- Email: [email protected]
All-Suture Anchors: Biomechanical Analysis of Pullout Strength, Displacement, and Failure Mode
All-Suture Anchors: Biomechanical Analysis of Pullout Strength, Displacement, and Failure Mode F. Alan Barber, M.D., and Morley A. Herbert, Ph.D.
Purpose: To evaluate the biomechanical and design characteristics of all-suture anchors. Methods: All-suture anchors were tested in fresh porcine cortical bone and biphasic polyurethane foam blocks by cyclic loading (10-100 N for 200 cycles), followed by destructive testing parallel to the insertion axis at 12.5 mm/s. Endpoints included ultimate failure load, displacement at 100 and 200 cycles, stiffness, and failure mode. Anchors tested included JuggerKnot (1.4, 1.5, and 2.8), Iconix (1, 2, and 3), Y-knot (1.3, 1.8, and 2.8), Q-Fix (1.8 and 2.8), and Draw Tight (1.8 and 3.2). Results: The mean ultimate failure strength of the triple-loaded anchors (564 42 N) was significantly greater than the mean ultimate failure strength of the double-loaded anchors (465 33 N) (P ¼ .017), and the double-loaded anchors were stronger than the single-loaded anchors (256 35 N) (P < .0001). No difference was found between the results in porcine bone and biphasic polyurethane foam. None of these anchors demonstrated 5 mm or 10 mm of displacement during cyclic loading. The Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix (P ¼ .025) but not the Iconix and Draw Tight (P > .05). The most common failure mode varied and was suture breaking for the Q-Fix (97%), JuggerKnot (81%), and Iconix anchors (58%), anchor pullout with the Draw Tight (76%), whereas the Y-Knot was 50% suture breaking and 50% anchor pullout. Conclusions: The ultimate failure load of an all-suture anchor is correlated directly with its number of sutures. With cyclic loading, the Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix but not the Iconix and Draw Tight. JuggerKnot (81%) and Q-Fix (97%) anchors failed by suture breaking, whereas the Draw Tight anchor failed by anchor pullout (76%). Clinical Relevance: All-suture anchors vary in strength and performance, and these factors may influence clinical success. Biphasic polyurethane foam is a validated model for suture anchor testing.
T
he importance of suture anchors to assist in the attachment of tendons, ligaments, and other soft tissue to bone is readily apparent, and these devices are used widely for most minimally invasive techniques, especially in the upper extremity. Over time, suture anchors have been improved with fully threaded designs, distally placed eyelets, the ability to
From Plano Orthopedic Sports Medicine and Spine Center (F.A.B.), Plano; and Advanced Surgical Institutes, Medical City Dallas Hospital (M.A.H.), Dallas, Texas, U.S.A. The authors report the following potential conflicts of interest or sources of funding: F.A.B. and M.A.H. receive support from Stryker Endoscopy, Smith & Nephew Endoscopy, ConMed-Linvatec, Biomet Sports Medicine, and Cayenne Medical (funding for Instron costs from the anchor makers). This study was funded by support from the F. Alan Barber, M.D., F.A.C.S., Research Foundation, and the donation of anchors by Biomet Sports Medicine, Stryker Endoscopy, ConMed Linvatec, Smith & Nephew, and Parcus Medical. Received June 6, 2016; accepted September 26, 2016. Address correspondence to F. Alan Barber, M.D., Plano Orthopedic Sports Medicine and Spine Center, 5228 West Plano Parkway, Plano, TX 75093, U.S.A. Ó 2016 by the Arthroscopy Association of North America. Published by Elsevier. All rights reserved. 0749-8063/16518/$36.00 http://dx.doi.org/10.1016/j.arthro.2016.09.031
accommodate multiple sutures, the use of first biodegradable and later biocomposite materials, and knotless designs. Suture anchors made completely of suture material are a recent development.1,2 These all-suture anchors are based on one of more ultra-highmolecular-weight polyethylene (UHMWPE)-containing sutures. The anchor portion of the device typically consists of a sleeve or tape also made from suture material through which the UHMWPE containing suture is woven. When the all-suture anchor is inserted into bone and the main suture pulled, the sleeve or tape is cinched up to compress against the overlying cortical bone creating a “ball,” which serves as the anchor. As with conventional anchors, these all-suture anchors are configured for glenoid-labral and tuberosity-tendon applications. The glenoid anchors are smaller and usually have 1 or at most 2 sutures.1,3 The rotator cuff anchors are larger and are either double or triple loaded. All-suture anchors are radiolucent, nondegradable, and concern has been raised about the development of cyst formation at the anchor site.4 An understanding of the failure mode and strength characteristics of these all-suture anchors is necessary for the surgeon considering their use. The purpose of this
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol
-,
No
-
(Month), 2016: pp 1-9
1
2
F. A. BARBER AND M. A. HERBERT
study was to evaluate the biomechanical and design characteristics of all-suture anchors. Our hypothesis was that these anchors do not have time zero failure profiles or biomechanical properties that are materially different from conventional suture anchors. A secondary hypothesis was that the all-suture anchor performance could be reliably assessed in either porcine bone or in biphasic polyurethane foam blocks.
