- Email: [email protected]
Comparative Testing by Cyclic Loading of Rotator Cuff Suture Anchors Containing Multiple High-Strength Sutures
Comparative Testing by Cyclic Loading of Rotator Cuff Suture Anchors Containing Multiple High-Strength Sutures F. Alan Barber, M.D., Onur Hapa, M.D., and James A. Bynum, M.D. Purpose: To compare isolated medial-row with isolated lateral-row anchor performance by use of cyclic loading followed by destructive testing in an in vitro cadaveric model. Methods: Using 16 human cadaveric humeri without tendons, we rotated 4 medial-row (Bio-Corkscrew FT [Arthrex, Naples FL], CrossFT PK [ConMed Linvatec, Largo, FL], TwinFix PK FT [Smith & Nephew Endoscopy, Andover, MA], and Healix PK [DePuy Mitek, Raynham, MA]) and 4 lateral-row (PopLok PK [ConMed Linvatec], PEEK [polyetheretherketone] PushLock [Arthrex], Footprint PEEK [Smith & Nephew Endoscopy], and Versalok [DePuy Mitek]) anchors among different medial (articular cartilage edge) and lateral greater tuberosity sites (anterior, central, posterior). All medial anchors were inserted into the humeral head at an angle no greater than 45°. All lateral anchors were inserted “over the top,” nearly planar to the superior humeral surface. After preloading, the constructs were cycled 500 times from 10 to 60 N at 1 Hz with the loads applied to the accompanying sutures. Those constructs surviving cycling were destructively tested. Cyclic displacement, ultimate load, and failure mode were recorded. Results: In this laboratory setting, most displacement occurred in the first 100 cycles except for the Footprint anchor. Lateral-row anchors had greater mean displacements (2.6 mm) than medial-row anchors (1.2 mm) at 100 cycles and between 100 and 500 cycles (1.8 mm v 0.75 mm). Lateral-row anchors also had more total displacement (4.4 mm) than medial-row anchors (1.9 mm). A 5-mm displacement gap, defined as failure, was not seen in the Bio-Corkscrew FT, TwinFix PK FT, and Versalok anchors. Ultimate failure loads ranged from 163 N (Footprint) to 308 N (Versalok) (P ⬍ .05). The principal failure mode was anchor pullout, followed by eyelet breakage. Medial-row eyelet failures only occurred after 500 cycles at loads higher than each anchor’s mean failure load. Eyelet failure for lateral-row anchors occurred before 500 cycles and at failure loads lower than each anchor’s mean. Conclusions: Lateral row anchors benefit from medial row anchors for their security, and because of design differences demonstrate more displacement. When lateral-row anchors fail at the eyelet, it is at lower failure loads, while if medial-row anchors fail at the eyelet, it is at higher loads. Clinical Relevance: Anchors designed to function as lateral-row fixation provide fixation strength inferior to that of medial-row anchors and are more likely to be subject to suture slippage.
S
uture anchors loaded with multiple high-strength sutures are now commonly used for rotator cuff repair.1-3 Comparative cyclic load testing in line with the
From the Plano Orthopedic Sports Medicine and Spine Center (F.A.B.), Plano, Texas; A. Kagan Orthopedics and Sports Medicine (J.A.B.), Fort Myers, Florida, U.S.A.; and Department of Orthopaedics and Traumatology, Mustafa Kemal University (O.H.), Antakya, Turkey. Supported by ConMed Linvatec, Largo, Florida, and DePuy Mitek, Raynham, Massachusetts. The authors have received support from ConMed Linvatec, DePuy Mitek, and Smith & Nephew Endoscopy, Andover, Massachusetts, exceeding US $500 related to this research. Research was performed at ConMed Linvatec. Received September 20, 2009; accepted March 4, 2010. Address correspondence and reprint requests to F. Alan Barber, M.D., Plano Orthopedic Sports Medicine and Spine Center 5228 W Plano Pkwy, Plano, TX 75093, U.S.A. © 2010 by the Arthroscopy Association of North America 0749-8063/9544/$36.00 doi:10.1016/j.arthro.2010.03.007
axis of the tendon is a testing mode that is consistent with the clinical situation.4-6 Single-pull ultimate failure loads do not reflect the clinical condition. Cyclic loading such as occurs with arm motion during rehabilitation or during activities of daily living is more reflective of the stresses on the rotator cuff repair occurring during the postoperative rehabilitation period. Repair displacement of 5 to 10 mm with cyclic loads would in reality be a clinical failure. Double-row constructs have gained considerable attention related to the repair of rotator cuff tendons.7-15 Tests that reflect how suture anchor constructs perform under cyclic loads provide more practical information about how successful they might be when used clinically. However, little evidence exists that compares the performance of the typical medial-row anchor with the more recently developed lateral anchors. In addition, the potential exists for the use of lateral-row anchors as independent fixation devices.
S134 Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 26, No 9 (September, Suppl 1), 2010: pp S134-S141
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS
S135
The purpose of this study was to compare isolated medial-row anchor with isolated lateral-row anchor performance by use of cyclic loading followed by destructive testing in an in vitro cadaveric model. Our hypothesis was that suture anchors designed for medial- or lateral-row rotator cuff applications would show different biomechanical behavior when subjected to cyclic loading. METHODS We prepared 16 humeri from 8 human donors by stripping all the soft tissue from the bones. The ages of these humeri were recorded. No donors with bone metastases, generalized bone disease, lymphoma, myeloma, or leukemia were included. The mean age was 80 years (range, 70 to 96 years), and there was only 1 female donor. Three medial and three lateral sites were identified at the greater tuberosity of the proximal humerus for anchor insertion (Fig 1). The 3 medial sites were in the anterior, central, and posterior greater tuberosity located immediately adjacent to the articular cartilage. The 3 lateral sites were also in the anterior, central, and posterior greater tuberosity but located in a lateral position consistent with doublerow cuff fixations using either conventional techniques or suture-bridging techniques. After specimen preparation and anchor insertion, the specimens were placed in an MTS 810 servohydraulic testing machine (MTS Systems, Eden Prairie,
FIGURE 1. Three medial-row and three lateral-row sites were identified at the greater tuberosity of the proximal humerus for anchor insertion. © 2010 F. Alan Barber, used by permission.
FIGURE 2. Each humerus was fixed at the shaft with the long axis of the humerus placed at approximately 135° to the load actuator. © 2010 F. Alan Barber, used by permission.
MN) with a 1-kN load cell. Each humerus was fixed at the shaft with the long axis of the humerus placed at approximately 135° to the load actuator (Fig 2). All anchors were threaded with 2 sutures containing ultrahigh–molecular weight polyethylene. The 2 sutures from each anchor were grasped with equal tension by a pneumatic suture grip with the same suture gauge length (45 mm) from the humeral head surface. The adjustable angle fixture was positioned at a 45° angle to more closely simulate the anatomic direction of the load applied by the rotator cuff without abrasion from the humeral head.6 The constructs were preloaded with 10 N at 1 N/s. The preload was held for 5 seconds, and then the constructs were cycled from 10 to 60 N at 1 Hz for 500 cycles. For those constructs surviving the 500 cycles, this was followed by a single pull to failure at 33 mm/s, consistent with prior reports.4,16 The data were plotted and cyclic displacement, ultimate load, and mode of failure recorded. Gap formation in a rotator cuff repair is a significant concern. Two displacement gaps were evaluated for these cyclic repair constructs: 5 mm and 10 mm. Because there was no tendon in this construct and the MTS machine grip was attached to the sutures with a standard gauge length (40 mm), the gap measured was the elongation documented in the system after the preload and at various intervals in the cyclic loading.
