- Email: [email protected]
Mode of failure for rotator cuff repair with suture anchors identified at revision surgery
Mode of failure for rotator cuff repair with suture anchors identified at revision surgery Craig A. Cummins, MD,a,b and George A. C. Murrell, MBBS, DPhil,b Barrington, IL, and Kogarah, Australia
Rotator cuff tears are a common cause of shoulder pain and dysfunction. After surgical repair, there is a significant re-tear rate (25%-90%). The aim of this study was to determine the primary mode of mechanical failure for rotator cuffs repaired with suture anchors at the time of revision rotator cuff repair. We prospectively followed 342 consecutive torn rotator cuffs, repaired by a single surgeon using suture anchors and a mattress-suturing configuration. Of those shoulders, 21 (6%) subsequently underwent a revision rotator cuff repair by the original surgeon, and 1 underwent a second revision repair. Intraoperative findings, including the mode of failure, were systematically recorded at revision surgery and compared with the findings at the primary repair. In addition, 81 primary rotator cuff repairs had a radiographic and fluoroscopic evaluation at a mean of 37 weeks after repair to assess for any loosening or migration of the anchors. At revision rotator cuff repair, the predominant mode of failure was tendon pulling through sutures (19/22 shoulders) (P ⬍ .001). Two recurrent tears occurred in a new location adjacent to the previous repair, and one anchor was found loose in the supraspinatus tendon. The mean size of the rotator cuff tear was larger at the revision surgery (P ⫽ .043), the tendon quality ranked poorer (P ⫽ .013), and the tendon mobility decreased (P ⫽ .002), as compared with the index procedure. The radiographs and fluoroscopic examination showed that all 335 anchors in 81 patients were in bone. Rotator cuff repairs with suture anchors that underwent revision surgery failed mechanically by three mechanisms, the most common of which was tendon pulling through sutures. This sugFrom the Lake Cook Orthopaedic Associates, Barrington, Ill,a and Orthopaedic Research Institute, Kogarah, Australia.b Funding provided by Mitek Australia and St George Hospital/ South Eastern Sydney Area Health Service. Reprint requests: Craig A. Cummins, MD, Lake Cook Orthopaedic Associates, 27401 W Hwy 22, Suite 125, Barrington, IL 60010 (E-mail: [email protected]). Copyright © 2003 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2003/$35.00 ⫹ 0 doi:10.1067/mse.2003.21
128
gests that the weak link in rotator cuff repairs with suture anchors and horizontal mattress sutures, as determined at revision surgery, is the tendon-suture interface. (J Shoulder Elbow Surg 2003;12:128-33.)
R
otator cuff tears are a common cause of shoulder pain and dysfunction. Symptoms include shoulder pain, pain and difficulty with overhead activities, and pain at night. Surgical repair attempts to restore the normal anatomic relationship of the rotator cuff tendons. Unfortunately, rotator cuff repair re-tear rates as high as 90%6 have been reported, with most studies demonstrating a re-tear rate from 25% to 35%.11,15,17,18,21 Patients whose rotator cuff repair fails have poorer shoulder function than those with intact repairs.11,15 Furthermore, clinical results deteriorate with repeated attempts at a rotator cuff repair.2,9 Potential causes for rotator cuff repairs retearing include anchors pulling out of bone, suture material breakage, surgical knot loosening, tendon pulling through sutures, or a tear developing in a new location. Although there have been investigations into the fixation strength of the tendon repair,1,5,13,14,16,19,20 to our knowledge, the in vivo mode(s) by which rotator cuff repairs re-tear has not been reported. The aim of this investigation was to determine the primary mode of mechanical failure of rotator cuff repairs with suture anchors in those patients who undergo revision rotator cuff tendon repairs. MATERIALS AND METHODS Three hundred forty-two consecutive patients with symptomatic rotator cuff tears underwent a rotator cuff repair between January 1996 and May 2001, by a single surgeon, with a standardized operative technique using suture anchors to reattach the torn tendon(s) to bone. In addition to the gathering of intraoperative data, patients were assessed systematically preoperatively and at 1 and 6 weeks, 3 and 6 months, and 2 years and as otherwise required postoperatively. During the postoperative course, patients were evaluated for a rotator cuff tendon re-tear if they (1) reported a traumatic event (eg, fall) followed by increased subjective pain and weakness or (2) had persistent, subjective pain and weakness, as well as weakness of the supraspinatus on manual muscle testing. In general, if a patient was chrono-
J Shoulder Elbow Surg Volume 12, Number 2
Cummins and Murrell
129
logically (⬍60 years) or physiologically younger, more active, or requiring improvement in function as well as pain relief, our tendency was to be more aggressive in evaluating for a rotator cuff re-tear. In this population we evaluated patients with new radiographs and ultrasound examinations. If the ultrasound showed a re-tear of the rotator cuff, a discussion was undertaken with the patient regarding nonsurgical and surgical treatment of the recurrent tendon tear. Those patients electing to proceed with a repeat attempt at rotator cuff repair then underwent a revision rotator cuff repair.