Methods All-suture anchors of each type to be evaluated were obtained from their respective manufacturer for testing and implanted into 2 different substrates. The goal was to have 10 samples of each anchor type and size for placement in the diaphyseal cortex of fresh adult porcine femurs obtained from a local abattoir.1,3 Each anchor tested was rotated among at least 5 different femurs, and the insertion sites also were rotated among different positions on the diaphysis to minimize the potential effect of variations in bone thickness. No more than 2 anchors of each type were placed in any one femur. Later, as the literature began to demonstrate the validity of testing implants in polyurethane foam, whenever possible, an additional 10 samples of each anchor type and size were placed in biphasic polyurethane foam blocks (Pacific Research Laboratories, Vashon, WA). The biphasic polyurethane foam material was selected to replicate the environment of the cortical bone in the shoulder. The greater tuberosity and glenoid may offer different cortical thickness, especially if abrasion of the anchor insertion site is done in the greater tuberosity. For an all-suture anchor to deploy, however, it does require some cortex against which to be pulled and compressed. A robust biphasic “cortex” was selected to eliminate this feature as a point of failure in this test system. This biphasic material consisted of a solid rigid polyurethane foam block with a density of 12 pcf with a fiber filled epoxy coating similar in density, hardness and strength to cortical bone laminated on top. Appropriately sized drill holes were placed through the dense simulated cortex to allow access by the anchor into the polyurethane foam beneath. Every all-suture anchor tested was based on UHMWPE-containing suture. The all-suture anchors were separated in either the porcine bone or biphasic polyurethane foam block by at least 1 cm from any adjacent anchor to prevent crack propagation between drill holes during testing. All anchors were inserted according to the manufacturer’s instructions with the appropriate instruments by the senior author (F.A.B.) or by a company product manager or engineer if available to assure adequate familiarity with the anchor and instrumentation. Both anchor insertion and pullout testing were conducted with the bones or biphasic
polyurethane foam blocks at room temperature in a non-aqueous environment. The biomechanical testing was performed (M.A.H.) by securing the femurs or foam blocks holding the anchors to a platform directly under the actuator arm of a mechanical materials testing machine (model 3345; Instron, Canton, MA). The femurs were placed in a specially prepared aluminum box that supported the bone and automatically aligned the anchors’ sutures directly under the actuator arm of the mechanical materials testing machine. The blocks were secured on the sides by clamps secured to the base plate. Thus, the load applied was always in line with the axis of anchor insertion. The sutures were secured in the upper hydraulic fixture with a constant gauge length. The upper arm of the Instron machine was positioned so that there was no load on the device and then under program control, a preload of 10 N was applied. After the preload, a cyclic load alternating between 10 N and 100 N was applied at 0.5 Hz for 200 cycles or until failure occurred. A gauge length of 30 mm was used for all tests to start and the gauge length after the initial cycle considered the baseline. After completion of 200 cycles, destructive testing was performed at a displacement rate of 12.5 mm/s. Data sampling of load and displacement was obtained at 100 samples per second. The number of cycles needed to reach both 5 mm and 10 mm of displacement was recorded if it occurred.1,3,5,6 Endpoints of this study included anchor dimensions and characteristics, the ultimate load at failure, displacement at 100 and 200 cycles (initial displacement was calculated at 10 