- Email: [email protected]
Increasing the insertion depth of suture anchors for rotator cuff repair does not improve biomechanical stability
Increasing the insertion depth of suture anchors for rotator cuff repair does not improve biomechanical stability Andrew T. Mahar, MS,a,b Bradford S. Tucker, MD,b Vidyadhar V. Upasani, BS,c Richard S. Oka, BA,a and Robert A. Pedowitz, MD, PhD,b San Diego, CA
The purpose of this study was to evaluate whether deeper-than-recommended insertion of a suture anchor within the rotator cuff footprint of human cadaveric humeri affects fixation characteristics. Metallic 5-mm screw-in anchors loaded with a single No. 2 suture were placed in the infraspinatus footprint of 8 human cadaveric humeri at standard and deep depths. Specimens were cyclically loaded from 10 to 45 N for 500 cycles and then loaded to failure. Cylic displacement, failure load, and failure mode were compared. All deep anchors became flush within a few cycles, and both anchor depths displaced and rotated at the bone surface. Displacement of the deep anchors was significantly greater than that of standard anchors. There was no difference in failure load. Cyclic testing showed significant displacement, regardless of anchor position, possibly leading to gap formation of the repair. Deep placement of suture anchors for increased purchase caused greater displacement and is not recommended. (J Shoulder Elbow Surg 2005;14: 626-630.)
T
he early postoperative integrity of the rotator cuff after repair is a major determinant of a good functional result. Previous studies have reported cuff defects after repair,8 and in some cases greater than 50% of repairs contain residual defects when more than just the supraspinatus is involved.7 These defects may be due, in part, to anchor movement, suture breakage, knot slippage, or suture pulling through tendon. Recently, Meyer et al9 reported lower failure loads when suture was tested by use of an anchor eyelet compared with a smooth hook. Rupp et al12 From the aOrthopedic Biomechanics Research Center, Children’s Hospital, and bDepartment of Orthopaedic Surgery and cSchool of Medicine, University of California, San Diego. Anchors and financial support supplied by Stryker Sports Medicine, Mahwah, NJ. Reprint requests: Andrew Mahar, MS, MC5054, 3020 Children’s Way, San Diego, CA 92123 (E-mail: [email protected]). Copyright © 2005 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2005/$30.00 doi:10.1016/j.jse.2005.03.011
626
also reported suture failure at the eyelet with sensitivity to the suture-eyelet orientation. Barber et al1 also found suture breakage to be the primary mode of failure upon anchor retrieval in a sheep model. These authors concluded that abrasive wear at the sutureeyelet interface was responsible for the failures and recommended suture-sparing channels or other methods by which to protect the suture. One possible method by which to shield the suture-eyelet interface would be to hinge the suture over the cortical margin so that the suture remains axially loaded in reference to the suture anchor. However, suture failures via cutting through bone was previously reported by Rossouw et al10 when sutures were draped over the greater tuberosity and placed under physiologic cyclic loading. However, this study focused more on anchor location in the greater tuberosity than anchor depth. To address the issue of anchor depth, a recent biomechanical study found that a deep anchor (6 mm below cortical margin) had higher failure loads than a standarddepth anchor (3 mm below cortical margin) in a cadaveric bovine model.5 However, the deep anchors were found to cut through the cortical margin within the first 10 cycles, a result similar to that reported by Rossouw et al. Roth et al11 reported failures associated with anchor migration to the cortical surface within the first 100 cycles that resulted in the anchor being partially exposed. A partially exposed anchor may then induce suture wear on the anchor eyelet, predisposing it to early failure, as well as possibly damaging the nearby cartilage and soft tissue. Thus, anchor migration, suture degradation at the anchor eyelet, and suture cutting through bone represent potential ways in which postoperative rehabilitation may cause residual repair deficiencies. The results from both Rossouw et al10 and Roth et al11 present concerns regarding the optimal position or depth for an anchor to withstand the early rehabilitation period. From an in vitro perspective, some cyclic testing involved loads up to 180 N,6 although the most common loading method was based on the estimation of submaximal load on individual anchors of Burkhart et al.4 The estimation of approximately 38 N of load per anchor repair may adequately simulate the early postoperative rehabili-
J Shoulder Elbow Surg Volume 14, Number 6