Abbreviations: Bio-C, Bio-Corkscrew FT; TwinFix, TwinFix FT PEEK; Healix, Healix PEEK; PushLock, PEEK PushLock; P-Lok, PopLok PEEK; Fprnt, Footprint PK; V-Lok, Versalok.
PushLock P-Lok Fprnt V-Lok V-Lok PushLock P-Lok Fprnt Fprnt V-Lok PushLock P-Lok PushLock P-Lok Fprnt V-Lok V-Lok PushLock P-Lok Fprnt Fprnt V-Lok PushLock P-Lok PushLock P-Lok Fprnt V-Lok V-Lok PushLock P-Lok Fprnt Fprnt V-Lok PushLock P-Lok P-Lok Fprnt V-Lok PushLock P-Lok Fprnt V-Lok PushLock P-Lok
Healix CrossFT TwinFix CrossFT TwinFix Bio-C TwinFix Bio-C Healix Healix Bio-C CrossFT Healix TwinFix CrossFT CrossFT TwinFix Bio-C TwinFix Bio-C Healix Healix Bio-C CrossFT Healix TwinFix CrossFT CrossFT TwinFix Bio-C TwinFix Bio-C Healix Healix Bio-C CrossFT Healix TwinFix CrossFT CrossFT TwinFix Bio-C TwinFix Bio-C Healix
Medial-row anchors Anterior Bio-C Middle Healix Posterior CrossFT Lateral-row anchors Anterior PushLock Middle V-Lok Posterior Fprnt
15 14 13 12 11 10 9 Location
1
2
3
4
5
6
7
8
Bone
Outline of 48 Anchor Specimen Assignments to Each Bone, Row, and Location
As indicated in the mode-of-failure data, this principally reflected movement at the anchor-bone interface. The data generated were evaluated to determine the mean cyclic displacement at the start of the test, at 100 cycles, and at 500 cycles; the mean yield load (start point at failure); and the ultimate load-to-failure strength for each anchor type. The anchor tests reflecting a 5- and 10-mm gap formation were also noted. The mode of failure was also recorded (anchor pullout, eyelet failure, suture breakage). We rotated 4 different medial anchors and 4 different lateral anchors among these sites in the 8 matched cadaveric pairs, with each site receiving a single anchor. The order of anchor site insertion was varied to reduce any variation between pairs of humeri or within pairs (Table 1). To minimize bias caused by either the insertion site or the differences in bone density of the different bones, the anchors were rotated such that each anchor was placed in each of the 3 bone insertion sites the same number of times, and every anchor was inserted into 1 bone of each pair at least once. Because there were 3 test sites in each of 16 shoulders (48 sites) and 4 anchors rotated among them, 12 tests of each anchor were obtained. The rotation among the matched pairs and total of 16 different humeri also reduced the effect of bone density differences that could affect the results. All medial anchors were placed in the humeral head at an angle no greater than 45° to angle toward the subchondral bone.4,16 All lateral anchors were inserted “over the top” of the greater tuberosity, nearly planar to the superior humeral surface, so that the lateral anchors also approached the subchondral bone for optimum fixation. The 4 medial anchors tested were the Bio-Corkscrew FT (Arthrex, Naples FL), the CrossFT PEEK (polyetheretherketone) (ConMed Linvatec, Largo, FL), the TwinFix PK FT (Smith & Nephew Endoscopy, Andover, MA), and the Healix PEEK (DePuy Mitek, Raynham, MA) (Fig 3). The Bio-Corkscrew FT is a poly-L-lactic acid screw anchor with a major diameter of 5.5 mm, a minor diameter of 3.65 mm, and a length of 14.9 mm (Fig 3). It has an eyelet made from suture lodged in the anchor’s central core. This suture eyelet is made from either No. 4 braided polyester or No. 2 FiberWire (Arthrex). The anchors tested held 2 No. 2 high-strength sutures containing ultrahigh–molecular weight polyethylene (FiberWire). The CrossFT PEEK is a screw anchor with a distal crossbar eyelet, and it accommodates 3 No. 2 highstrength sutures (Hi-Fi; ConMed Linvatec) (Fig 3).
16
F. A. BARBER ET AL.
TABLE 1.
S136
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS
S137
FIGURE 3. Four medial-row anchors were tested (left to right): Bio-Corkscrew FT, CrossFT PEEK, TwinFix PK FT, and Healix. © 2010 F. Alan Barber, used by permission.
There is a single screw thread that extends from the proximal to distal end and a second interleaved thread at the proximal end to maximize cortical compression. The major diameter is 5.5 mm, the minor diameter is 3.8 mm, and the length is 17 mm. The TwinFix PK FT 5.5 anchor (Smith & Nephew Endoscopy) is a threaded screw anchor made from PEEK with proximal threads to engage the cortical bone and is double or triple loaded with No. 2 highstrength sutures (UltraBraid; Smith & Nephew Endoscopy) that pass through the center of the anchor to loop through a distal “eyelet” (Fig 3). The tested version had 2 sutures. It has a major diameter of 5.5 mm, a minor diameter of 3.5 mm, and a length of 15 mm.
FIGURE 4. Including the Versalok anchor (Fig 5), 4 lateral-row anchors tested (left to right): Footprint PEEK anchor, PopLok PEEK, and PEEK PushLock (Arthrex). © 2010 F. Alan Barber, used by permission.
FIGURE 5. The Versalok anchor is an expanding bolt with 2 separate components, a titanium pin that is forced inside a PEEK outer sleeve. © 2010 F. Alan Barber, used by permission.
The Healix PEEK anchor tested is made from PEEK and has an outer diameter of 5.5 mm and an inner diameter of 3.9 mm (Fig 3). The anchor is 18 mm long with 10 threads (5 distal cancellous threads and 5 proximal cortical threads). These cortical threads are more tightly spaced than the cancellous threads. It was loaded with 2 No. 2 high-strength sutures (Orthocord; DePuy Mitek) that passed down the central anchor core to loop over a distal crossbar “eyelet.” The 4 lateral anchors tested were the PopLok PEEK (ConMed Linvatec), the PEEK PushLock (Arthrex), the Footprint PEEK anchor (Smith & Nephew Endoscopy), and the Versalok (DePuy Mitek) (Figs 4 and 5). The PopLok PEEK 4.5 is an expanding bolt–type anchor that is 4.5 mm in diameter and 15.5 mm long (Fig 4). It can accommodate up to 4 strands of any No. 2 high-strength suture. It has 2 wings that deploy to hold the anchor in position, and once deployed, the diameter across the wing tips is 9 mm and the length shortens to 11 mm. The PEEK PushLock (Arthrex) is 5.5 mm in diameter and has 2 separate PEEK components: a distal closed “eyelet” and a body with a flight of 9 screw threads around a hollow core through which the driver shaft and a holding suture pass (Fig 4). The closed distal eyelet is 2 mm wide and can hold several sutures or a single 2-mm tape. This eyelet in the anchors tested was loaded with 2 No. 2 high-strength sutures (Hi-Fi). The length of the body is 18.5 mm with an eyelet length of 5.6 mm. The Footprint PK 5.5 is made of PEEK material