cycles), stiffness, and mode of failure (M.A.H.) (anchor pull out, eyelet/suture cut out, or suture breakage). The anchors tested varied in size, suture material and number, and characteristics, which are listed in Table 1. All companies producing anchors were invited to provide their anchors for this test. Those who were willing donated an appropriate number of anchors along with the associated insertion equipment and personnel to assist in the anchor insertion. Some anchors were limited in their supply and not available for testing in the polyurethane foam blocks. The anchors tested are listed in the paragraphs to follow. JuggerKnot 1.4 mm, 1.5 mm, and 2.8 mm all-suture anchors (Biomet Sports Medicine, Warsaw, IN) (Fig 1) are composed of 1 or 2 (blue/white) size No. 1 or 2 braided UHMWPE sutures that pass through a flexible tube of No. 6 braided polyester material measuring 2 mm wide and between 20 and 24 mm in length. The size of the polyester tube varies depending on the anchor. This sleeve bunches up under a cortical surface, creating a ball that provides the anchoring effect. The JuggerKnot 1.4 has a single No. 1 braided UHMWPE suture, whereas the JuggerKnot 1.5 has a single No. 2
3
ANALYSIS OF ALL-SUTURE ANCHORS Table 1. All-Suture Anchor Physical Characteristics Anchor JuggerKnot 1.4 JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-knot 1.3 Y-knot 1.8 Y-knot 2.8 Q-Fix 1.8 Q-Fix 2.8 Draw Tight 1.8 Draw Tight 3.2
Anchor Material Polyester Polyester Polyester Polyester Polyester Polyester UHMWPE UHMWPE UHMWPE Polyester Polyester UHMWPE UHMWPE
Suture Size #1 #2 #2 #2 #2 #2 #2 #2 #2 #2 #2 #2 #2
Suture Number Single Single Double Single Double Triple Single Double Double/triple Single Double Single Double
Drill Diameter, mm 1.4 1.5 2.9 1.4 2.3 2.3 1.3 1.8 2.8 2.0 3.0 1.8 3.2
Sleeve Length, mm 20.3 24.4 25.4 25 25 25 25 25 32 15 20 17.8 28.6
Drill Length, mm 24 24 20.8 20 20 20 21 21 23 22.3 26.8 19 19
UHMWPE, ultra-high-molecular-weight polyethylene.
braided UHMWPE suture. The JuggerKnot 2.8 is double loaded with No. 2 braided UHMWPE sutures. Iconix (1, 2, and 3) all-suture anchors (Stryker Endoscopy, San Jose, CA) (Fig 2) are provided in 3 versions, the Iconix 1, 2, and 3, reflecting the number of strands of No. 2 braided UHMWPE suture they possess. These sutures are woven 3 times through a flat, flexible tube of braided polyester and pushed into a predrilled hole. When tensioned, the braided tube collapses into a “clover leaf” configuration, creating the anchor. Y-Knot (1.3 mm 1.8 mm, and 2.8 mm) all-suture anchors (ConMed Linvatec, Largo, FL) (Fig 3) are of different sizes, containing 1, 2, or 3 No. 2 braided UHMWPE sutures. In contrast to the JuggerKnot and Iconix anchors, the Y-Knot’s anchor portion is created from a flat braided UHMWPE tube threaded with the No. 2 UHMWPE suture. When pulled securely against
Fig 1. The JuggerKnot anchors shown are, from left to right, 2.9, 1.5, and 1.4. The associated No. 1 or 2 ultra-highmolecular-weight polyethylene (UHMWPE) sutures pass through a 2-mm wide flexible No. 6 braided polyester tube. This tube bunches into a ball, creating the subcortical anchor. The JuggerKnot 1.4 has one No. 1 UHMWPE suture, the JuggerKnot 1.5 has one No. 2 UHMWPE suture, and the 2.8 has two No. 2 UHMWPE sutures. (Ó 2016 F. Alan Barber. All Rights Reserved.)
the overlying cortical bone, this device creates a ball of UHMWPE material. The Y-Knot 2.8 is designed for rotator cuff applications and is provided as a doubleloaded or triple-loaded anchor. The Q-Fix (1.8 mm and 2.8 mm) all-suture anchor (Smith & Nephew, Andover, MA) (Fig 4) is based on a single (the 1.8 version) or double (the 2.8 version) No. 2 braided UHMWPE suture that is woven through a braided polyester tape. As with the others, the ball of material serves as the anchor when pulled against cortical bone. The 1.8-mm Q-Fix is designed for glenoid and acetabular rim applications, whereas the 2.8-mm Q-Fix is designed for rotator cuff repairs.