tation period. In their article, optimal placement required placing the suture anchor at 45° to the greater tuberosity surface, previously described as the dead man’s angle.3 Arthroscopic placement of anchors in the dead man’s angle can make accurate estimations of anchor depth difficult. It is possible that anchors placed arthroscopically achieve a variety of depths outside of what was surgically intended. However, this relationship has not been addressed in the biomechanical literature by use of a human cadaveric model with the specific independent variable being anchor depth. The purpose of this study was to evaluate whether placement of suture anchors in a deeper-than-recommended position within the rotator cuff footprint of human cadaveric humeri adversely affects the sutureanchor-bone interface. Specifically, the following hypotheses were tested: (1) suture anchors placed in a deep position would fail by cutting of the suture through the bone, and (2) suture anchors placed with the eyelet flush to the bone surface would fail by suture failure at the anchor eyelet. METHODS Soft tissue was dissected from 8 fresh-frozen human cadaveric proximal humeri (6 male and 2 female specimens; mean age, 87 ⫾ 3 years) to expose the rotator cuff footprint, taking care to avoid decortication of the insertion area. Bone mineral density (BMD) measurements were not available at the time of harvest. However, this patient population represents an osteoporotic group that would potentially have anchor migration as a result of poor bone quality and were, therefore, considered viable and relevant to the study scheme. Stryker 5-mm metallic screw-in anchors (Stryker Sports Medicine, Mahwah, NJ) loaded with a single No. 2 braided nonabsorbable polyester suture were placed in the infraspinatus footprint at two depths (standard and deep) (n ⫽ 8 tests per depth). The infraspinatus footprint was chosen because some of the specimens had tears of the supraspinatus with adjacent bone sclerosis. The infraspinatus tendon and footprint were intact in all specimens tested. Therefore, for experimental purposes, the infraspinatus footprint was used to improve the consistency of bone quality. Anchor depths were randomly placed in the footprint of each specimen into an anterior or posterior position in an attempt to account for varying levels of osteoporosis across specimens. The anchors were placed at least 1 cm apart and at an angle of 45° from the bone surface, per the recommended angle for anchor stability of Burkhart.3 Standard anchor depth was defined as the position where the laser line of the insertion device was flush with the surface of the bone. This placed the anchor eyelet 3 mm below the cortical surface. Deep anchors were inserted 3 mm deeper than the standard anchors (eyelet depth, 6 mm). Although anchor bite may influence surgeons to select anchor position, it was believed that bite was potentially too subjective. It was believed that consistent depth was better at defining anchor location within and across samples to
Mahar et al
627
Figure 1 Experimental test setup (by use of the synthetic humerus model).
address the hypothesis best. For all anchors, eyelets were aligned parallel to the direction of pull according to the laser line on the insertion device. Equal numbers of standard and deep anchor positions were evaluated in both anterior and posterior positions. To concentrate the suture-bone-anchor interface as the variable of failure, all sutures were tied with 11 square knots over a 6-cm-diameter dowel to create a closed loop. The suture loop was then attached to an MTS 858 machine (MTS Systems Corporation, Eden Prairie, MN) over a smooth metallic rod (Figure 1). For this testing, the machine’s maximum capacity was 2500 N with 16-bit response resolution and 12-bit acquisition resolution. The distal humerus was potted in a 2-part epoxy resin (Bondo-Marhyde, Atlanta, GA) and held in place with a custom-designed fixation jig. The sutures were placed under 10 N of preload with 0 mm of suture-anchor-bone displacement. The hand-tied knot was placed equidistant from the bone hole exit and the smooth metallic rod on the MTS machine during load application. Though not indicative of surgical knot placement, the knot was located away from the repair area to prevent knot impingement on the metallic rod or human anatomy. The location of the 2 suture strands touching the bone hole exit was recorded via an ink mark to describe the location of failure better. Each specimen was cyclically loaded from 10 to 45 N at 0.5 Hz to a maximum of 500 cycles and then loaded at 0.5 mm/s to failure. This loading scheme was based on the work of Burkhart et al4 that estimated each repair to withstand approximately 37.7 N of load during maximum force generation (302 N) of the supraspinatus/infraspinatus. The direction of pull was parallel to the surface of the bone and 90° from anchor insertion.3 This oblique orientation of loading more closely mimics the rotator cuff line of action when the anchor is directed 45° from the bone surface. Displacement (in millimeters) of the suture-anchor construct and force (in Newtons) were sampled at 10 Hz for the duration of each test. Displacement after 5 cycles and after cyclic testing were calculated, as was ultimate failure load.
628
Mahar et al
J Shoulder Elbow Surg November/December 2005
Figure 2 A, Deep (left) and standard (right) anchor position before testing. B, Deep (left) and standard (right) anchor position after testing indicating both translation to the cortical margin and rotation at the surface in the plane of loading.