Surgical technique: Primary rotator cuff repair The surgical repair of the rotator cuff tear was performed open and included partial splitting of the deltoid in line with its fibers, detachment of the coracoacromial ligament at the level of the anterior acromion, and an anterior acromioplasty and bursectomy. The torn rotator cuff tendons were then repaired as closely as possible to their normal anatomic positions. If the mobility of the involved tendons was not sufficient to achieve a tension-free repair, appropriate soft-tissue releases were performed (subacromial and extraarticular adhesions, coracohumeral ligament, shoulder capsule, rotator cuff interval). Suture anchors were used to reattach the tendons to bone. The two suture anchors used in this investigation were the Mitek Rotator Cuff QuickAnchors (Mitek Surgical Products, Norwood, Mass) and Arthrotek Collarless Harpoon (3.2 mm) suture anchors (Biomet, Inc, Warsaw, Ind). In the majority of cases, the suture anchors were directly impacted into the bone of the proximal humerus without predrilling the bone. After insertion of the anchors, the suture material was tested in order to set the anchor against subchondral bone as well as to ensure secure initial bony fixation. For larger tears, two rows of suture anchors were used to allow dispersion of the forces across a larger surface area and increase the tendon-to-bone interface. The suture material used with the above anchors was No. 2, braided, nonabsorbable polyester suture. The tendons were grasped with the suture material by a horizontal mattress-stitch configuration (Figure 1). Bone troughs were not used in the repair. If a side-to-side repair was performed, a No. 2 Ethibond suture (Ethicon Inc, Edinburgh, Scotland) was used, with the suture being passed in a simple stitch pattern. Postoperatively, the patients were placed on a gradually progressive rehabilitation protocol. They were started on immediate passive range-of-motion pendulum exercises. After the first postoperative visit at day 8, they were started on active assisted forward flexion, external rotation, and abduction range-of-motion exercises. Active range of motion was emphasized at the 6-week postoperative office visit. Active overhead activities and lifting 5 kg or more were usually initiated 3 months postoperatively.
Data collection Intraoperative data were collected on standardized forms and included shoulder range of motion and stability. The cross-sectional size of the rotator cuff tear was estimated and recorded on a specifically designed diagram3 (Figure 2). The rotator cuff tendon quality and mobility were
Figure 1 Illustration demonstrating the horizontal mattress suture pattern used in this investigation to grasp the tendon during a rotator cuff repair.
rated as poor, fair, good, or excellent, with a numerical value attached to each grade (1, poor; 2, fair; 3, good; and 4, excellent). The number and type of suture anchors used in the rotator cuff repair, as well as any additional procedures, were recorded.
Revision rotator cuff repair The surgical procedure (positioning, approach, and exposure) for all revision rotator cuff repairs was similar to that of primary repair and was performed by the same surgeon. The intraoperative findings and mode of mechanical failure of the original repair were identified and recorded on the same data sheet used in the primary repair (Figure 2). The location of the recurrent rotator cuff tear was documented and compared with the site of the original tear. If a suture anchor was protruding from bone or loose in the soft tissues, the mode of failure was characterized as occurring at the anchor-bone interface. We were unable to assess whether suture anchor migration in bone had contributed to the tendon repair failure. If the suture material and surgical knot were intact but no longer located in the tendon, the mode of failure was characterized as occurring at the suture-tendon interface. We were not able to assess accurately whether the suture material had partially pulled through the tendon or whether the surgical knot had partially slipped but remained intact. If the suture material was broken or the knot had come apart, the mode of failure was recorded as such.
Radiographic suture anchor location Ethics approval was obtained from the institutional ethics committee for the radiographic portion of this investigation.
130
Cummins and Murrell
J Shoulder Elbow Surg March/April 2003
Figure 2 Illustration of the diagram used to record intraoperative findings with respect to the size and location of rotator cuff tears.
It was undertaken to determine whether the low incidence of suture anchor failure identified in the patients who went on to undergo revision rotator cuff repair was consistent with rotator cuff repair patients of the same surgeon who did not undergo a revision rotator cuff repair. This investigation involved 200 consecutive rotator cuff repair patients. These patients are a subset of the original 342 patients who met the inclusion criteria of at least 2 years’ follow-up. An extensive attempt was made to obtain at least 2 years’ follow-up on all 200 patients; however, we were only able to contact 154 of the patients. Four patients were known to be deceased, and two had severely dementia. All patients were offered radiographic and fluoroscopic imaging of their surgical shoulder. Eighty-one patients gave written informed consent and had the imaging tests performed. One of those 81 patients had an additional rotator cuff repair performed on the contralateral shoulder, and another patient went on to have a revision rotator cuff repair on the same shoulder. At the time of the imaging tests, the mean age of the patients was 68 ⫾ 11 years. Thirty-six patients were men and forty-seven were women. Sixty-two of the cases involved the right shoulder and twenty-one the left. The imaging tests were performed a mean of 37 ⫾ 9 months (range, 20-63 months) after the rotator cuff repair. The radiographic images obtained included an anteriorposterior radiograph of the glenohumeral joint, in neutral rotation, both with the arm at the patient’s side and with the arm abducted 20°. In addition, real-time fluoroscopic im-
ages of the involved shoulder were obtained to assess for loose suture anchors. Radiographic investigations were evaluated by a musculoskeletal radiologist. An anchor was characterized as loose if any portion of the anchor protruded from the bone. No attempt was made to assess for migration of the suture anchors in the bone.
Statistical methods Results are reported as mean values ⫾ SD. Comparisons between groups were made with Fisher exact tests, 2-way paired Student t tests, Wilcoxon rank sum tests, 1-way analysis of variance, and Tukey tests, with corrections made for multiple comparisons with the use of Sigma Stat, version 2.0 (Jandel Corporation, San Rafael, Calif). For all statistical tests, statistical significance was set at an ␣ error of .05 and a  error of .2.
RESULTS The mean age of patients at the time of primary rotator cuff repair was 65 ⫾ 13 years (range, 40-89 years). Revision rotator cuff repair
Of the original 342 patients, 21 (6%) subsequently underwent a revision rotator cuff repair by the same surgeon; 1 patient had a second revision surgery.