Fig 2. The Iconix anchors shown include, from left to right, the single-loaded Iconix 1 compressed into the “clover leaf” pattern, the Iconix 1 uncompressed, the double-loaded Iconix 2, and the triple-loaded Iconix 3. At the bottom is a double-loaded Iconix 2 demonstrating how the sutures are woven 3 times through a flat, flexible braided polyester tube. (Ó 2016 F. Alan Barber. All Rights Reserved.)
4
F. A. BARBER AND M. A. HERBERT
Fig 3. The Y-knot anchors shown are, from left to right, the triple-loaded Y-Knot 2.8 demonstrating how the sutures are woven through the flat braided ultra-high-molecularweight polyethylene tube, the single-threaded Y-knot 1.3, dual-threaded Y-knot 1.8, and again the triple-loaded Y-knot 2.8 on its inserter. At the bottom is a dual-threaded Y-knot 1.8 compressed, demonstrating the compressed configuration, which as a “W” appearance. (Ó 2016 F. Alan Barber. All Rights Reserved.)
anchor. The mean ultimate failure strength of the triple loaded anchors (564N 42; 95% confidence interval [CI] 509.5-618.8) was significantly greater than the mean ultimate failure strength of the double-loaded anchors (465 33; 95% CI 417.5-512.2) (P ¼ .017), and the double-loaded anchors were stronger than the single-loaded anchors (256 35; 95% CI 219.6-291.8) (P < .0001). Because the triple-loaded anchors are larger and designed for rotator cuff applications, it follows that the glenoid (smaller and single-loaded) anchors have lower mean ultimate failure strengths than the rotator cuff anchors. A comparison of the data generated in the porcine cortical bone and biphasic polyurethane foam blocks demonstrated no difference in the means between ultimate failure loads for the individual anchors but much narrower ranges and standard deviations. The displacement of the constructs was recorded during cyclic loading. The mean cyclic displacement for each anchor is reported in Table 3. Displacement of 5 mm is considered a significant threshold in a biomechanical test.5,6 None of these anchors reached either 5-mm or 10-mm displacement. The Y-Knot demonstrated greater
The Draw Tight (1.8 mm and 3.2 mm) device (Parcus Medical, Sarasota, FL) (Fig 5) is based on a single or double No. 2 braided UHMWPE suture. In contrast to the other all-suture based anchors, the Draw Tight has a small polyether ether ketone tip through which the suture and the sleeve pass, which facilitates insertion. The sleeve through which the suture is woven is braided UHMWPE. In the single-suture version the polyether ether ketone tip measures 1.9 mm 4.9 mm, whereas in the double-suture version it measures 3.0 mm 6.8 mm. Statistical Analysis Statistical analysis was carried out with SAS software (SAS Institute, Cary, NC). Data were read directly from the Excel spreadsheets (Microsoft, Redmond, WA) with the use of dynamic data exchange and then analyzed with SAS. Test data were analyzed for means and standard deviations. A Student t-test was used to determine differences between groups. An analysis of variance was used to test for effects from bone and failure mode differences before combining the data. Statistical significance was set at P < .05.
Results The ultimate failure loads for the tested anchors are reported in Table 2 for the porcine bone and when the anchors were available the biphasic polyurethane foam blocks. Anchor ultimate load at failure directly correlated to the number of sutures associated with the
Fig 4. The Q-Fix anchors shown are, from left to right, the single-loaded Q-Fix 1.8, the double-loaded Q-Fix 2.8, and the double-loaded Q-Fix 2.8 compressed into a ball. All sutures are No. 2 braided ultra-high-molecular-weight polyethylene which is woven through a braided polyester tape. Note how with compression the anchor width expands markedly to create a subcortical anchor. (Ó 2016 F. Alan Barber. All Rights Reserved.)
5
ANALYSIS OF ALL-SUTURE ANCHORS
The failure mode for these anchor tests is reported (Table 4). The attempt to test the Iconix 3 in biphasic polyurethane foam was unsuccessful because the biphasic resin fragmented and delaminated during cyclic loading, and only 3 tests survived the initial cycle. It should be noted that the JuggerKnot (81%) and Q-Fix (97%) anchors failed predominantly by the suture breaking whereas the Draw Tight anchor failed by anchor pullout (76%). A majority of the Iconix anchors failed by suture breaking, and the Y-Knot was evenly distributed in failure mode between suture breaking and anchor pullout.