When suture failure occurred, the site of failure was determined in reference to the anchor eyelet by use of the previously placed ink marks. Displacement data after 5 cycles and 500 cycles, as well as failure load, were analyzed by a 1-way analysis of variance (P ⬍ .05) (Statistica, Inc, Tulsa, OK).
RESULTS A single deep anchor failed during cyclic loading by pulling out of bone midway through the test. This test was subsequently eliminated from any statistical testing, leaving 7 samples in this group. During cyclic loading tests, all deep anchors became flush within 5 cycles or less, and both the deep and standard anchors not only displaced but also rotated at the bone surface (Figures 2, A and B). After only 5 cycles, displacement of the deep anchors (5.7 ⫾ 1.9 mm) approached, but did not achieve, statistical significance compared with the standard anchors (3.9 ⫾ 0.8 mm). At the completion of the cyclic loading tests, the mean total displacement of the deep anchors (8.4 ⫾ 2.4 mm) was significantly greater than that of the standard anchors (5.7 ⫾ 1.4 mm) (P ⬍ .02). During failure testing, the ultimate failure load of the deep anchors (144.4 ⫾ 13.9 N) was statistically similar to that of the standard position (142.9 ⫾ 18.9 N) (P ⫽ .9). The mode of ultimate failure for the 7 deep anchors was as follows: 6 constructs failed at the suture knot and 1 failed by anchor pullout from the bone. For the standard anchor group, 4 failed at the suture knot, 3 failed at the suture eyelet, and 1 failed by bone pullout. DISCUSSION Rotator cuff defects after repair have been reported and have been attributed to a variety of scenarios.7,8 Previous studies have postulated that anchor migration,11 suture failure at the anchor eyelet,9,12 or suture
cutting through the bone10,11 may be responsible for some of these residual deficiencies. Some surgeons may possibly be tempted to alleviate some of these issues by placing the anchor deeper within the greater tuberosity, which may be an especially attractive option in osteoporotic bone. It is also possible that surgeons, when using arthroscopic techniques, may inadvertently place the suture anchor deeper than the manufacturer’s recommendation. A previous study on suture anchor depth, in a bovine model, found that the suture materials in a deep anchor construct cut through the bone at the cortical margin.5 This study did not report evidence of anchor migration; however, there may have been less anchor migration as a result of the similarity of BMD between young bovine bone (0.8 g/cm3) and young human bone (0.76 g/cm3) (unpublished data; Upasani V; January 2005). The current study attempted to address the clinically relevant issue of anchor depth and its effect on suture-anchor-bone interface biomechanics in a human cadaveric model. Data from this study revealed significantly greater construct displacement for the deep anchors (approximately 8 mm) compared with standard anchors (approximately 6 mm) after 500 cycles of physiologic loading between 10 and 45 N. However, Burkhart et al4 described a tissue gap formation of greater than 3 mm as clinical failure, meaning that both depths in this study produced unacceptable amounts of construct displacement. In fact, when the first 5 cycles of displacement were examined, both deep anchors (approximately 6 mm) and standard anchors (approximately 4 mm) exceeded this 3-mm criterion for an acceptable repair. These large construct displacements may be related to the level of osteoporosis in each specimen. An attempt was made to counter this by using anchors at each depth within each specimen. The frequency of anchor pullout (12.5%) within the specimens indicates that the level of osteoporosis in the specimens may have
Mahar et al
J Shoulder Elbow Surg Volume 14, Number 6
been relatively high. However, the primary modes of failure were at either the knot or anchor eyelet and may indicate adequate bony purchase. Unfortunately, measurements of BMD were unavailable at the time of testing. There have been conflicting studies stating that BMD does not influence suture anchor pullout strength,2 whereas others report a high correlation between failure force and BMD.13 BMD measures will be used in subsequent studies from our group to understand the relationship better between BMD and suture anchor pullout. Despite these concerns in our study, a simple conclusion may still be drawn from the displacement data. With all other variables being constant (length of loop, type of knot, loading method, position of humerus, amount of suture elongation, amount of knot slippage), differences in construct displacement between anchors at different depths may be best attributed to translation and rotation of the anchor within the greater tuberosity. Fluoroscopic images to confirm this phenomenon were not taken but will be included in subsequent studies. Although the methodologic variables were kept constant across specimens, some may not necessarily reflect the clinical scenario. The loop used in this study was much larger (approximately 9⫻) than the diameter observed surgically. Loop elongation or loss of internal knot fixation could adversely affect the calculation of total construct displacement. However, Ethibond suture (Ethicon, Somerville, NJ) was selected for its high stiffness to limit stretch in the construct. The knot was backed up with 11 hand-tied surgeon’s knots to ensure a stable knot. Because physiologic loads (10-45 N) were used for this study, it was believed that this load scheme would not induce significant suture elongation or loss of knot security and that all construct displacement would be related to the suture-bone-anchor interface. There were no instances of adverse results during the cyclic phase of testing, other than a single sample in which the anchor was pulled from its insertion. However, anchor migration to the cortical surface or rotation at the cortical surface was noted for every anchor for every specimen. For these reasons, the total construct displacement should have been primarily related to anchor migration/rotation or suture cutting through bone (or both). Ultimate failure loads in this study were not influenced by anchor depth, and each withstood approximately 140 N of force. Burkhart et al4 estimated a total supraspinatus/infraspinatus force of 302 N, based on measurements of muscle crosssectional area and a force production constant. This group also estimated a load per suture of approximately 38 N. Thus, it would seem that either anchor position, when loaded with a single suture, should withstand the estimation of Burkhart
629
et al. The deep anchors failed either at the knot or by anchor pullout, indicating that the deep position protected the suture-eyelet interface to some degree. It was qualitatively noted after testing that the deep anchors had translated to the cortical margin and also appeared to have rotated in the plane of loading. Although the deep anchor position now replicated the initial standard anchor placement, the loading history and suture-eyelet interface demonstrated slight differences in mechanical behavior. The standard anchors also had instances of knot failure and anchor pullout. However, nearly 50% of standard anchors (3/8) failed at the sutureeyelet interface, confirming differences in mechanical behavior between anchor depths. This failure rate indicates a weak point that had similarly been reported by Meyer et al9 and Rupp et al.12 The standard anchors appeared to have translated and rotated at the cortical margin, resulting in an anchor head above the cortical margin, potentially damaging soft-tissue structures and articular cartilage. Whether anchors are delivered to a deep position purposely to increase purchase or incidentally because of arthroscopic placement difficulties, there appears to be no benefit derived from the deep placement in a human cadaveric model. In a similar study using a bovine model,5 data for deep anchors also showed early clinical failure, reinforcing the concern with this technique. It may be possible that a younger population with increased bone density may benefit from this procedure. There are no clear data indicating this, however. The early postoperative rehabilitation protocol may also influence anchor position within the greater tuberosity and should be carefully considered with regard to age and activity level of the patient. REFERENCES
1. Barber FA, Cawley P, Prudich JF. Suture anchor failure strength—an in vivo study. Arthroscopy 1993;9:647-52. 2. Barber FA, Feder SM, Burkhart SS, Ahrens J. The relationship of suture anchor failure and bone density to proximal humerus location: a cadaveric study. Arthroscopy 1997;13:340-5. 3. Burkhart SS. The deadman theory of suture anchors: observations along a south Texas fence line. Arthroscopy 1995;11: 119-23. 4. Burkhart SS, Wirth MA, Simonich M, et al. Knot security in simple sliding knots and its relationship to rotator cuff repair: how secure must the knot be? Arthroscopy 2000;16:202-7. 5. Bynum CK, Lee S, Mahar AT, Tasto J, Pedowitz RA. Failure mode of suture anchors as a function of insertion depth. Am J Sports Med 2005;33:1030-4. 6. Goradia VK, Mullen DJ, Boucher HR, Parks BG, O’Donnell JB. Cyclic loading of rotator cuff repairs: a comparison of bioabsorbable tacks with metal suture anchors and transosseous sutures. Arthroscopy 2001;17:360-4. 7. Harryman DT II, 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.
630
Mahar et al
8. Liu SH, Baker CL. Arthroscopically assisted rotator cuff repair: correlation of functional results with integrity of the cuff. Arthroscopy 1994;10:54-60. 9. Meyer DC, Nyffeler RW, Fucentese SF, Gerber C. Failure of suture material at suture anchor eyelets. Arthroscopy 2002;18:1013-9. 10. Rossouw DJ, McElroy BJ, Amis AA, Emery RJ. A biomechanical evaluation of suture anchors in repair of the rotator cuff. J Bone Joint Surg Br 1997;79:458-61. 11. Roth CA, Bartolozzi AR, Ciccotti MG, et al. Failure properties of
J Shoulder Elbow Surg November/December 2005
suture anchors in the glenoid and the effects of cortical thickness. Arthroscopy 1998;14:186-91. 12. Rupp S, Georg T, Gauss C, Kohn D, Seil R. Fatigue testing of suture anchors. Am J Sports Med 2002;30:239-47. 13. Tingart MJ, Apreleva M, Lehtinen J, Zurakowski D, Warner JJ. Anchor design and bone mineral density affect the pull-out strength of suture anchors in rotator cuff repair: which anchors are best to use in patients with low bone quality? Am J Sports Med. 2004;32:1466-73.