Cummins and Murrell
J Shoulder Elbow Surg Volume 12, Number 2
131
tissue quality at the time of revision, 9 had the same quality, and 2 had better tendon quality. In addition, tendon mobility at the time of revision repair was rated as poorer than at the primary repair (P ⫽ .002), with 13 cases having poorer tendon mobility, 5 having the same mobility, and 2 having better mobility at the revision surgery. Of the 22 revision rotator cuff repairs, 19 used Mitek suture anchors and 3 used Arthrotek Collarless Harpoon suture anchors. Radiographic suture anchor location
Figure 3 Distribution of rotator cuff repair mode of failure, as determined at the time of revision rotator cuff repair, in 22 revision cases: tendon pulling through suture (A), tear in a new location (B), suture anchors pulling out of bone (C), and suture breakage or suture knot (D). The modes of failure were compared with each other by the Fisher exact test (***P ⬍ .001).
There was no statistically significant difference between the age of the 21 patients who required a revision rotator cuff repair and the original 342 patients. Twelve of the patients were men and nine were women. Fourteen of the revision surgeries involved the right shoulder and eight the left. Fourteen patients recalled a precipitating event that they thought contributed to the tendon repair re-tearing; seven patients could not recall such an event, and in one patient this information was not recorded. The mean time between the primary and revision tendon repairs was 7.7 ⫾ 8.2 weeks (range, 1-35 weeks). The primary mode of failure identified at the time of revision surgery was tendon pulling through sutures (P ⬍ .001) (Figure 3); this mode of failure was identified in 19 of 22 revision cases. Of the 22 tendon re-tears, 2 occurred in a new area of the tendon adjacent to where the primary repair was performed. In only one case was a suture anchor not in bone but in the supraspinatus tendon. The etiology of the loose suture anchor may have been improper seating of the anchor into bone at the time of the original repair. This case involved a small tear measuring 0.5 cm2; because of the small size of the tear, visualization may not have been ideal during anchor insertion. No case of suture breakage was identified. The mean size of the rotator cuff tear was larger at the revision repair (8.9 ⫾ 7.5 cm2) than at the primary repair (6.3 ⫾ 5.6 cm2) (P ⫽ .043). In 15 cases, the tear was larger at revision surgery; in 2 cases, it was the same size; and in 5 cases, it was smaller. The rotator cuff tendon tissue was rated as being of poorer quality at revision surgery than at primary repair (P ⫽ .013); 11 of the cases demonstrated poorer tendon
The mean number of anchors used per surgical repair was 4.0 ⫾ 1.7 (range, 2-11 anchors/case). Overall, a total of 335 anchors were implanted in those 81 patients (Mitek anchors in 316 and Arthrotek Collarless Harpoon anchors in 19). The intraoperative findings in these patients included a mean rotator cuff tear size of 4.4 ⫾ 4.5 cm2 (range, 1-25 cm2), a tendon tissue quality rating of 2.6 ⫾ 0.9, and a tendon mobility rating of 2.8 ⫾ 1.0 (range, 1-4) (Figure 4). The findings of this investigation demonstrated that all suture anchors were confined to the bone of the proximal humerus; no portion of an anchor protruded from the bone on imaging tests. DISCUSSION The results of this investigation identified three mechanisms of mechanical failure of rotator cuff tendon repairs at revision surgery, including suture pulling through tendon, a tear in a new location, and suture anchor pulling out of bone. In 19 of the 22 revision rotator cuff repairs, the mode by which the original repair failed was tendon pulling through sutures. In the majority of rotator cuff tendon re-tears, the sutures, surgical knots, and anchors were intact but were no longer within the tendon. This suggests that the weak link in rotator cuff repairs with suture anchors was the tendon-suture interface. The answer to the question of whether these results can be generalized to all rotator cuff repairs in which a similar repair technique is used is undetermined. These in vivo results are consistent with the authors’ ex vivo investigations with an ovine subscapularis model, in which the predominant mode of failure with the use of suture anchors and a mattress suture configuration was tendon pulling through the sutures. Initial fixation strength could be enhanced by using more anchors or more complex suture patterns in the tendon.7,8 In two cases the rotator cuff tendon tear was identified in a new location adjacent to the area that was previously repaired. Although this was not truly a re-tear of the surgical repair, it was a re-tear of a previously repaired rotator cuff tendon and, therefore, was included in the analysis. It is possible that
132
Cummins and Murrell
J Shoulder Elbow Surg March/April 2003
Figure 4 Graphs demonstrating rotator cuff tear size (square centimeters) (A), tendon quality (B), and tendon mobility (C), at the time of rotator cuff repair, in 81 subjects, imaged for anchor location 37 @ 9 weeks after rotator cuff repair.