Discussion
Fig 5. The Draw Tight anchors tested included, from left to right, the Draw Tight 1.8 and Draw Tight 3.2. Both are shown double threaded with No. 2 braided ultra-high-molecularweight polyethylene suture. The Draw Tight anchor has a polyether ether ketone tip through which the suture and the sleeve pass which facilitates insertion. (Ó 2016 F. Alan Barber. All Rights Reserved.)
displacement than the JuggerKnot and Q-Fix (P ¼ .025) but not the Iconix and Draw Tight (P > .05). None of the other anchors demonstrated a statistical difference in displacement. The stiffness was not statistically different for the different anchor groups.
For the all-suture anchors, ultimate load at failure correlated directly with the number of UHMWPEcontaining sutures in the anchor. Triple-loaded anchors (564 42 N) were stronger than double-loaded anchors (465 33 N) (P ¼ .017), which in turn were stronger than single-loaded anchors (256 35 N) (P < .0001). Triple (and possibly double)-loaded all-suture anchors are larger and generally designed for rotator cuff repair whereas the smaller, single (and possibly double)-loaded anchors generally are designed for glenoid applications. None of the all-suture anchors reached a clinically significant 5-mm displacement during cyclic loading. Individual anchor performances were compared, and the Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix (P ¼ .025) but not the Iconix and Draw Tight (P > .05). The anchor dimensions and characteristics are listed in Table 1. What is notable about the all-suture anchors is the greater length and longer associated drills needed for insertion when compared with conventional suture anchors.1,3
Table 2. Data From Load to Failure Testing in Both Porcine Bone and When Available Polyurethane Foam Is Presented Porcine Cortical Bone Anchor JuggerKnot 1.4 JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-knot 1.3 Y-knot 1.8 Y-knot 2.8 Q-Fix 1.8 Q-Fix 2.8 Draw Tight 1.8 Draw Tight 3.2
N 10 14 14 11 12 12 12 10 20 19 20 20 20 19
Failure Load, N 239.1 15.1 (215-263) 290.5 15.3 (263-325) 519.3 51.4 (455-620) 208.7 69 (121-298) 468.3 570.3 249.7 477.4 602.9 346.1 495.1 290.3 418.5
45.2 (412-548) 47 (510-654) 53.5 (161-302) 78.7 (265-590) 159 (288-765) 92 (285-602) 87.9 (288-585) 127.7 (171-514) 111.1 (134-548)
NOTE. Ranges are in parentheses.
Displacement, mm 0.22 0.22 0.22 0.31 0.23 0.20 0.45 0.33 0.55 0.19 0.23 0.30 0.30
Polyurethane Foam Stiffness, N/mm 198 98 57 3.3 76 8.3 65.2 2.1 82.9 88.9 64.8 74.1 84.1 55 56.8 41.5 49.4
5.7 4.7 4.2 5.6 3.7 3.4 6.8 5.6 10.5
Failure Load, N
Displacement, mm
Stiffness, N/mm
234.7 73.6 (115-351)
0.23
72.5 6.1
519.9 50.8 (434-584)
0.43
86.0 4.9
0.23 0.19 0.11 0.16 0.24 0.20
75.4 81.8 55 56.8 41.5 49.4
151.8 531.1 657.1 291.8 494.7 262.7 190.8
12.6 (135-174) 33.3 (485-590) 128.1 (324-766) 17.6 (258-316) 106 (213-578) 91.5 (148-401) 98.8 (134-451)
7.5 2.5 3.4 6.8 5.6 10.5
6
F. A. BARBER AND M. A. HERBERT
Table 3. Displacement During Cyclic Loading After 100 Cycles and 200 Cycles Anchor JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-Knot 1.3 Y-Knot 1.8 Y-knot 2.9 Draw Tight 1.8 Draw Tight 3.2 Q-Fix 1.8 Q-Fix 2.8
Displacement 100 Cycles, mm 1.17 0.46 1.22 0.47 1.56 0.37 1.32 0.36 1.24 0.33 1.95 0.79 1.67 0.51 2.97 1.55 1.82 0.36 2.32 0.8 1.03 0.24 1.35 0.31
Tests 14 14 11 12 12 19 20 19 20 19 20 19
Displacement 200 Cycles, mm 1.39 0.57 1.44 0.57 1.87 0.48 1.55 0.39 1.44 0.35 2.40 0.96 2.0 0.6 3.52 1.79 2.12 0.42 2.62 0.65 1.22 0.27 1.58 0.34
Failure mode was evaluated, and major differences were found in the common modes of failure for these all-suture anchors. Both JuggerKnot (81%) and Q-Fix (97%) anchors failed predominantly by the suture breaking, whereas the Draw Tight anchor failed by anchor pullout (76%). This finding may have clinical significance. It is accepted that suture breaking or the suture pulling out of the anchor is clinically better than having the anchor itself pull out of the bone.3 This is not to minimize the fact that even a loose suture in the joint can be destructive.7 With an all-suture based anchor, anchor pullout can result in a suture-ball within the joint, which can cause unappreciated damage because these are not readily identified on routine radiographs. It should be pointed out that this destructive potential also exists with polyether ether ketone, biodegradable, and biocomposite anchors. This polyurethane foam model was biphasic with a dense polymer resin to simulate cortical bone against which the all-suture anchor was secured. The polyurethane foam modeled the human cortical bone, providing a test substrate that demonstrated consistent performance with low standard deviations. Differences exist between the anchor insertion sites in the glenoid and greater tuberosity. Patient age plays a role, because Table 4. Mode of Failure for the Anchors Tested Anchor JuggerKnot 1.4 JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-Knot 1.3 Y-Knot 1.8 Y-knot 2.9 Q-Fix 1.8 Q-Fix 2.8 Draw Tight 1.8 Draw Tight 3.2