The purpose of this study was to evaluate whether deeper-than-recommended insertion of a suture anchor within the rotator cuff footprint of human cadaveric humeri affects fixation characteristics. Metallic 5-mm screw-in anchors loaded with a single No. 2 suture were placed in the infraspinatus footprint of 8 human cadaveric humeri at standard and deep depths. Specimens were cyclically loaded from 10 to 45 N for 500 cycles and then loaded to failure. Cylic displacement, failure load, and failure mode were compared. All deep anchors became flush within a few cycles, and both anchor depths displaced and rotated at the bone surface. Displacement of the deep anchors was significantly greater than that of standard anchors. There was no difference in failure load. Cyclic testing showed significant displacement, regardless of anchor position, possibly leading to gap formation of the repair. Deep placement of suture anchors for increased purchase caused greater displacement and is not recommended. (J Shoulder Elbow Surg 2005;14: 626-630.)
T
he early postoperative integrity of the rotator cuff after repair is a major determinant of a good functional result. Previous studies have reported cuff defects after repair,8 and in some cases greater than 50% of repairs contain residual defects when more than just the supraspinatus is involved.7 These defects may be due, in part, to anchor movement, suture breakage, knot slippage, or suture pulling through tendon. Recently, Meyer et al9 reported lower failure loads when suture was tested by use of an anchor eyelet compared with a smooth hook. Rupp et al12 From the aOrthopedic Biomechanics Research Center, Children’s Hospital, and bDepartment of Orthopaedic Surgery and cSchool of Medicine, University of California, San Diego. Anchors and financial support supplied by Stryker Sports Medicine, Mahwah, NJ. Reprint requests: Andrew Mahar, MS, MC5054, 3020 Children’s Way, San Diego, CA 92123 (E-mail: [email protected]). Copyright © 2005 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-2746/2005/$30.00 doi:10.1016/j.jse.2005.03.011
626
also reported suture failure at the eyelet with sensitivity to the suture-eyelet orientation. Barber et al1 also found suture breakage to be the primary mode of failure upon anchor retrieval in a sheep model. These authors concluded that abrasive wear at the sutureeyelet interface was responsible for the failures and recommended suture-sparing channels or other methods by which to protect the suture. One possible method by which to shield the suture-eyelet interface would be to hinge the suture over the cortical margin so that the suture remains axially loaded in reference to the suture anchor. However, suture failures via cutting through bone was previously reported by Rossouw et al10 when sutures were draped over the greater tuberosity and placed under physiologic cyclic loading. However, this study focused more on anchor location in the greater tuberosity than anchor depth. To address the issue of anchor depth, a recent biomechanical study found that a deep anchor (6 mm below cortical margin) had higher failure loads than a standarddepth anchor (3 mm below cortical margin) in a cadaveric bovine model.5 However, the deep anchors were found to cut through the cortical margin within the first 10 cycles, a result similar to that reported by Rossouw et al. Roth et al11 reported failures associated with anchor migration to the cortical surface within the first 100 cycles that resulted in the anchor being partially exposed. A partially exposed anchor may then induce suture wear on the anchor eyelet, predisposing it to early failure, as well as possibly damaging the nearby cartilage and soft tissue. Thus, anchor migration, suture degradation at the anchor eyelet, and suture cutting through bone represent potential ways in which postoperative rehabilitation may cause residual repair deficiencies. The results from both Rossouw et al10 and Roth et al11 present concerns regarding the optimal position or depth for an anchor to withstand the early rehabilitation period. From an in vitro perspective, some cyclic testing involved loads up to 180 N,6 although the most common loading method was based on the estimation of submaximal load on individual anchors of Burkhart et al.4 The estimation of approximately 38 N of load per anchor repair may adequately simulate the early postoperative rehabili-
J Shoulder Elbow Surg Volume 14, Number 6