S138
F. A. BARBER ET AL.
with an outer diameter of 5.5 mm and a length of 14.9 mm (Fig 4). The inner diameter tapers from 4.5 mm at the proximal end to 3.7 mm at the distal end. Four No. 2 sutures from other anchors may be loaded into the distal eyelet by use of a passing suture provided with the anchor. These sutures are then secured by an inner plug that is tightened down to securely lock the sutures inside the anchor once appropriate tension is applied to the sutures. The Versalok anchor is an expanding bolt with 2 separate components (Fig 5). The pre-deployment outer diameter is 4.9 mm and the length is 27 mm, but the insertion depth is 31 mm. During deployment, the titanium pin is forced inside the PEEK outer sleeve. This expands the outer diameter of the implant to 6.3 mm, shortens the length of the device to 17 mm, and traps any sutures passed through the titanium section. We obtained 12 samples of each of the anchors. We performed anchor insertion using the manufacturers’ instruments and instructions, and anchors were advanced to a depth designated by a laser line on the appropriate inserter. All anchors were placed at least 1 cm apart and in the correct third of the tuberosity to avoid crack propagation between insertion sites. Anchors were inserted so that their eyelets were oriented toward the bone and consequently in the direction of the subsequent pull. Statistical Analysis A 1-way analysis of variance was conducted to compare each of the 4 endpoints. The analysis of variance included the Tukey test for pair-wise comparisons. An a priori power analysis was performed that found 12 tests for each anchor would be necessary to show a 50-N difference with 80% power.4,16-18 A mixed-effects re-
TABLE 2.
gression model was conducted to compare the groups. P ⬍ .05 was considered significant.
RESULTS The 16 cadaveric humeri were obtained from 8 individuals. One pair of shoulders was more osteoporotic than the others, and the anchors pulled out during the first cycle. The humeral heads had large cysts, and the cortical bone failed, which resulted in the anchor pullout observed. This individual was a 71-year-old man who died of respiratory failure. We inserted all anchors using the manufacturers’ specifications, insertion techniques, and insertion equipment. After the preload was applied, cyclic loading data were obtained. The initial displacement (occurring during the preload), displacement after 100 cycles, and displacement after 500 cycles were recorded for every test (Table 2). As has been reported in other studies, most of the total displacement observed occurred in the first 100 cycles, except for the Footprint.4,17 The TwinFix PK FT showed almost as much displacement in the first 100 cycles as in the final 200 cycles. Ultimate failure loads were recorded (Table 2) and ranged from a mean of 163 N to a mean of 308 N. Table 2 includes the load recorded at the time of failure, including those anchors that failed during the cyclic phase. The Footprint was weaker than the Versalok (P ⬍ .05). Two displacement gap thresholds were evaluated after cyclic loading of these anchors: 5 mm and 10 mm. None of the anchors surviving the 500 cycles showed 10 mm of displacement. However, some tests did reach the 5-mm threshold (Table 3).
Initial Displacement, Displacement After 100 cycles, and Displacement After 500 Cycles
Anchor
Initial Displacement (mm)
Displacement After 100 Cycles (mm)
Displacement After 500 Cycles (mm)
Displacement Within First 100 Cycles (mm)
Total Displacement (mm)
Ultimate Load (N)
SEM
Range
Bio-Corkscrew FT CrossFT PEEK TwinFix PK FT Healix PEEK PopLok PEEK PushLock Footprint Versalok
5.6 6.5 6.2 6.6 11.4 9.5 10.3 9.8
6.8 7.9 7.3 7.5 13.3 13.8 13.2 11
7.3 9 8.3 7.9 14.8 15.1 16.9 11.6
1.2 1.4 1.1 0.9 1.9 4.3 2.9 1.2
1.7 2.5 2.1 1.3 3.4 5.6 6.6 1.8
209.3 269.3 263.7 295.9 209.8 201.7 163.2* 308
27.9 35.0 30.9 38.1 20.2 42.7 31.4 18.5
66-343 64-425 47-427 80-459 104-304 44-413 28-296 212-426
*The Footprint was weaker than the Versalok (P ⬍ .05).
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS The 5-mm displacement gap threshold was the only “failure level” observed in this testing (Table 3). The Bio-Corkscrew FT, TwinFix PK FT, and Versalok did not have any tests exceeding 5 mm of displacement. The other anchors had 1 test exceeding the 5-mm displacement threshold and the PushLock had 2. What may be more telling is that only 6 of 12 PushLock anchors and 5 of 12 Footprint anchors survived 500 cycles, of which 2 PushLock anchors and 1 Footprint anchor exceeded 5 mm of displacement (Table 3). The Versalok anchor was the only lateral-row anchor for which all anchors tested survived 500 cycles. The mode of failure was also recorded (anchor pullout, suture breakage, and eyelet failure) (Table 4). In addition, the 1 specimen with very osteoporotic bone resulted in such weak anchors that most failed when being connected to the MTS testing device. This event was recorded as bone failure. The principal mode of failure was by anchor pullout. DISCUSSION The purpose of this study is to test medial-row and lateral-row suture anchors under cyclic loading conditions. Advocates of double-row rotator cuff repair highlight the added strength provided by the additional fixation points and the larger repair footprint created with more tissue coverage of the greater tuberosity.8-11 The presumption is that these factors lead to better tendonto-bone healing. One common double-row rotator technique connects the sutures from the medial anchor row in a crossing fashion to the anchors of the lateral row.6,19-21 This is referred to as the cross-stitch or suture-bridge technique. To accomplish this, the lateral-row anchors must be able to accept sutures from the previously placed anchors in their eyelet and then lock these sutures in place. These lateral-row anchors must have a cinching or locking eyelet mechanism.
TABLE 3. Number of Anchors Completing Cyclic Loading Phase and Those Showing 5 mm of Displacement Anchor
Tests Reaching 500 Cycles
Tests Exceeding 5 mm
Bio-Corkscrew FT CrossFT PEEK TwinFix PK FT Healix PEEK PopLok PEEK PushLock Footprint Versalok
8 8 10 10 9 6 5 12
0 1 0 1 1 2 1 0
TABLE 4.
S139
Mode of Failure
Anchor
Anchor Pullout
Suture Broke
Eyelet Broke
Bone Failure
Bio-Corkscrew FT CrossFT PEEK TwinFix PK FT Healix PEEK PopLok PEEK PushLock Footprint Versalok
10 12 9 10 11 7 8 10
0 0 1 0 0 0 0 2
1 0 2 1 0 3 3 0
1 0 1 1 1 2 1 0
Trabecular, cortical, and total bone density is greater near the articular cartilage but declines at more lateral positions on the greater tuberosity (P ⬍ .01), making the lateral anchor locations inherently weaker insertion sites and more likely to be associated with tendon-bone gap formation.22 In recognition of this, significant design differences exist between the new anchors used for a lateral-row application and those used for medial rows. These design differences take into account the variable nature of the bone into which the anchors are inserted, the mechanism for holding the suture, the number of sutures accommodated, and possibly, the anchor size. Previous cyclic loading studies have reported that more displacement (repair separation) occurs during the initial 100 test cycles than occurs between 100 and 500 cycles.4,17 In our test there was a clear difference in the behavior between the medial-row and lateralrow anchors. Whereas more displacement occurred in the first 100 cycles for all anchors except the Footprint (a lateral-row anchor), the amount of that displacement was much greater for the lateral-row anchors than the medial-row anchors (mean displacement of 1.2 mm at 100 cycles for the medial-row anchors compared with 2.6 mm for the lateral-row anchors). The displacement shown in the 100 cycle–to–500 cycle portions of the test was also much greater for the lateral-row anchors (mean, 1.8 mm) than the medialrow anchors (mean, 0.75 mm). The total displacement observed for all anchors was also greater for the lateral-row anchors (mean, 4.4 mm) compared with the medial-row anchors (mean, 1.9 mm), and the mean 4.4 mm of displacement approached the 5.0 mm of displacement defined as a failure. This means that both medial- and lateral-row anchor performance was consistent with prior studies. What is significant is that the lateral-row anchors, perhaps as a result of their design, were found to slip much more than the medialrow anchors both in the first 100 cycles, when most of