Tests 20 14 14 23 22 12 19 20 19 20 19 18 19
Anchor Pullout 9
Anchor Break
3 1
9 11
14 7 8 1 12 16
1 2
Suture Break 11 14 14 11 21 1 5 13 11 20 18 5 1
Change, mm 0.22 0.22 0.31 0.23 0.20 0.45 0.33 0.55 0.30 0.33 0.19 0.23
Displaced, 5 mm/10 mm 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0
typically glenoid anchors are used to treat instability, which occurs in younger individuals, whereas greater tuberosity anchors are used to treat rotator cuff tears occurring in older individuals. Consequently, the glenoid bone has a more robust cortex ideally suited to activate an all-suture anchor, whereas the thinner greater tuberosity cortex may be compromised by surface abrading designed to promote healing. The biphasic polyurethane foam model was selected to simulate a femoral diaphyseal test environment that would efficiently activate the all-suture anchor deployment mechanism. The desire was to remove the substrate as a mode of failure and simulate the biologic bone model against which the foam data were compared. Polyurethane foam has been used many times for this type of testing.8-12 The use of a polyurethane foam block has been validated for tibial boneepatellar tendonebone graft testing.13 In that study, the human cadaver tibia demonstrated far more variability than the foam block and did not secure the bone plugs as effectively as the polyurethane foam altering the mode of failure. Its use for suture anchors has also been documented12,14,15 as well as testing the effects of suture anchor orientation.16 Polyurethane foam is much less expensive than a biologic specimen, readily available, with no infection risk, and no disposal challenges. These all-suture anchors typically have a 1- to 3-mm wide sleeve of suture material (either UHMPWE or braided polyester) through which the base suture is passed. The Iconix anchor and the Y-Knot have their base sutures woven though the sleeve, whereas the JuggerKnot anchors have the suture passed down the sleeve lumen. This influences the shape of the anchor when it is tightened (Figs 2-4). Displacement of 5 mm or more often is considered a significant threshold in a biomechanical test.5,6 None of the anchors tested demonstrated displacement of 5 mm or more during cyclic loading. The minimum load required for a rotator cuff repair has not been established clinically and probably can
ANALYSIS OF ALL-SUTURE ANCHORS
vary with the bone and rotator cuff tendon quality; however, a load of 250 N has been identified by Mazzocca et al.17 and others5,18 as the load that is required for the rehabilitation phase of rotator cuff tendon repair. The anchors designed for rotator cuff applications successfully exceeded this level. Specifically, the JuggerKnot 2.9 (519 N), Iconix 3 (570 N), Y-knot 2.8 (602N), Q-Fix 2.8 (495N), and Draw Tight 3.2 (418 N) all had mean failure loads above the 250 N level. The biologic bone model used was not in metaphyseal bone as might be found at the greater tuberosity. Another distinctive feature of these all-suture anchors is the drill length. As noted in Table 1, the drill lengths range from 19 to 27 mm. This drill length is longer than most conventional suture anchors.1,3 Although probably not a significant concern for rotator cuff applications, a long drill can penetrate the superior glenoid cortex associated with a SLAP repair, reaching the suprascapular nerve, or the inferior glenoid, reaching the axillary nerve.19 The most common conventional suture anchor design is a fully threaded screw-in anchor designed to achieve its greatest purchase in the cortical and subcortical bone. The biomechanical problems associated placing anchors deeper than the cortex are well known.20,21 The drill hole for the all-suture anchors are not filled with rigid polymer or metal, thus