tation period. In their article, optimal placement required placing the suture anchor at 45° to the greater tuberosity surface, previously described as the dead man’s angle.3 Arthroscopic placement of anchors in the dead man’s angle can make accurate estimations of anchor depth difficult. It is possible that anchors placed arthroscopically achieve a variety of depths outside of what was surgically intended. However, this relationship has not been addressed in the biomechanical literature by use of a human cadaveric model with the specific independent variable being anchor depth. The purpose of this study was to evaluate whether placement of suture anchors in a deeper-than-recommended position within the rotator cuff footprint of human cadaveric humeri adversely affects the sutureanchor-bone interface. Specifically, the following hypotheses were tested: (1) suture anchors placed in a deep position would fail by cutting of the suture through the bone, and (2) suture anchors placed with the eyelet flush to the bone surface would fail by suture failure at the anchor eyelet. METHODS Soft tissue was dissected from 8 fresh-frozen human cadaveric proximal humeri (6 male and 2 female specimens; mean age, 87 ⫾ 3 years) to expose the rotator cuff footprint, taking care to avoid decortication of the insertion area. Bone mineral density (BMD) measurements were not available at the time of harvest. However, this patient population represents an osteoporotic group that would potentially have anchor migration as a result of poor bone quality and were, therefore, considered viable and relevant to the study scheme. Stryker 5-mm metallic screw-in anchors (Stryker Sports Medicine, Mahwah, NJ) loaded with a single No. 2 braided nonabsorbable polyester suture were placed in the infraspinatus footprint at two depths (standard and deep) (n ⫽ 8 tests per depth). The infraspinatus footprint was chosen because some of the specimens had tears of the supraspinatus with adjacent bone sclerosis. The infraspinatus tendon and footprint were intact in all specimens tested. Therefore, for experimental purposes, the infraspinatus footprint was used to improve the consistency of bone quality. Anchor depths were randomly placed in the footprint of each specimen into an anterior or posterior position in an attempt to account for varying levels of osteoporosis across specimens. The anchors were placed at least 1 cm apart and at an angle of 45° from the bone surface, per the recommended angle for anchor stability of Burkhart.3 Standard anchor depth was defined as the position where the laser line of the insertion device was flush with the surface of the bone. This placed the anchor eyelet 3 mm below the cortical surface. Deep anchors were inserted 3 mm deeper than the standard anchors (eyelet depth, 6 mm). Although anchor bite may influence surgeons to select anchor position, it was believed that bite was potentially too subjective. It was believed that consistent depth was better at defining anchor location within and across samples to
Mahar et al
627
Figure 1 Experimental test setup (by use of the synthetic humerus model).
address the hypothesis best. For all anchors, eyelets were aligned parallel to the direction of pull according to the laser line on the insertion device. Equal numbers of standard and deep anchor positions were evaluated in both anterior and posterior positions. To concentrate the suture-bone-anchor interface as the variable of failure, all sutures were tied with 11 square knots over a 6-cm-diameter dowel to create a closed loop. The suture loop was then attached to an MTS 858 machine (MTS Systems Corporation, Eden Prairie, MN) over a smooth metallic rod (Figure 1). For this testing, the machine’s maximum capacity was 2500 N with 16-bit response resolution and 12-bit acquisition resolution. The distal humerus was potted in a 2-part epoxy resin (Bondo-Marhyde, Atlanta, GA) and held in place with a custom-designed fixation jig. The sutures were placed under 10 N of preload with 0 mm of suture-anchor-bone displacement. The hand-tied knot was placed equidistant from the bone hole exit and the smooth metallic rod on the MTS machine during load application. Though not indicative of surgical knot placement, the knot was located away from the repair area to prevent knot impingement on the metallic rod or human anatomy. The location of the 2 suture strands touching the bone hole exit was recorded via an ink mark to describe the location of failure better. Each specimen was cyclically loaded from 10 to 45 N at 0.5 Hz to a maximum of 500 cycles and then loaded at 0.5 mm/s to failure. This loading scheme was based on the work of Burkhart et al4 that estimated each repair to withstand approximately 37.7 N of load during maximum force generation (302 N) of the supraspinatus/infraspinatus. The direction of pull was parallel to the surface of the bone and 90° from anchor insertion.3 This oblique orientation of loading more closely mimics the rotator cuff line of action when the anchor is directed 45° from the bone surface. Displacement (in millimeters) of the suture-anchor construct and force (in Newtons) were sampled at 10 Hz for the duration of each test. Displacement after 5 cycles and after cyclic testing were calculated, as was ultimate failure load.
628
Mahar et al
J Shoulder Elbow Surg November/December 2005
Figure 2 A, Deep (left) and standard (right) anchor position before testing. B, Deep (left) and standard (right) anchor position after testing indicating both translation to the cortical margin and rotation at the surface in the plane of loading.
When suture failure occurred, the site of failure was determined in reference to the anchor eyelet by use of the previously placed ink marks. Displacement data after 5 cycles and 500 cycles, as well as failure load, were analyzed by a 1-way analysis of variance (P ⬍ .05) (Statistica, Inc, Tulsa, OK).