S140
F. A. BARBER ET AL.
the slippage occurs, and continuing into the interval between 100 and 500 cycles. No anchor that survived 500 cycles showed 10 mm of displacement. As noted in these data, more medial-row anchors completed the 500 cycles than lateral-row anchors. Early lateral-row anchor failure may be because of osteoporotic bone at the insertion site or anchor design. This also suggests that these lateral anchor designs work better in conjunction with medial-row anchors, which take up most of the tension. A standalone application for the lateral-row anchors should be considered with caution. The mean ultimate failure loads (Table 2) ranged from 163 to 308 N. These figures include those anchors that failed during cyclic loading. Statistical analysis of these load values showed only 1 significant result: the Versalok anchor was statistically stronger than the Footprint anchor (P ⬍ .05). The principal mode of failure for all anchors was by anchor pullout (Table 4). The next most common failure mode was eyelet breakage. The failure load at which eyelet breakage was recorded was distinctly different for medial-row anchors and lateral-row anchors. The lateral-row anchors PushLock and Footprint both had 3 tests fail by eyelet breakage. Eyelet failure in all but 1 of these 6 lateral-row anchors occurred at less than 500 cycles and consequently was at a lower failure load than the mean for the particular anchor (PushLock eyelet failures at 83 N v PushLock mean of 253 N and Footprint eyelet failures at 105 N v Footprint mean of 253 N). The medial-row anchors’ behavior was in marked contrast. All eyelet failures in medial-row anchors occurred after successfully completing the 500 cycles and, except for 1 anchor, at loads higher than the mean failure load for that anchor. The lateral-row anchors tested differ from the medialrow anchors tested in both fundamental design and performance. Lateral-row anchors accommodate more suture strands, as well as sutures from other anchors, and fix these sutures with cinching or locking mechanisms. Medial-row anchors use conventional eyelet posts and rely on tied knots for suture security. These design differences allow the sutures held in lateralrow anchors to slip if the loads are high enough. In contrast, the sutures in medial-row anchors are dependent on the knot security to avoid slippage. The isolated use of an anchor designed for lateral-row applications is likely to provide fixation inferior to that of a medial-row anchor with a greater likelihood of suture slippage in a cyclic loading model. However, the function of the medial- and lateral-row anchors tested
in this study was not tested specifically in terms of rotator cuff loading in this study setup. Strengths of this study include the use of paired specimens of human bone, the distribution of the anchors among 3 different medial and 3 different lateral insertion sites, and the cyclic loading protocol that applied stress in line with the direction of a physiologic tendon pull. Limitations of the study include the fact that testing was performed at room temperature and not in a water bath to more accurately simulate the clinical setting. A tendon repair was not performed, eliminating what is probably the most common cuff repair failure site. Space did not allow placement of 1 sample of each of the 4 different anchors tested in each location of every pair, but the paired specimens allowed for consistency of the bone insertion environments. The advanced age of the specimens could have affected the bone quality, although rotator cuff repair is common in older populations. One pair of humeri was very osteoporotic, and early anchor failure was observed. No radiographs were obtained to exclude patients with large cysts. This in vivo test cannot be applied directly to the clinical setting. This construct did not test single- or doublerow repairs but rather the isolated performance of anchors typically assigned a medial- or lateral-row function in the humeral site consistent with these roles. These data represent a time 0 response and do not take into account healing and the postoperative rehabilitation program. CONCLUSIONS Anchors designed to function as lateral-row fixation provide fixation strength inferior to that of medial-row anchors and are more likely to be subject to suture slippage. When lateral-row anchors fail at the eyelet, it is at lower failure loads, whereas if medial-row anchors fail at the eyelet, it is at higher loads. REFERENCES 1. Adams CR, Schoolfield JD, Burkhart SS. The results of arthroscopic subscapularis tendon repairs. Arthroscopy 2008;24: 1381-1389. 2. Castagna A, Garofalo R, Conti M, Borroni M, Snyder SJ. Arthroscopic rotator cuff repair using a triple-loaded suture anchor and a modified Mason-Allen technique (Alex stitch). Arthroscopy 2007;23:440.e1-440.e4. Available online at www.arthroscopyjournal.org. 3. Krishnan SG, Harkins DC, Schiffern SC, Pennington SD, Burkhead WZ. Arthroscopic repair of full-thickness tears of the rotator cuff in patients younger than 40 years. Arthroscopy 2008;24:324-328.
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS 4. Barber FA, Coons DA, Ruiz-Suarez M. Cyclic load testing of biodegradable suture anchors containing 2 high-strength sutures. Arthroscopy 2007;23:355-360. 5. Bynum CK, Lee S, Mahar A, Tasto J, Pedowitz R. Failure mode of suture anchors as a function of insertion depth. Am J Sports Med 2005;33:1030-1034. 6. Burkhart SS, Adams CR, Schoolfield JD. A biomechanical comparison of 2 techniques of footprint reconstruction for rotator cuff repair: The SwiveLock-FiberChain construct versus standard double-row repair. Arthroscopy 2009;25:274-281. 7. Spang JT, Buchmann S, Brucker PU, et al. A biomechanical comparison of 2 transosseous-equivalent double-row rotator cuff repair techniques using bioabsorbable anchors: Cyclic loading and failure behavior. Arthroscopy 2009;25:872-879. 8. Trantalis JN, Boorman RS, Pletsch K, Lo IK. Medial rotator cuff failure after arthroscopic double-row rotator cuff repair. Arthroscopy 2008;24:727-731. 9. Reardon DJ, Maffulli N. Clinical evidence shows no difference between single- and double-row repair for rotator cuff tears. Arthroscopy 2007;23:670-673. 10. Nelson CO, Sileo MJ, Grossman MG, Serra-Hsu F. Single-row modified Mason-Allen versus double-row arthroscopic rotator cuff repair: A biomechanical and surface area comparison. Arthroscopy 2008;24:941-948. 11. Cole BJ, ElAttrache NS, Anbari A. Arthroscopic rotator cuff repairs: An anatomic and biomechanical rationale for different suture-anchor repair configurations. Arthroscopy 2007;23:662669. 12. Mahar A, Tamborlane J, Oka R, Esch J, Pedowitz RA. Singlerow suture anchor repair of the rotator cuff is biomechanically equivalent to double-row repair in a bovine model. Arthroscopy 2007;23:1265-1270. 13. Ji JH, Kim WY, Ra KH. Arthroscopic double-row suture anchor fixation of minimally displaced greater tuberosity frac-
14.
15.
16.
17. 18. 19. 20.
21. 22.