these re-tears occurred because the rotator cuff tendon degeneration was more diffuse than the localized area of the tendon tear that was originally repaired. Because patients with rotator cuff tears are often elderly, the proximal humerus may be osteoporotic, resulting in poor fixation of the anchor to bone. On the basis of this premise and the results of ex vivo investigations,5,10,13,16 several authors have stated that suture anchors are contraindicated in elderly patients with osteoporotic bone.12 However, in our study only 1 of 22 revision rotator cuff repairs was associated with a loose suture anchor, and this may have been the result of poor placement rather than the anchor pulling out from the bone. The results of the radiographic portion of this investigation also support the findings that suture anchors pulling out of bone is an uncommon mechanism of mechanical failure, even in elderly patients. The strengths of the investigation relate to the data being collected in a prospective fashion and the same repair technique being consistently used in all tendon repairs. Furthermore, the results from the findings at revision surgery were confirmed, in part, by the findings from the radiographic portion of the investigation
and are also consistent with ex vivo investigations that demonstrated the site of mechanical failure of rotator cuffs repaired with suture anchors to be at the suturetendon interface.4,8 Both a criticism and strength of the study is that the results are only valid for rotator cuff repairs that use suture anchors and a mattress suture pattern. Although the applicability of the findings to other repair techniques is limited, decreasing the number of variables in the study resulted in more reproducible findings. A criticism of this investigation is that the mode of failure in patients who required revision rotator cuff repair may not be representative of the mode of failure in all patients who had mechanical failure of a rotator cuff repair. It is highly likely that many other patients had mechanical failure of the rotator cuff repair yet remained relatively asymptomatic and therefore did not undergo revision surgery.16 Furthermore, other patients may have become symptomatic after a mechanical failure of the rotator cuff repair but either they did not undergo revision repair or another surgeon performed the revision rotator cuff repair. It is possible that mechanical failure in these patients oc-
Cummins and Murrell
J Shoulder Elbow Surg Volume 12, Number 2
curs by a different mechanism. However, short of performing a second-look operation on all rotator cuff repair failures, documenting the mode of failure at revision surgery was the most accurate method that we could think of to assess how tendon repairs fail in vivo. Another criticism of this study is that we were not able to assess accurately anchor migration in bone, partial failure at the suture-tendon interface, or surgical knot loosening. Potentially, these mechanisms could lead to the tendon no longer being apposed to the bone of the greater tuberosity. Theoretically, one could evaluate anchor migration postoperatively by carefully controlling radiographic positioning of patients. As this attempts to identify 3-dimensional anchor migration on 2-dimensional images, it would require detailed computer programming and mapping technology to be performed with reasonable accuracy. This evaluation was not performed in this investigation. Despite this limitation, the radiographic portion of the investigation demonstrated all anchors to be confined to the bone of the proximal humerus. In summary, this study has demonstrated that the predominant mode of failure in patients undergoing revision surgery after a rotator cuff repair with suture anchors and mattress sutures was suture pulling through tendon. In this investigation, suture anchors pulling out of bone did not play a significant role in recurrent tendon tears, even in elderly patients. We are grateful to all patients who participated in the study and to Dr Carl Bryant, St George Private Hospital Imaging, for radiologic evaluation. REFERENCES
1. Amis AA. The strength of artificial ligament anchorages. A comparative experimental study. J Bone Joint Surg Br 1988;70:397403. 2. Bigliani LU, Cordasco FA, McIlveen SJ, Musso ES. Operative treatment of failed repairs of the rotator cuff. J Bone Joint Surg Am 1992;74:1505-15. 3. Bryant L, Shnier R, Bryant C, Murrell GAC. A comparison of clinical estimation, ultrasonography, magnetic resonance imaging, and arthroscopy in determining the size of rotator cuff tears. J Shoulder Elbow Surg 2002;11:219-24. 4. Burkhart SS, Pagan JLD, Wirth MA, Athanasiou KA. Cyclic loading of anchor-based rotator cuff repairs: confirmation of the
5. 6. 7.
8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
133
tension overload phenomenon and comparison of suture anchor fixation with transosseous fixation. Arthroscopy 1997;13:720-4. Caldwell GL, Warner JJP, Miller MD, et al. Strength of fixation with transosseous sutures in rotator cuff repair. J Bone Joint Surg Am 1997;79:1064-8. Calvert PT, Packer NP, Stoker DJ, Bayley JIL, Kessel L. Arthrography of the shoulder after operative repair of the torn rotator cuff. J Bone Joint Surg Br 1986;68:147-50. Cummins CA, Strickland S, Appleyard RC, et al. Rotator cuff repair: an ex vivo mechanical study comparing trans-osseous sutures, suture anchors and bio-absorbable screws. Presented at Orthopaedic Research Society; Feb; Dallas, Tex. Cummins CA, Strickland S, Appleyard RC, et al. Rotator cuff repair: an ex vivo mechanical study comparing trans-osseous sutures, suture anchors and bio-absorbable screws. Presented at American Shoulder and Elbow Society Biennial Meeting; April 2002; Orlando, Fla. DeOrio JK, Cofield RH. Results of a second attempt at surgical repair of a failed initial rotator-cuff repair. J Bone Joint Surg Am 1984;66:563-7. France EP, Paulos LE, Harner CD, Straight CB. Biomechanical evaluation of rotator cuff fixation methods. Am J Sports Med 1989;17:176-81. Gazielly DF, Gleyze P, Montagnon C. Functional and anatomical results after rotator cuff repair. Clin Orthop Rel Res 1994; 304:43-53. Gerber C. Massive rotator cuff tears. In: Iannotti JP, Williams GR. Disorders of the shoulder: diagnosis and management. Philadelphia: Lippincott, Williams and Wilkins; 1988. p. 57-92. Gerber C, Schneeberger AG, Beck M, Schlegel URS. Mechanical strength of repairs of the rotator cuff. J Bone Joint Surg Br 1994;76:371-80. Gerber C, Schneeberger AG, Perren SM, Nyffeler RW. Experimental rotator cuff repair. A preliminary study. J Bone Joint Surg Am 1999;81:1281-90. Harryman DT, Mack LA, Wang KY, et al. Repairs of the rotator cuff. Correlation of functional results with integrity of the cuff. J Bone Joint Surg Am 1991;73:982-9. Hecker AT, Shea M, Hayhurst JO, et al. Pull-out strength of suture anchors for rotator cuff and Bankart lesion repairs. Am J Sports Med 1993;21:874-9. Knudsen HB, Gelineck J, Sojbjerg JO, et al. Functional and magnetic resonance imaging evaluation after single-tendon rotator cuff reconstruction. J Shoulder Elbow Surg 1999;8:242-6. Liu SH, Baker CL. Arthroscopically assisted rotator cuff repair: correlation of functional results with integrity of the cuff. Arthroscopy 1994;10:54-60. Robertson DB, Daniel DM, Biden E. Soft tissue fixation to bone. Am J Sports Med 1986;14:398-403. Sward L, Hughes JS, Amis AA, Wallace WA. The strength of surgical repairs of the rotator cuff. A biomechanical study on cadavers. J Bone Joint Surg Br 1992;74:585-8. Wulker N, Melzer C, Wirth CJ. Shoulder surgery for rotator cuff tears. Ultrasonographic 3-year follow-up of 97 cases. Acta Orthop Scand 1991;62:142-7.