RESULTS A single deep anchor failed during cyclic loading by pulling out of bone midway through the test. This test was subsequently eliminated from any statistical testing, leaving 7 samples in this group. During cyclic loading tests, all deep anchors became flush within 5 cycles or less, and both the deep and standard anchors not only displaced but also rotated at the bone surface (Figures 2, A and B). After only 5 cycles, displacement of the deep anchors (5.7 ⫾ 1.9 mm) approached, but did not achieve, statistical significance compared with the standard anchors (3.9 ⫾ 0.8 mm). At the completion of the cyclic loading tests, the mean total displacement of the deep anchors (8.4 ⫾ 2.4 mm) was significantly greater than that of the standard anchors (5.7 ⫾ 1.4 mm) (P ⬍ .02). During failure testing, the ultimate failure load of the deep anchors (144.4 ⫾ 13.9 N) was statistically similar to that of the standard position (142.9 ⫾ 18.9 N) (P ⫽ .9). The mode of ultimate failure for the 7 deep anchors was as follows: 6 constructs failed at the suture knot and 1 failed by anchor pullout from the bone. For the standard anchor group, 4 failed at the suture knot, 3 failed at the suture eyelet, and 1 failed by bone pullout. DISCUSSION Rotator cuff defects after repair have been reported and have been attributed to a variety of scenarios.7,8 Previous studies have postulated that anchor migration,11 suture failure at the anchor eyelet,9,12 or suture
cutting through the bone10,11 may be responsible for some of these residual deficiencies. Some surgeons may possibly be tempted to alleviate some of these issues by placing the anchor deeper within the greater tuberosity, which may be an especially attractive option in osteoporotic bone. It is also possible that surgeons, when using arthroscopic techniques, may inadvertently place the suture anchor deeper than the manufacturer’s recommendation. A previous study on suture anchor depth, in a bovine model, found that the suture materials in a deep anchor construct cut through the bone at the cortical margin.5 This study did not report evidence of anchor migration; however, there may have been less anchor migration as a result of the similarity of BMD between young bovine bone (0.8 g/cm3) and young human bone (0.76 g/cm3) (unpublished data; Upasani V; January 2005). The current study attempted to address the clinically relevant issue of anchor depth and its effect on suture-anchor-bone interface biomechanics in a human cadaveric model. Data from this study revealed significantly greater construct displacement for the deep anchors (approximately 8 mm) compared with standard anchors (approximately 6 mm) after 500 cycles of physiologic loading between 10 and 45 N. However, Burkhart et al4 described a tissue gap formation of greater than 3 mm as clinical failure, meaning that both depths in this study produced unacceptable amounts of construct displacement. In fact, when the first 5 cycles of displacement were examined, both deep anchors (approximately 6 mm) and standard anchors (approximately 4 mm) exceeded this 3-mm criterion for an acceptable repair. These large construct displacements may be related to the level of osteoporosis in each specimen. An attempt was made to counter this by using anchors at each depth within each specimen. The frequency of anchor pullout (12.5%) within the specimens indicates that the level of osteoporosis in the specimens may have
Mahar et al
J Shoulder Elbow Surg Volume 14, Number 6
been relatively high. However, the primary modes of failure were at either the knot or anchor eyelet and may indicate adequate bony purchase. Unfortunately, measurements of BMD were unavailable at the time of testing. There have been conflicting studies stating that BMD does not influence suture anchor pullout strength,2 whereas others report a high correlation between failure force and BMD.13 BMD measures will be used in subsequent studies from our group to understand the relationship better between BMD and suture anchor pullout. Despite these concerns in our study, a simple conclusion may still be drawn from the displacement data. With all other variables being constant (length of loop, type of knot, loading method, position of humerus, amount of suture elongation, amount of knot slippage), differences in construct displacement between anchors at different depths may be best attributed to translation and rotation of the anchor within the greater tuberosity. Fluoroscopic images to confirm this phenomenon were not taken but will be included in subsequent studies. Although the methodologic variables were kept constant across specimens, some may not necessarily reflect the clinical scenario. The loop used in this study was much larger (approximately 9⫻) than the diameter observed surgically. Loop elongation or loss of internal knot fixation could adversely affect the calculation of total construct displacement. However, Ethibond suture (Ethicon, Somerville, NJ) was selected for its high stiffness to limit stretch in the construct. The knot was backed up with 11 hand-tied surgeon’s knots to ensure