S141
tures. Arthroscopy 2007;23:1133.e1-1133.e4. Available online at www.arthroscopyjournal.org. Grasso A, Milano G, Salvatore M, Falcone G, Deriu L, Fabbriciani C. Single-row versus double-row arthroscopic rotator cuff repair: A prospective randomized clinical study. Arthroscopy 2009;25:4-12. Toussaint B, Schnaser E, Lafosse L, Bahurel J, Gobezie R. A new approach to improving the tissue grip of the medial-row repair in the suture-bridge technique: The “modified lasso-loop stitch.” Arthroscopy 2009;25:691-695. Coons DA, Barber FA, Herbert MA. Triple-loaded singleanchor stitch configurations: An analysis of cyclically loaded suture-tendon interface security. Arthroscopy 2006;22:11541158. Barber FA, Coons DA, Ruiz-Suarez M. Cyclic load testing and ultimate failure strength of biodegradable glenoid anchors. Arthroscopy 2008;24:224-228. Ruiz-Suarez M, Aziz-Jacobo J, Barber FA. Cyclic load testing and ultimate failure strength of hip arthroscopy suture anchors. Arthroscopy 2010;26:762–768. Kim KC, Rhee KJ, Shin HD. Deformities associated with the suture-bridge technique for full-thickness rotator cuff tears. Arthroscopy 2008;24:1251-1257. Lorbach O, Bachelier F, Vees J, Kohn D, Pape D. Cyclic loading of rotator cuff reconstructions: Single-row repair with modified suture configurations versus double-row repair. Am J Sports Med 2008;36:1504-1510. Song HS, Williams GR Jr. Arthroscopic reduction and fixation with suture-bridge technique for displaced or comminuted greater tuberosity fractures. Arthroscopy 2008;24:956-960. Tingart MJ, Apreleva M, Zurakowski D, Warner JJ. Pullout strength of suture anchors used in rotator cuff repair. J Bone Joint Surg Am 2003;85:2190-2198.
S
uture anchors loaded with multiple high-strength sutures are now commonly used for rotator cuff repair.1-3 Comparative cyclic load testing in line with the
From the Plano Orthopedic Sports Medicine and Spine Center (F.A.B.), Plano, Texas; A. Kagan Orthopedics and Sports Medicine (J.A.B.), Fort Myers, Florida, U.S.A.; and Department of Orthopaedics and Traumatology, Mustafa Kemal University (O.H.), Antakya, Turkey. Supported by ConMed Linvatec, Largo, Florida, and DePuy Mitek, Raynham, Massachusetts. The authors have received support from ConMed Linvatec, DePuy Mitek, and Smith & Nephew Endoscopy, Andover, Massachusetts, exceeding US $500 related to this research. Research was performed at ConMed Linvatec. Received September 20, 2009; accepted March 4, 2010. Address correspondence and reprint requests to F. Alan Barber, M.D., Plano Orthopedic Sports Medicine and Spine Center 5228 W Plano Pkwy, Plano, TX 75093, U.S.A. © 2010 by the Arthroscopy Association of North America 0749-8063/9544/$36.00 doi:10.1016/j.arthro.2010.03.007
axis of the tendon is a testing mode that is consistent with the clinical situation.4-6 Single-pull ultimate failure loads do not reflect the clinical condition. Cyclic loading such as occurs with arm motion during rehabilitation or during activities of daily living is more reflective of the stresses on the rotator cuff repair occurring during the postoperative rehabilitation period. Repair displacement of 5 to 10 mm with cyclic loads would in reality be a clinical failure. Double-row constructs have gained considerable attention related to the repair of rotator cuff tendons.7-15 Tests that reflect how suture anchor constructs perform under cyclic loads provide more practical information about how successful they might be when used clinically. However, little evidence exists that compares the performance of the typical medial-row anchor with the more recently developed lateral anchors. In addition, the potential exists for the use of lateral-row anchors as independent fixation devices.
S134 Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 26, No 9 (September, Suppl 1), 2010: pp S134-S141
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS
S135
The purpose of this study was to compare isolated medial-row anchor with isolated lateral-row anchor performance by use of cyclic loading followed by destructive testing in an in vitro cadaveric model. Our hypothesis was that suture anchors designed for medial- or lateral-row rotator cuff applications would show different biomechanical behavior when subjected to cyclic loading. METHODS We prepared 16 humeri from 8 human donors by stripping all the soft tissue from the bones. The ages of these humeri were recorded. No donors with bone metastases, generalized bone disease, lymphoma, myeloma, or leukemia were included. The mean age was 80 years (range, 70 to 96 years), and there was only 1 female donor. Three medial and three lateral sites were identified at the greater tuberosity of the proximal humerus for anchor insertion (Fig 1). The 3 medial sites were in the anterior, central, and posterior greater tuberosity located immediately adjacent to the articular cartilage. The 3 lateral sites were also in the anterior, central, and posterior greater tuberosity but located in a lateral position consistent with doublerow cuff fixations using either conventional techniques or suture-bridging techniques. After specimen preparation and anchor insertion, the specimens were placed in an MTS 810 servohydraulic testing machine (MTS Systems, Eden Prairie,
FIGURE 1. Three medial-row and three lateral-row sites were identified at the greater tuberosity of the proximal humerus for anchor insertion. © 2010 F. Alan Barber, used by permission.
FIGURE 2. Each humerus was fixed at the shaft with the long axis of the humerus placed at approximately 135° to the load actuator. © 2010 F. Alan Barber, used by permission.
MN) with a 1-kN load cell. Each humerus was fixed at the shaft with the long axis of the humerus placed at approximately 135° to the load actuator (Fig 2). All anchors were threaded with 2 sutures containing ultrahigh–molecular weight polyethylene. The 2 sutures from each anchor were grasped with equal tension by a pneumatic suture grip with the same suture gauge length (45 mm) from the humeral head surface. The adjustable angle fixture was positioned at a 45° angle to more closely simulate the anatomic direction of the load applied by the rotator cuff without abrasion from the humeral head.6 The constructs were preloaded with 10 N at 1 N/s. The preload was held for 5 seconds, and then the constructs were cycled from 10 to 60 N at 1 Hz for 500 cycles. For those constructs surviving the 500 cycles, this was followed by a single pull to failure at 33 mm/s, consistent with prior reports.4,16 The data were plotted and cyclic displacement, ultimate load, and mode of failure recorded. Gap formation in a rotator cuff repair is a significant concern. Two displacement gaps were evaluated for these cyclic repair constructs: 5 mm and 10 mm. Because there was no tendon in this construct and the MTS machine grip was attached to the sutures with a standard gauge length (40 mm), the gap measured was the elongation documented in the system after the preload and at various intervals in the cyclic loading.
Abbreviations: Bio-C, Bio-Corkscrew FT; TwinFix, TwinFix FT PEEK; Healix, Healix PEEK; PushLock, PEEK PushLock; P-Lok, PopLok PEEK; Fprnt, Footprint PK; V-Lok, Versalok.