Rotator cuff tears are a common cause of shoulder pain and dysfunction. After surgical repair, there is a significant re-tear rate (25%-90%). The aim of this study was to determine the primary mode of mechanical failure for rotator cuffs repaired with suture anchors at the time of revision rotator cuff repair. We prospectively followed 342 consecutive torn rotator cuffs, repaired by a single surgeon using suture anchors and a mattress-suturing configuration. Of those shoulders, 21 (6%) subsequently underwent a revision rotator cuff repair by the original surgeon, and 1 underwent a second revision repair. Intraoperative findings, including the mode of failure, were systematically recorded at revision surgery and compared with the findings at the primary repair. In addition, 81 primary rotator cuff repairs had a radiographic and fluoroscopic evaluation at a mean of 37 weeks after repair to assess for any loosening or migration of the anchors. At revision rotator cuff repair, the predominant mode of failure was tendon pulling through sutures (19/22 shoulders) (P ⬍ .001). Two recurrent tears occurred in a new location adjacent to the previous repair, and one anchor was found loose in the supraspinatus tendon. The mean size of the rotator cuff tear was larger at the revision surgery (P ⫽ .043), the tendon quality ranked poorer (P ⫽ .013), and the tendon mobility decreased (P ⫽ .002), as compared with the index procedure. The radiographs and fluoroscopic examination showed that all 335 anchors in 81 patients were in bone. Rotator cuff repairs with suture anchors that underwent revision surgery failed mechanically by three mechanisms, the most common of which was tendon pulling through sutures. This sugFrom the Lake Cook Orthopaedic Associates, Barrington, Ill,a and Orthopaedic Research Institute, Kogarah, Australia.b Funding provided by Mitek Australia and St George Hospital/ South Eastern Sydney Area Health Service. Reprint requests: Craig A. Cummins, MD, Lake Cook Orthopaedic Associates, 27401 W Hwy 22, Suite 125, Barrington, IL 60010 (E-mail: [email protected]). Copyright © 2003 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2003/$35.00 ⫹ 0 doi:10.1067/mse.2003.21
128
gests that the weak link in rotator cuff repairs with suture anchors and horizontal mattress sutures, as determined at revision surgery, is the tendon-suture interface. (J Shoulder Elbow Surg 2003;12:128-33.)
R
otator cuff tears are a common cause of shoulder pain and dysfunction. Symptoms include shoulder pain, pain and difficulty with overhead activities, and pain at night. Surgical repair attempts to restore the normal anatomic relationship of the rotator cuff tendons. Unfortunately, rotator cuff repair re-tear rates as high as 90%6 have been reported, with most studies demonstrating a re-tear rate from 25% to 35%.11,15,17,18,21 Patients whose rotator cuff repair fails have poorer shoulder function than those with intact repairs.11,15 Furthermore, clinical results deteriorate with repeated attempts at a rotator cuff repair.2,9 Potential causes for rotator cuff repairs retearing include anchors pulling out of bone, suture material breakage, surgical knot loosening, tendon pulling through sutures, or a tear developing in a new location. Although there have been investigations into the fixation strength of the tendon repair,1,5,13,14,16,19,20 to our knowledge, the in vivo mode(s) by which rotator cuff repairs re-tear has not been reported. The aim of this investigation was to determine the primary mode of mechanical failure of rotator cuff repairs with suture anchors in those patients who undergo revision rotator cuff tendon repairs. MATERIALS AND METHODS Three hundred forty-two consecutive patients with symptomatic rotator cuff tears underwent a rotator cuff repair between January 1996 and May 2001, by a single surgeon, with a standardized operative technique using suture anchors to reattach the torn tendon(s) to bone. In addition to the gathering of intraoperative data, patients were assessed systematically preoperatively and at 1 and 6 weeks, 3 and 6 months, and 2 years and as otherwise required postoperatively. During the postoperative course, patients were evaluated for a rotator cuff tendon re-tear if they (1) reported a traumatic event (eg, fall) followed by increased subjective pain and weakness or (2) had persistent, subjective pain and weakness, as well as weakness of the supraspinatus on manual muscle testing. In general, if a patient was chrono-
J Shoulder Elbow Surg Volume 12, Number 2
Cummins and Murrell
129
logically (⬍60 years) or physiologically younger, more active, or requiring improvement in function as well as pain relief, our tendency was to be more aggressive in evaluating for a rotator cuff re-tear. In this population we evaluated patients with new radiographs and ultrasound examinations. If the ultrasound showed a re-tear of the rotator cuff, a discussion was undertaken with the patient regarding nonsurgical and surgical treatment of the recurrent tendon tear. Those patients electing to proceed with a repeat attempt at rotator cuff repair then underwent a revision rotator cuff repair.