a stable knot. Because physiologic loads (10-45 N) were used for this study, it was believed that this load scheme would not induce significant suture elongation or loss of knot security and that all construct displacement would be related to the suture-bone-anchor interface. There were no instances of adverse results during the cyclic phase of testing, other than a single sample in which the anchor was pulled from its insertion. However, anchor migration to the cortical surface or rotation at the cortical surface was noted for every anchor for every specimen. For these reasons, the total construct displacement should have been primarily related to anchor migration/rotation or suture cutting through bone (or both). Ultimate failure loads in this study were not influenced by anchor depth, and each withstood approximately 140 N of force. Burkhart et al4 estimated a total supraspinatus/infraspinatus force of 302 N, based on measurements of muscle crosssectional area and a force production constant. This group also estimated a load per suture of approximately 38 N. Thus, it would seem that either anchor position, when loaded with a single suture, should withstand the estimation of Burkhart
629
et al. The deep anchors failed either at the knot or by anchor pullout, indicating that the deep position protected the suture-eyelet interface to some degree. It was qualitatively noted after testing that the deep anchors had translated to the cortical margin and also appeared to have rotated in the plane of loading. Although the deep anchor position now replicated the initial standard anchor placement, the loading history and suture-eyelet interface demonstrated slight differences in mechanical behavior. The standard anchors also had instances of knot failure and anchor pullout. However, nearly 50% of standard anchors (3/8) failed at the sutureeyelet interface, confirming differences in mechanical behavior between anchor depths. This failure rate indicates a weak point that had similarly been reported by Meyer et al9 and Rupp et al.12 The standard anchors appeared to have translated and rotated at the cortical margin, resulting in an anchor head above the cortical margin, potentially damaging soft-tissue structures and articular cartilage. Whether anchors are delivered to a deep position purposely to increase purchase or incidentally because of arthroscopic placement difficulties, there appears to be no benefit derived from the deep placement in a human cadaveric model. In a similar study using a bovine model,5 data for deep anchors also showed early clinical failure, reinforcing the concern with this technique. It may be possible that a younger population with increased bone density may benefit from this procedure. There are no clear data indicating this, however. The early postoperative rehabilitation protocol may also influence anchor position within the greater tuberosity and should be carefully considered with regard to age and activity level of the patient. REFERENCES
1. Barber FA, Cawley P, Prudich JF. Suture anchor failure strength—an in vivo study. Arthroscopy 1993;9:647-52. 2. Barber FA, Feder SM, Burkhart SS, Ahrens J. The relationship of suture anchor failure and bone density to proximal humerus location: a cadaveric study. Arthroscopy 1997;13:340-5. 3. Burkhart SS. The deadman theory of suture anchors: observations along a south Texas fence line. Arthroscopy 1995;11: 119-23. 4. Burkhart SS, Wirth MA, Simonich M, et al. Knot security in simple sliding knots and its relationship to rotator cuff repair: how secure must the knot be? Arthroscopy 2000;16:202-7. 5. Bynum CK, Lee S, Mahar AT, Tasto J, Pedowitz RA. Failure mode of suture anchors as a function of insertion depth. Am J Sports Med 2005;33:1030-4. 6. Goradia VK, Mullen DJ, Boucher HR, Parks BG, O’Donnell JB. Cyclic loading of rotator cuff repairs: a comparison of bioabsorbable tacks with metal suture anchors and transosseous sutures. Arthroscopy 2001;17:360-4. 7. Harryman DT II, 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.
630
Mahar et al
8. Liu SH, Baker CL. Arthroscopically assisted rotator cuff repair: correlation of functional results with integrity of the cuff. Arthroscopy 1994;10:54-60. 9. Meyer DC, Nyffeler RW, Fucentese SF, Gerber C. Failure of suture material at suture anchor eyelets. Arthroscopy 2002;18:1013-9. 10. Rossouw DJ, McElroy BJ, Amis AA, Emery RJ. A biomechanical evaluation of suture anchors in repair of the rotator cuff. J Bone Joint Surg Br 1997;79:458-61. 11. Roth CA, Bartolozzi AR, Ciccotti MG, et al. Failure properties of
J Shoulder Elbow Surg November/December 2005
suture anchors in the glenoid and the effects of cortical thickness. Arthroscopy 1998;14:186-91. 12. Rupp S, Georg T, Gauss C, Kohn D, Seil R. Fatigue testing of suture anchors. Am J Sports Med 2002;30:239-47. 13. Tingart MJ, Apreleva M, Lehtinen J, Zurakowski D, Warner JJ. Anchor design and bone mineral density affect the pull-out strength of suture anchors in rotator cuff repair: which anchors are best to use in patients with low bone quality? Am J Sports Med. 2004;32:1466-73.