PushLock P-Lok Fprnt V-Lok V-Lok PushLock P-Lok Fprnt Fprnt V-Lok PushLock P-Lok PushLock P-Lok Fprnt V-Lok V-Lok PushLock P-Lok Fprnt Fprnt V-Lok PushLock P-Lok PushLock P-Lok Fprnt V-Lok V-Lok PushLock P-Lok Fprnt Fprnt V-Lok PushLock P-Lok P-Lok Fprnt V-Lok PushLock P-Lok Fprnt V-Lok PushLock P-Lok
Healix CrossFT TwinFix CrossFT TwinFix Bio-C TwinFix Bio-C Healix Healix Bio-C CrossFT Healix TwinFix CrossFT CrossFT TwinFix Bio-C TwinFix Bio-C Healix Healix Bio-C CrossFT Healix TwinFix CrossFT CrossFT TwinFix Bio-C TwinFix Bio-C Healix Healix Bio-C CrossFT Healix TwinFix CrossFT CrossFT TwinFix Bio-C TwinFix Bio-C Healix
Medial-row anchors Anterior Bio-C Middle Healix Posterior CrossFT Lateral-row anchors Anterior PushLock Middle V-Lok Posterior Fprnt
15 14 13 12 11 10 9 Location
1
2
3
4
5
6
7
8
Bone
Outline of 48 Anchor Specimen Assignments to Each Bone, Row, and Location
As indicated in the mode-of-failure data, this principally reflected movement at the anchor-bone interface. The data generated were evaluated to determine the mean cyclic displacement at the start of the test, at 100 cycles, and at 500 cycles; the mean yield load (start point at failure); and the ultimate load-to-failure strength for each anchor type. The anchor tests reflecting a 5- and 10-mm gap formation were also noted. The mode of failure was also recorded (anchor pullout, eyelet failure, suture breakage). We rotated 4 different medial anchors and 4 different lateral anchors among these sites in the 8 matched cadaveric pairs, with each site receiving a single anchor. The order of anchor site insertion was varied to reduce any variation between pairs of humeri or within pairs (Table 1). To minimize bias caused by either the insertion site or the differences in bone density of the different bones, the anchors were rotated such that each anchor was placed in each of the 3 bone insertion sites the same number of times, and every anchor was inserted into 1 bone of each pair at least once. Because there were 3 test sites in each of 16 shoulders (48 sites) and 4 anchors rotated among them, 12 tests of each anchor were obtained. The rotation among the matched pairs and total of 16 different humeri also reduced the effect of bone density differences that could affect the results. All medial anchors were placed in the humeral head at an angle no greater than 45° to angle toward the subchondral bone.4,16 All lateral anchors were inserted “over the top” of the greater tuberosity, nearly planar to the superior humeral surface, so that the lateral anchors also approached the subchondral bone for optimum fixation. The 4 medial anchors tested were the Bio-Corkscrew FT (Arthrex, Naples FL), the CrossFT PEEK (polyetheretherketone) (ConMed Linvatec, Largo, FL), the TwinFix PK FT (Smith & Nephew Endoscopy, Andover, MA), and the Healix PEEK (DePuy Mitek, Raynham, MA) (Fig 3). The Bio-Corkscrew FT is a poly-L-lactic acid screw anchor with a major diameter of 5.5 mm, a minor diameter of 3.65 mm, and a length of 14.9 mm (Fig 3). It has an eyelet made from suture lodged in the anchor’s central core. This suture eyelet is made from either No. 4 braided polyester or No. 2 FiberWire (Arthrex). The anchors tested held 2 No. 2 high-strength sutures containing ultrahigh–molecular weight polyethylene (FiberWire). The CrossFT PEEK is a screw anchor with a distal crossbar eyelet, and it accommodates 3 No. 2 highstrength sutures (Hi-Fi; ConMed Linvatec) (Fig 3).
16
F. A. BARBER ET AL.
TABLE 1.
S136
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS
S137
FIGURE 3. Four medial-row anchors were tested (left to right): Bio-Corkscrew FT, CrossFT PEEK, TwinFix PK FT, and Healix. © 2010 F. Alan Barber, used by permission.
There is a single screw thread that extends from the proximal to distal end and a second interleaved thread at the proximal end to maximize cortical compression. The major diameter is 5.5 mm, the minor diameter is 3.8 mm, and the length is 17 mm. The TwinFix PK FT 5.5 anchor (Smith & Nephew Endoscopy) is a threaded screw anchor made from PEEK with proximal threads to engage the cortical bone and is double or triple loaded with No. 2 highstrength sutures (UltraBraid; Smith & Nephew Endoscopy) that pass through the center of the anchor to loop through a distal “eyelet” (Fig 3). The tested version had 2 sutures. It has a major diameter of 5.5 mm, a minor diameter of 3.5 mm, and a length of 15 mm.
FIGURE 4. Including the Versalok anchor (Fig 5), 4 lateral-row anchors tested (left to right): Footprint PEEK anchor, PopLok PEEK, and PEEK PushLock (Arthrex). © 2010 F. Alan Barber, used by permission.
FIGURE 5. The Versalok anchor is an expanding bolt with 2 separate components, a titanium pin that is forced inside a PEEK outer sleeve. © 2010 F. Alan Barber, used by permission.
The Healix PEEK anchor tested is made from PEEK and has an outer diameter of 5.5 mm and an inner diameter of 3.9 mm (Fig 3). The anchor is 18 mm long with 10 threads (5 distal cancellous threads and 5 proximal cortical threads). These cortical threads are more tightly spaced than the cancellous threads. It was loaded with 2 No. 2 high-strength sutures (Orthocord; DePuy Mitek) that passed down the central anchor core to loop over a distal crossbar “eyelet.” The 4 lateral anchors tested were the PopLok PEEK (ConMed Linvatec), the PEEK PushLock (Arthrex), the Footprint PEEK anchor (Smith & Nephew Endoscopy), and the Versalok (DePuy Mitek) (Figs 4 and 5). The PopLok PEEK 4.5 is an expanding bolt–type anchor that is 4.5 mm in diameter and 15.5 mm long (Fig 4). It can accommodate up to 4 strands of any No. 2 high-strength suture. It has 2 wings that deploy to hold the anchor in position, and once deployed, the diameter across the wing tips is 9 mm and the length shortens to 11 mm. The PEEK PushLock (Arthrex) is 5.5 mm in diameter and has 2 separate PEEK components: a distal closed “eyelet” and a body with a flight of 9 screw threads around a hollow core through which the driver shaft and a holding suture pass (Fig 4). The closed distal eyelet is 2 mm wide and can hold several sutures or a single 2-mm tape. This eyelet in the anchors tested was loaded with 2 No. 2 high-strength sutures (Hi-Fi). The length of the body is 18.5 mm with an eyelet length of 5.6 mm. The Footprint PK 5.5 is made of PEEK material
S138
F. A. BARBER ET AL.
with an outer diameter of 5.5 mm and a length of 14.9 mm (Fig 4). The inner diameter tapers from 4.5 mm at the proximal end to 3.7 mm at the distal end. Four No. 2 sutures from other anchors may be loaded into the distal eyelet by use of a passing suture provided with the anchor. These sutures are then secured by an inner plug that is tightened down to securely lock the sutures inside the anchor once appropriate tension is applied to the sutures. The Versalok anchor is an expanding bolt with 2 separate components (Fig 5). The pre-deployment outer diameter is 4.9 mm and the length is 27 mm, but the insertion depth is 31 mm. During deployment, the titanium pin is forced inside the PEEK outer sleeve. This expands the outer diameter of the implant to 6.3 mm, shortens the length of the device to 17 mm, and traps any sutures passed through the titanium section. We obtained 12 samples of each of the anchors. We performed anchor insertion using the manufacturers’ instruments and instructions, and anchors were advanced to a depth designated by a laser line on the appropriate inserter. All anchors were placed at least 1 cm apart and in the correct third of the tuberosity to avoid crack propagation between insertion sites. Anchors were inserted so that their eyelets were oriented toward the bone and consequently in the direction of the subsequent pull. Statistical Analysis A 1-way analysis of variance was conducted to compare each of the 4 endpoints. The analysis of variance included the Tukey test for pair-wise comparisons. An a priori power analysis was performed that found 12 tests for each anchor would be necessary to show a 50-N difference with 80% power.4,16-18 A mixed-effects re-
TABLE 2.
gression model was conducted to compare the groups. P ⬍ .05 was considered significant.