Surgical technique: Primary rotator cuff repair The surgical repair of the rotator cuff tear was performed open and included partial splitting of the deltoid in line with its fibers, detachment of the coracoacromial ligament at the level of the anterior acromion, and an anterior acromioplasty and bursectomy. The torn rotator cuff tendons were then repaired as closely as possible to their normal anatomic positions. If the mobility of the involved tendons was not sufficient to achieve a tension-free repair, appropriate soft-tissue releases were performed (subacromial and extraarticular adhesions, coracohumeral ligament, shoulder capsule, rotator cuff interval). Suture anchors were used to reattach the tendons to bone. The two suture anchors used in this investigation were the Mitek Rotator Cuff QuickAnchors (Mitek Surgical Products, Norwood, Mass) and Arthrotek Collarless Harpoon (3.2 mm) suture anchors (Biomet, Inc, Warsaw, Ind). In the majority of cases, the suture anchors were directly impacted into the bone of the proximal humerus without predrilling the bone. After insertion of the anchors, the suture material was tested in order to set the anchor against subchondral bone as well as to ensure secure initial bony fixation. For larger tears, two rows of suture anchors were used to allow dispersion of the forces across a larger surface area and increase the tendon-to-bone interface. The suture material used with the above anchors was No. 2, braided, nonabsorbable polyester suture. The tendons were grasped with the suture material by a horizontal mattress-stitch configuration (Figure 1). Bone troughs were not used in the repair. If a side-to-side repair was performed, a No. 2 Ethibond suture (Ethicon Inc, Edinburgh, Scotland) was used, with the suture being passed in a simple stitch pattern. Postoperatively, the patients were placed on a gradually progressive rehabilitation protocol. They were started on immediate passive range-of-motion pendulum exercises. After the first postoperative visit at day 8, they were started on active assisted forward flexion, external rotation, and abduction range-of-motion exercises. Active range of motion was emphasized at the 6-week postoperative office visit. Active overhead activities and lifting 5 kg or more were usually initiated 3 months postoperatively.
Data collection Intraoperative data were collected on standardized forms and included shoulder range of motion and stability. The cross-sectional size of the rotator cuff tear was estimated and recorded on a specifically designed diagram3 (Figure 2). The rotator cuff tendon quality and mobility were
Figure 1 Illustration demonstrating the horizontal mattress suture pattern used in this investigation to grasp the tendon during a rotator cuff repair.
rated as poor, fair, good, or excellent, with a numerical value attached to each grade (1, poor; 2, fair; 3, good; and 4, excellent). The number and type of suture anchors used in the rotator cuff repair, as well as any additional procedures, were recorded.
Revision rotator cuff repair The surgical procedure (positioning, approach, and exposure) for all revision rotator cuff repairs was similar to that of primary repair and was performed by the same surgeon. The intraoperative findings and mode of mechanical failure of the original repair were identified and recorded on the same data sheet used in the primary repair (Figure 2). The location of the recurrent rotator cuff tear was documented and compared with the site of the original tear. If a suture anchor was protruding from bone or loose in the soft tissues, the mode of failure was characterized as occurring at the anchor-bone interface. We were unable to assess whether suture anchor migration in bone had contributed to the tendon repair failure. If the suture material and surgical knot were intact but no longer located in the tendon, the mode of failure was characterized as occurring at the suture-tendon interface. We were not able to assess accurately whether the suture material had partially pulled through the tendon or whether the surgical knot had partially slipped but remained intact. If the suture material was broken or the knot had come apart, the mode of failure was recorded as such.
Radiographic suture anchor location Ethics approval was obtained from the institutional ethics committee for the radiographic portion of this investigation.
130
Cummins and Murrell
J Shoulder Elbow Surg March/April 2003
Figure 2 Illustration of the diagram used to record intraoperative findings with respect to the size and location of rotator cuff tears.
It was undertaken to determine whether the low incidence of suture anchor failure identified in the patients who went on to undergo revision rotator cuff repair was consistent with rotator cuff repair patients of the same surgeon who did not undergo a revision rotator cuff repair. This investigation involved 200 consecutive rotator cuff repair patients. These patients are a subset of the original 342 patients who met the inclusion criteria of at least 2 years’ follow-up. An extensive attempt was made to obtain at least 2 years’ follow-up on all 200 patients; however, we were only able to contact 154 of the patients. Four patients were known to be deceased, and two had severely dementia. All patients were offered radiographic and fluoroscopic imaging of their surgical shoulder. Eighty-one patients gave written informed consent and had the imaging tests performed. One of those 81 patients had an additional rotator cuff repair performed on the contralateral shoulder, and another patient went on to have a revision rotator cuff repair on the same shoulder. At the time of the imaging tests, the mean age of the patients was 68 ⫾ 11 years. Thirty-six patients were men and forty-seven were women. Sixty-two of the cases involved the right shoulder and twenty-one the left. The imaging tests were performed a mean of 37 ⫾ 9 months (range, 20-63 months) after the rotator cuff repair. The radiographic images obtained included an anteriorposterior radiograph of the glenohumeral joint, in neutral rotation, both with the arm at the patient’s side and with the arm abducted 20°. In addition, real-time fluoroscopic im-
ages of the involved shoulder were obtained to assess for loose suture anchors. Radiographic investigations were evaluated by a musculoskeletal radiologist. An anchor was characterized as loose if any portion of the anchor protruded from the bone. No attempt was made to assess for migration of the suture anchors in the bone.
Statistical methods Results are reported as mean values ⫾ SD. Comparisons between groups were made with Fisher exact tests, 2-way paired Student t tests, Wilcoxon rank sum tests, 1-way analysis of variance, and Tukey tests, with corrections made for multiple comparisons with the use of Sigma Stat, version 2.0 (Jandel Corporation, San Rafael, Calif). For all statistical tests, statistical significance was set at an ␣ error of .05 and a  error of .2.
RESULTS The mean age of patients at the time of primary rotator cuff repair was 65 ⫾ 13 years (range, 40-89 years). Revision rotator cuff repair
Of the original 342 patients, 21 (6%) subsequently underwent a revision rotator cuff repair by the same surgeon; 1 patient had a second revision surgery.
Cummins and Murrell
J Shoulder Elbow Surg Volume 12, Number 2
131
tissue quality at the time of revision, 9 had the same quality, and 2 had better tendon quality. In addition, tendon mobility at the time of revision repair was rated as poorer than at the primary repair (P ⫽ .002), with 13 cases having poorer tendon mobility, 5 having the same mobility, and 2 having better mobility at the revision surgery. Of the 22 revision rotator cuff repairs, 19 used Mitek suture anchors and 3 used Arthrotek Collarless Harpoon suture anchors. Radiographic suture anchor location
Figure 3 Distribution of rotator cuff repair mode of failure, as determined at the time of revision rotator cuff repair, in 22 revision cases: tendon pulling through suture (A), tear in a new location (B), suture anchors pulling out of bone (C), and suture breakage or suture knot (D). The modes of failure were compared with each other by the Fisher exact test (***P ⬍ .001).
