] Bone graft to coracoid process Trillat [11] Osteotomy with displacement of coracoid process Bristow–Helfet [12] Transfer of coracoid process with the first attached tendons to the neck of the scapula
Fig. 2. Schematic drawings of axial shoulder MRI sections in a healthy subject (a) and after anterior gienoid bone block insertion (b). A fatty marrow equivalent, high signal intensity bone block is surrounded by a large, low signal intensity scar.
4
M. Vahlensieck et al. / European Journal of Radiology 30 (1999) 2–4
Fig. 3. Oblique-sagittal T1-weighted spin-echo MRI image (TR = 600, TE=20 ms) of a patient 16 years after anterior glenoid bone block insertion. Anteriorly to the glenoid fossa, lies the bone block with signal intensity equivalent to fatty marrow (curved arrow) surrounded by low signal intensity scar tissue (arrows). This tissue was also of low signal on T*-weighted images (not shown). C, coracoid process. 2
The insertion of a bone block onto the anterior glenoid seems to induce the development of a soft-tissue mass, which, given its low signal intensity on T1- and T*-weighted images, appears to consist of fibrous scar 2 tissue. Since this mass develops also in the cases of bone block resorption and dislocation, it is considered to be a regular postoperative finding, and most suggestive of importance for the therapeutic principle in the Eden– Hybinette operation. The fibrous mass obviously acts as mechanical stabilization of the shoulder and prevents recurrent luxation. However, since we did not encounter patients in whom the operation failed, the last consideration cannot be proven confidently. The radiographical appearance of the bone block itself cannot be used as an indicator for therapeutic success, since persistent, resorbed and dislocated bone block were encountered independently from the clinical outcome. One limitation of our study is the lack of histologic confirmation of the ventral mass to represent fibrous scar tissue. However, due to the signal behavior of low intensity on both T1- and T2-weighted MRI images, as well as the location next to the bone block insertion site, no other conclusion seems to be likely. Shrinking of the ventral capsular soft tissues after shoulder instability operations, due to development of scar tissue, was also reported in a computed tomography-arthrographic study by Kreitner et al. [5]. Since fibrous scar tissue is not directly visible on radiographs, MRI is superior for
Fig. 4. Oblique-coronal T1-weighted spin-echo MRI image (TR = 600, TE=20 ms) of a patient 16 years after anterior glenoid bone block insertion. The bone block images as a mostly low signal intensity structure, consistent with predominantly cortical bone and sclerosis (arrow), surrounded by low signal intensity scar tissue, with thickening of the inferior joint capsule at the axilar recess (curved arrow). The signal intensity of the surrounding mass was also low on T2*-weighted images (not shown).
follow-up of patients with insufficient clinical results after such operations.
References [1] Eden R. Zur operation der habituellen schulterluxation unter mitteilung eines neuen verfahrens bei abriß am inneren pfannenrand. Dtsch Z Chir 1918;144:269 – 75. [2] Hybinette S. De la transplantation d’un fragment osseux pour reme´dier aux luxations re´cidivantes de l’e´paule: constatations et re´sults ope´ratoires. Acta Chir Scand 1932;71:411–20. [3] Owen RS, lannotti JP, Kneeland JB, Dalinka MK, Deren JA, Olega L. Shoulder after surgery: MR imaging with surgical validation. Radiology 1993;186:443 – 7. [4] Evancho AM, Stiles RG, Fajman WA, Flower SP, Macha T, Brunner MC, Fleming L. MR imaging diagnosis of rotator cuff tears. Am J Radiol 1988;151:751 – 4. [5] Kreitner KF, Lehmann M, Mildenberger P, Kirschner P, Schild HH. CT-Arthrographie der schulter nach operativer versorgung ventraler Instabilita¨ten. Fortschr Ro¨ntgenstr 1991;155:299–304. [6] Resnick D, Goergen TG. In: Resnick D, editor. Diagnosis of bone and joint disorders. Philadelphia, USA: Saunders Company, 1995:2700. [7] Bankart A. Recurrent or habitual dislocation of the shoulder joint. Br Med J 1923;2:1132 – 43. [8] Osmond-Clarke H. Recurrent dislocation of the shoulder: the Putti – Platt operation. J Bone J Surg Br Vol 1948;30:19–31. [9] Magnuson PB, Stack JK. Bilateral habitual dislocation of the shoulders in twins, afamily tendency. JAMA 1940;144:2103–14. [10] Oudard P. La luxation re´cidivante de l’e´paule (varie´te´ antero-interne). Procede´ ope´ratoire. J Chir 1924;23:1924–7. [11] Trillat A. Traitment de la luxation re´cidivante de l’e´paule: Considerations techniques. Lyon Chir 1954;49:986 –90. [12] Helfet AJ. Coracoid transplantation for recurring dislocation of the shoulder. Bone J Surg Br Vol 1958;40:198 –203.