RESULTS The 16 cadaveric humeri were obtained from 8 individuals. One pair of shoulders was more osteoporotic than the others, and the anchors pulled out during the first cycle. The humeral heads had large cysts, and the cortical bone failed, which resulted in the anchor pullout observed. This individual was a 71-year-old man who died of respiratory failure. We inserted all anchors using the manufacturers’ specifications, insertion techniques, and insertion equipment. After the preload was applied, cyclic loading data were obtained. The initial displacement (occurring during the preload), displacement after 100 cycles, and displacement after 500 cycles were recorded for every test (Table 2). As has been reported in other studies, most of the total displacement observed occurred in the first 100 cycles, except for the Footprint.4,17 The TwinFix PK FT showed almost as much displacement in the first 100 cycles as in the final 200 cycles. Ultimate failure loads were recorded (Table 2) and ranged from a mean of 163 N to a mean of 308 N. Table 2 includes the load recorded at the time of failure, including those anchors that failed during the cyclic phase. The Footprint was weaker than the Versalok (P ⬍ .05). Two displacement gap thresholds were evaluated after cyclic loading of these anchors: 5 mm and 10 mm. None of the anchors surviving the 500 cycles showed 10 mm of displacement. However, some tests did reach the 5-mm threshold (Table 3).
Initial Displacement, Displacement After 100 cycles, and Displacement After 500 Cycles
Anchor
Initial Displacement (mm)
Displacement After 100 Cycles (mm)
Displacement After 500 Cycles (mm)
Displacement Within First 100 Cycles (mm)
Total Displacement (mm)
Ultimate Load (N)
SEM
Range
Bio-Corkscrew FT CrossFT PEEK TwinFix PK FT Healix PEEK PopLok PEEK PushLock Footprint Versalok
5.6 6.5 6.2 6.6 11.4 9.5 10.3 9.8
6.8 7.9 7.3 7.5 13.3 13.8 13.2 11
7.3 9 8.3 7.9 14.8 15.1 16.9 11.6
1.2 1.4 1.1 0.9 1.9 4.3 2.9 1.2
1.7 2.5 2.1 1.3 3.4 5.6 6.6 1.8
209.3 269.3 263.7 295.9 209.8 201.7 163.2* 308
27.9 35.0 30.9 38.1 20.2 42.7 31.4 18.5
66-343 64-425 47-427 80-459 104-304 44-413 28-296 212-426
*The Footprint was weaker than the Versalok (P ⬍ .05).
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS The 5-mm displacement gap threshold was the only “failure level” observed in this testing (Table 3). The Bio-Corkscrew FT, TwinFix PK FT, and Versalok did not have any tests exceeding 5 mm of displacement. The other anchors had 1 test exceeding the 5-mm displacement threshold and the PushLock had 2. What may be more telling is that only 6 of 12 PushLock anchors and 5 of 12 Footprint anchors survived 500 cycles, of which 2 PushLock anchors and 1 Footprint anchor exceeded 5 mm of displacement (Table 3). The Versalok anchor was the only lateral-row anchor for which all anchors tested survived 500 cycles. The mode of failure was also recorded (anchor pullout, suture breakage, and eyelet failure) (Table 4). In addition, the 1 specimen with very osteoporotic bone resulted in such weak anchors that most failed when being connected to the MTS testing device. This event was recorded as bone failure. The principal mode of failure was by anchor pullout. DISCUSSION The purpose of this study is to test medial-row and lateral-row suture anchors under cyclic loading conditions. Advocates of double-row rotator cuff repair highlight the added strength provided by the additional fixation points and the larger repair footprint created with more tissue coverage of the greater tuberosity.8-11 The presumption is that these factors lead to better tendonto-bone healing. One common double-row rotator technique connects the sutures from the medial anchor row in a crossing fashion to the anchors of the lateral row.6,19-21 This is referred to as the cross-stitch or suture-bridge technique. To accomplish this, the lateral-row anchors must be able to accept sutures from the previously placed anchors in their eyelet and then lock these sutures in place. These lateral-row anchors must have a cinching or locking eyelet mechanism.
TABLE 3. Number of Anchors Completing Cyclic Loading Phase and Those Showing 5 mm of Displacement Anchor
Tests Reaching 500 Cycles
Tests Exceeding 5 mm
Bio-Corkscrew FT CrossFT PEEK TwinFix PK FT Healix PEEK PopLok PEEK PushLock Footprint Versalok
8 8 10 10 9 6 5 12
0 1 0 1 1 2 1 0
TABLE 4.
S139
Mode of Failure
Anchor
Anchor Pullout
Suture Broke
Eyelet Broke
Bone Failure
Bio-Corkscrew FT CrossFT PEEK TwinFix PK FT Healix PEEK PopLok PEEK PushLock Footprint Versalok
10 12 9 10 11 7 8 10
0 0 1 0 0 0 0 2
1 0 2 1 0 3 3 0
1 0 1 1 1 2 1 0
Trabecular, cortical, and total bone density is greater near the articular cartilage but declines at more lateral positions on the greater tuberosity (P ⬍ .01), making the lateral anchor locations inherently weaker insertion sites and more likely to be associated with tendon-bone gap formation.22 In recognition of this, significant design differences exist between the new anchors used for a lateral-row application and those used for medial rows. These design differences take into account the variable nature of the bone into which the anchors are inserted, the mechanism for holding the suture, the number of sutures accommodated, and possibly, the anchor size. Previous cyclic loading studies have reported that more displacement (repair separation) occurs during the initial 100 test cycles than occurs between 100 and 500 cycles.4,17 In our test there was a clear difference in the behavior between the medial-row and lateralrow anchors. Whereas more displacement occurred in the first 100 cycles for all anchors except the Footprint (a lateral-row anchor), the amount of that displacement was much greater for the lateral-row anchors than the medial-row anchors (mean displacement of 1.2 mm at 100 cycles for the medial-row anchors compared with 2.6 mm for the lateral-row anchors). The displacement shown in the 100 cycle–to–500 cycle portions of the test was also much greater for the lateral-row anchors (mean, 1.8 mm) than the medialrow anchors (mean, 0.75 mm). The total displacement observed for all anchors was also greater for the lateral-row anchors (mean, 4.4 mm) compared with the medial-row anchors (mean, 1.9 mm), and the mean 4.4 mm of displacement approached the 5.0 mm of displacement defined as a failure. This means that both medial- and lateral-row anchor performance was consistent with prior studies. What is significant is that the lateral-row anchors, perhaps as a result of their design, were found to slip much more than the medialrow anchors both in the first 100 cycles, when most of