There was no statistically significant difference between the age of the 21 patients who required a revision rotator cuff repair and the original 342 patients. Twelve of the patients were men and nine were women. Fourteen of the revision surgeries involved the right shoulder and eight the left. Fourteen patients recalled a precipitating event that they thought contributed to the tendon repair re-tearing; seven patients could not recall such an event, and in one patient this information was not recorded. The mean time between the primary and revision tendon repairs was 7.7 ⫾ 8.2 weeks (range, 1-35 weeks). The primary mode of failure identified at the time of revision surgery was tendon pulling through sutures (P ⬍ .001) (Figure 3); this mode of failure was identified in 19 of 22 revision cases. Of the 22 tendon re-tears, 2 occurred in a new area of the tendon adjacent to where the primary repair was performed. In only one case was a suture anchor not in bone but in the supraspinatus tendon. The etiology of the loose suture anchor may have been improper seating of the anchor into bone at the time of the original repair. This case involved a small tear measuring 0.5 cm2; because of the small size of the tear, visualization may not have been ideal during anchor insertion. No case of suture breakage was identified. The mean size of the rotator cuff tear was larger at the revision repair (8.9 ⫾ 7.5 cm2) than at the primary repair (6.3 ⫾ 5.6 cm2) (P ⫽ .043). In 15 cases, the tear was larger at revision surgery; in 2 cases, it was the same size; and in 5 cases, it was smaller. The rotator cuff tendon tissue was rated as being of poorer quality at revision surgery than at primary repair (P ⫽ .013); 11 of the cases demonstrated poorer tendon
The mean number of anchors used per surgical repair was 4.0 ⫾ 1.7 (range, 2-11 anchors/case). Overall, a total of 335 anchors were implanted in those 81 patients (Mitek anchors in 316 and Arthrotek Collarless Harpoon anchors in 19). The intraoperative findings in these patients included a mean rotator cuff tear size of 4.4 ⫾ 4.5 cm2 (range, 1-25 cm2), a tendon tissue quality rating of 2.6 ⫾ 0.9, and a tendon mobility rating of 2.8 ⫾ 1.0 (range, 1-4) (Figure 4). The findings of this investigation demonstrated that all suture anchors were confined to the bone of the proximal humerus; no portion of an anchor protruded from the bone on imaging tests. DISCUSSION The results of this investigation identified three mechanisms of mechanical failure of rotator cuff tendon repairs at revision surgery, including suture pulling through tendon, a tear in a new location, and suture anchor pulling out of bone. In 19 of the 22 revision rotator cuff repairs, the mode by which the original repair failed was tendon pulling through sutures. In the majority of rotator cuff tendon re-tears, the sutures, surgical knots, and anchors were intact but were no longer within the tendon. This suggests that the weak link in rotator cuff repairs with suture anchors was the tendon-suture interface. The answer to the question of whether these results can be generalized to all rotator cuff repairs in which a similar repair technique is used is undetermined. These in vivo results are consistent with the authors’ ex vivo investigations with an ovine subscapularis model, in which the predominant mode of failure with the use of suture anchors and a mattress suture configuration was tendon pulling through the sutures. Initial fixation strength could be enhanced by using more anchors or more complex suture patterns in the tendon.7,8 In two cases the rotator cuff tendon tear was identified in a new location adjacent to the area that was previously repaired. Although this was not truly a re-tear of the surgical repair, it was a re-tear of a previously repaired rotator cuff tendon and, therefore, was included in the analysis. It is possible that
132
Cummins and Murrell
J Shoulder Elbow Surg March/April 2003
Figure 4 Graphs demonstrating rotator cuff tear size (square centimeters) (A), tendon quality (B), and tendon mobility (C), at the time of rotator cuff repair, in 81 subjects, imaged for anchor location 37 @ 9 weeks after rotator cuff repair.
these re-tears occurred because the rotator cuff tendon degeneration was more diffuse than the localized area of the tendon tear that was originally repaired. Because patients with rotator cuff tears are often elderly, the proximal humerus may be osteoporotic, resulting in poor fixation of the anchor to bone. On the basis of this premise and the results of ex vivo investigations,5,10,13,16 several authors have stated that suture anchors are contraindicated in elderly patients with osteoporotic bone.12 However, in our study only 1 of 22 revision rotator cuff repairs was associated with a loose suture anchor, and this may have been the result of poor placement rather than the anchor pulling out from the bone. The results of the radiographic portion of this investigation also support the findings that suture anchors pulling out of bone is an uncommon mechanism of mechanical failure, even in elderly patients. The strengths of the investigation relate to the data being collected in a prospective fashion and the same repair technique being consistently used in all tendon repairs. Furthermore, the results from the findings at revision surgery were confirmed, in part, by the findings from the radiographic portion of the investigation
and are also consistent with ex vivo investigations that demonstrated the site of mechanical failure of rotator cuffs repaired with suture anchors to be at the suturetendon interface.4,8 Both a criticism and strength of the study is that the results are only valid for rotator cuff repairs that use suture anchors and a mattress suture pattern. Although the applicability of the findings to other repair techniques is limited, decreasing the number of variables in the study resulted in more reproducible findings. A criticism of this investigation is that the mode of failure in patients who required revision rotator cuff repair may not be representative of the mode of failure in all patients who had mechanical failure of a rotator cuff repair. It is highly likely that many other patients had mechanical failure of the rotator cuff repair yet remained relatively asymptomatic and therefore did not undergo revision surgery.16 Furthermore, other patients may have become symptomatic after a mechanical failure of the rotator cuff repair but either they did not undergo revision repair or another surgeon performed the revision rotator cuff repair. It is possible that mechanical failure in these patients oc-
Cummins and Murrell
J Shoulder Elbow Surg Volume 12, Number 2
curs by a different mechanism. However, short of performing a second-look operation on all rotator cuff repair failures, documenting the mode of failure at revision surgery was the most accurate method that we could think of to assess how tendon repairs fail in vivo. Another criticism of this study is that we were not able to assess accurately anchor migration in bone, partial failure at the suture-tendon interface, or surgical knot loosening. Potentially, these mechanisms could lead to the tendon no longer being apposed to the bone of the greater tuberosity. Theoretically, one could evaluate anchor migration postoperatively by carefully controlling radiographic positioning of patients. As this attempts to identify 3-dimensional anchor migration on 2-dimensional images, it would require detailed computer programming and mapping technology to be performed with reasonable accuracy. This evaluation was not performed in this investigation. Despite this limitation, the radiographic portion of the investigation demonstrated all anchors to be confined to the bone of the proximal humerus. In summary, this study has demonstrated that the predominant mode of failure in patients undergoing revision surgery after a rotator cuff repair with suture anchors and mattress sutures was suture pulling through tendon. In this investigation, suture anchors pulling out of bone did not play a significant role in recurrent tendon tears, even in elderly patients. We are grateful to all patients who participated in the study and to Dr Carl Bryant, St George Private Hospital Imaging, for radiologic evaluation. REFERENCES
1. Amis AA. The strength of artificial ligament anchorages. A comparative experimental study. J Bone Joint Surg Br 1988;70:397403. 2. Bigliani LU, Cordasco FA, McIlveen SJ, Musso ES. Operative treatment of failed repairs of the rotator cuff. J Bone Joint Surg Am 1992;74:1505-15. 3. Bryant L, Shnier R, Bryant C, Murrell GAC. A comparison of clinical estimation, ultrasonography, magnetic resonance imaging, and arthroscopy in determining the size of rotator cuff tears. J Shoulder Elbow Surg 2002;11:219-24. 4. Burkhart SS, Pagan JLD, Wirth MA, Athanasiou KA. Cyclic loading of anchor-based rotator cuff repairs: confirmation of the
5. 6. 7.
8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
133
tension overload phenomenon and comparison of suture anchor fixation with transosseous fixation. Arthroscopy 1997;13:720-4. Caldwell GL, Warner JJP, Miller MD, et al. Strength of fixation with transosseous sutures in rotator cuff repair. J Bone Joint Surg Am 1997;79:1064-8. Calvert PT, Packer NP, Stoker DJ, Bayley JIL, Kessel L. Arthrography of the shoulder after operative repair of the torn rotator cuff. J Bone Joint Surg Br 1986;68:147-50. Cummins CA, Strickland S, Appleyard RC, et al. Rotator cuff repair: an ex vivo mechanical study comparing trans-osseous sutures, suture anchors and bio-absorbable screws. Presented at Orthopaedic Research Society; Feb; Dallas, Tex. Cummins CA, Strickland S, Appleyard RC, et al. Rotator cuff repair: an ex vivo mechanical study comparing trans-osseous sutures, suture anchors and bio-absorbable screws. Presented at American Shoulder and Elbow Society Biennial Meeting; April 2002; Orlando, Fla. DeOrio JK, Cofield RH. Results of a second attempt at surgical repair of a failed initial rotator-cuff repair. J Bone Joint Surg Am 1984;66:563-7. France EP, Paulos LE, Harner CD, Straight CB. Biomechanical evaluation of rotator cuff fixation methods. Am J Sports Med 1989;17:176-81. Gazielly DF, Gleyze P, Montagnon C. Functional and anatomical results after rotator cuff repair. Clin Orthop Rel Res 1994; 304:43-53. Gerber C. Massive rotator cuff tears. In: Iannotti JP, Williams GR. Disorders of the shoulder: diagnosis and management. Philadelphia: Lippincott, Williams and Wilkins; 1988. p. 57-92. Gerber C, Schneeberger AG, Beck M, Schlegel URS. Mechanical strength of repairs of the rotator cuff. J Bone Joint Surg Br 1994;76:371-80. Gerber C, Schneeberger AG, Perren SM, Nyffeler RW. Experimental rotator cuff repair. A preliminary study. J Bone Joint Surg Am 1999;81:1281-90. Harryman DT, Mack LA, Wang KY, et al. Repairs of the rotator cuff. Correlation of functional results with integrity of the cuff. J Bone Joint Surg Am 1991;73:982-9. Hecker AT, Shea M, Hayhurst JO, et al. Pull-out strength of suture anchors for rotator cuff and Bankart lesion repairs. Am J Sports Med 1993;21:874-9. Knudsen HB, Gelineck J, Sojbjerg JO, et al. Functional and magnetic resonance imaging evaluation after single-tendon rotator cuff reconstruction. J Shoulder Elbow Surg 1999;8:242-6. Liu SH, Baker CL. Arthroscopically assisted rotator cuff repair: correlation of functional results with integrity of the cuff. Arthroscopy 1994;10:54-60. Robertson DB, Daniel DM, Biden E. Soft tissue fixation to bone. Am J Sports Med 1986;14:398-403. Sward L, Hughes JS, Amis AA, Wallace WA. The strength of surgical repairs of the rotator cuff. A biomechanical study on cadavers. J Bone Joint Surg Br 1992;74:585-8. Wulker N, Melzer C, Wirth CJ. Shoulder surgery for rotator cuff tears. Ultrasonographic 3-year follow-up of 97 cases. Acta Orthop Scand 1991;62:142-7.