S140
F. A. BARBER ET AL.
the slippage occurs, and continuing into the interval between 100 and 500 cycles. No anchor that survived 500 cycles showed 10 mm of displacement. As noted in these data, more medial-row anchors completed the 500 cycles than lateral-row anchors. Early lateral-row anchor failure may be because of osteoporotic bone at the insertion site or anchor design. This also suggests that these lateral anchor designs work better in conjunction with medial-row anchors, which take up most of the tension. A standalone application for the lateral-row anchors should be considered with caution. The mean ultimate failure loads (Table 2) ranged from 163 to 308 N. These figures include those anchors that failed during cyclic loading. Statistical analysis of these load values showed only 1 significant result: the Versalok anchor was statistically stronger than the Footprint anchor (P ⬍ .05). The principal mode of failure for all anchors was by anchor pullout (Table 4). The next most common failure mode was eyelet breakage. The failure load at which eyelet breakage was recorded was distinctly different for medial-row anchors and lateral-row anchors. The lateral-row anchors PushLock and Footprint both had 3 tests fail by eyelet breakage. Eyelet failure in all but 1 of these 6 lateral-row anchors occurred at less than 500 cycles and consequently was at a lower failure load than the mean for the particular anchor (PushLock eyelet failures at 83 N v PushLock mean of 253 N and Footprint eyelet failures at 105 N v Footprint mean of 253 N). The medial-row anchors’ behavior was in marked contrast. All eyelet failures in medial-row anchors occurred after successfully completing the 500 cycles and, except for 1 anchor, at loads higher than the mean failure load for that anchor. The lateral-row anchors tested differ from the medialrow anchors tested in both fundamental design and performance. Lateral-row anchors accommodate more suture strands, as well as sutures from other anchors, and fix these sutures with cinching or locking mechanisms. Medial-row anchors use conventional eyelet posts and rely on tied knots for suture security. These design differences allow the sutures held in lateralrow anchors to slip if the loads are high enough. In contrast, the sutures in medial-row anchors are dependent on the knot security to avoid slippage. The isolated use of an anchor designed for lateral-row applications is likely to provide fixation inferior to that of a medial-row anchor with a greater likelihood of suture slippage in a cyclic loading model. However, the function of the medial- and lateral-row anchors tested
in this study was not tested specifically in terms of rotator cuff loading in this study setup. Strengths of this study include the use of paired specimens of human bone, the distribution of the anchors among 3 different medial and 3 different lateral insertion sites, and the cyclic loading protocol that applied stress in line with the direction of a physiologic tendon pull. Limitations of the study include the fact that testing was performed at room temperature and not in a water bath to more accurately simulate the clinical setting. A tendon repair was not performed, eliminating what is probably the most common cuff repair failure site. Space did not allow placement of 1 sample of each of the 4 different anchors tested in each location of every pair, but the paired specimens allowed for consistency of the bone insertion environments. The advanced age of the specimens could have affected the bone quality, although rotator cuff repair is common in older populations. One pair of humeri was very osteoporotic, and early anchor failure was observed. No radiographs were obtained to exclude patients with large cysts. This in vivo test cannot be applied directly to the clinical setting. This construct did not test single- or doublerow repairs but rather the isolated performance of anchors typically assigned a medial- or lateral-row function in the humeral site consistent with these roles. These data represent a time 0 response and do not take into account healing and the postoperative rehabilitation program. CONCLUSIONS Anchors designed to function as lateral-row fixation provide fixation strength inferior to that of medial-row anchors and are more likely to be subject to suture slippage. When lateral-row anchors fail at the eyelet, it is at lower failure loads, whereas if medial-row anchors fail at the eyelet, it is at higher loads. REFERENCES 1. Adams CR, Schoolfield JD, Burkhart SS. The results of arthroscopic subscapularis tendon repairs. Arthroscopy 2008;24: 1381-1389. 2. Castagna A, Garofalo R, Conti M, Borroni M, Snyder SJ. Arthroscopic rotator cuff repair using a triple-loaded suture anchor and a modified Mason-Allen technique (Alex stitch). Arthroscopy 2007;23:440.e1-440.e4. Available online at www.arthroscopyjournal.org. 3. Krishnan SG, Harkins DC, Schiffern SC, Pennington SD, Burkhead WZ. Arthroscopic repair of full-thickness tears of the rotator cuff in patients younger than 40 years. Arthroscopy 2008;24:324-328.
CYCLIC LOADING OF ROTATOR CUFF SUTURE ANCHORS 4. Barber FA, Coons DA, Ruiz-Suarez M. Cyclic load testing of biodegradable suture anchors containing 2 high-strength sutures. Arthroscopy 2007;23:355-360. 5. Bynum CK, Lee S, Mahar A, Tasto J, Pedowitz R. Failure mode of suture anchors as a function of insertion depth. Am J Sports Med 2005;33:1030-1034. 6. Burkhart SS, Adams CR, Schoolfield JD. A biomechanical comparison of 2 techniques of footprint reconstruction for rotator cuff repair: The SwiveLock-FiberChain construct versus standard double-row repair. Arthroscopy 2009;25:274-281. 7. Spang JT, Buchmann S, Brucker PU, et al. A biomechanical comparison of 2 transosseous-equivalent double-row rotator cuff repair techniques using bioabsorbable anchors: Cyclic loading and failure behavior. Arthroscopy 2009;25:872-879. 8. Trantalis JN, Boorman RS, Pletsch K, Lo IK. Medial rotator cuff failure after arthroscopic double-row rotator cuff repair. Arthroscopy 2008;24:727-731. 9. Reardon DJ, Maffulli N. Clinical evidence shows no difference between single- and double-row repair for rotator cuff tears. Arthroscopy 2007;23:670-673. 10. Nelson CO, Sileo MJ, Grossman MG, Serra-Hsu F. Single-row modified Mason-Allen versus double-row arthroscopic rotator cuff repair: A biomechanical and surface area comparison. Arthroscopy 2008;24:941-948. 11. Cole BJ, ElAttrache NS, Anbari A. Arthroscopic rotator cuff repairs: An anatomic and biomechanical rationale for different suture-anchor repair configurations. Arthroscopy 2007;23:662669. 12. Mahar A, Tamborlane J, Oka R, Esch J, Pedowitz RA. Singlerow suture anchor repair of the rotator cuff is biomechanically equivalent to double-row repair in a bovine model. Arthroscopy 2007;23:1265-1270. 13. Ji JH, Kim WY, Ra KH. Arthroscopic double-row suture anchor fixation of minimally displaced greater tuberosity frac-
14.
15.
16.
17. 18. 19. 20.
21. 22.
S141
tures. Arthroscopy 2007;23:1133.e1-1133.e4. Available online at www.arthroscopyjournal.org. Grasso A, Milano G, Salvatore M, Falcone G, Deriu L, Fabbriciani C. Single-row versus double-row arthroscopic rotator cuff repair: A prospective randomized clinical study. Arthroscopy 2009;25:4-12. Toussaint B, Schnaser E, Lafosse L, Bahurel J, Gobezie R. A new approach to improving the tissue grip of the medial-row repair in the suture-bridge technique: The “modified lasso-loop stitch.” Arthroscopy 2009;25:691-695. Coons DA, Barber FA, Herbert MA. Triple-loaded singleanchor stitch configurations: An analysis of cyclically loaded suture-tendon interface security. Arthroscopy 2006;22:11541158. Barber FA, Coons DA, Ruiz-Suarez M. Cyclic load testing and ultimate failure strength of biodegradable glenoid anchors. Arthroscopy 2008;24:224-228. Ruiz-Suarez M, Aziz-Jacobo J, Barber FA. Cyclic load testing and ultimate failure strength of hip arthroscopy suture anchors. Arthroscopy 2010;26:762–768. Kim KC, Rhee KJ, Shin HD. Deformities associated with the suture-bridge technique for full-thickness rotator cuff tears. Arthroscopy 2008;24:1251-1257. Lorbach O, Bachelier F, Vees J, Kohn D, Pape D. Cyclic loading of rotator cuff reconstructions: Single-row repair with modified suture configurations versus double-row repair. Am J Sports Med 2008;36:1504-1510. Song HS, Williams GR Jr. Arthroscopic reduction and fixation with suture-bridge technique for displaced or comminuted greater tuberosity fractures. Arthroscopy 2008;24:956-960. Tingart MJ, Apreleva M, Zurakowski D, Warner JJ. Pullout strength of suture anchors used in rotator cuff repair. J Bone Joint Surg Am 2003;85:2190-2198.