creating the potential for these holes to act as stress risers increasing the risk of subsequent glenoid fracture.4 This risk is difficult to calculate, although there are case reports of glenoid fractures associated with suture anchors.22 The biomechanical superiority of suture anchors in comparison with bone tunnels was established in the 1990s,5,6,23,24 and one of the most significant modes of failure is the suture cutting through the bone with cyclic loading.16 The increased abrasiveness of some of the current UHMWPE-containing sutures and tapes is well known,25-27 and with an anchor located in a subcortical position the pivot point of the suture loading may result in some cutting through of the overlying cortex.16 The force required to adequately “set the anchor” manually was not determined in this study. Manually inserting the all-suture anchor and pulling back on the suture to “set the anchor” did not result in significant anchor displacement during cyclic loading. With the exception of the 2.9 Y-Knot anchor, none of the anchors approached the 5-mm displacement level (Table 3). The Y-Knot and Draw tight anchors did have some anchors with an absolute displacement of 3 mm, but this should not be considered clinically applicable.3,5,6,28,29 What was measured was the mean displacement change and that was less than 0.5 mm for almost all of these anchors (Table 3). Pfeiffer et al.4 performed a biomechanical and histological study in a canine glenoid rim model and compared the JuggerKnot 1.4 with a conventional
7
2.4-mm fully threaded anchor. At 8 weeks, all JuggerKnot 1.4 anchor sites became cyst-like cavities (mean of 6.3 mm in diameter) with dense lamellar bone rims in stark contrast to the conventional 2.4-mm fully threaded anchor sites which contained intact anchors (mean of 2.7 mm in diameter) closely approximated to trabeculae of the adjacent lamellar bone.4 Biomechanical testing demonstrated significantly greater displacement at ultimate failure load for the JuggerKnot anchors (13.7 6.6 mm) compared with the conventional anchors (3.2 0.5 mm) (P ¼ .0009) as well as greater mean cyclic displacement (2.9 mm compared with 1.3 mm, respectively; P ¼ .0001). Mazzocca et al.2 compared these 2 anchors biomechanically using human cadaver glenoids simulating tears of the anterior-inferior and posterior-inferior labrum. Although similar ultimate failure loads were demonstrated (JuggerKnot 146 N and the conventional anchor 172 N) for these anchors, the solid anchor required significantly higher loads to achieve 2 mm of labral displacement (JuggerKnot 39.2 N and the conventional anchor 84.1 N; P < .001). Dwyer et al.12 expanded on the model of Mazzocca et al. and compared the 1.3 Y-Knot with a conventional 3.1-mm screw-in anchor. These were compared in human cadaver glenoids, bovine bone, and polyurethane foam. The Y-Knot also was either hand tensioned or pretensioned by a machine to 60 N. In the human cadaveric glenoid, after cyclic loading the ultimate failure loads for the 1.3 Y-Knot (91 N if hand set or 145 N if pretensioned) were comparable with the conventional anchor (107 N). These load levels were significantly greater in bovine bone and in polyurethane foam, with no difference being demonstrated between the pretensioned 1.3 Y-Knot and the conventional anchor. Chiang et al.30 performed a biomechanical analysis of a biceps tenodesis in a human humerus comparing techniques using either a 1.3 Y-Knot all-suture anchor or an 8 23-mm biocomposite interference screw fixed 15 mm proximal to the inferior border of the pectoralis major tendon insertion. Interestingly the 1.3 Y-Knot demonstrated a mean failure load of 239.1 N which compares favorably with the mean failure load for the 1.3 Y-knot of 249 N in our study. The biocomposite interference screw failure load of 254.4 N was not statistically different (P ¼ .87). Limitations Study limitations include the fact that this was a nonarthroscopically performed test, done in nonaqueous conditions at room temperature with the use of porcine bone or biphasic polyurethane foam. Because of the limited device availability, not all of the anchors tested in porcine bone also were tested in biphasic polyurethane foam. Specifically, the
8
F. A. BARBER AND M. A. HERBERT
Juggerknot anchors were not available for testing in foam and the large Iconix 3 anchors broke the foam substrate; however, all the data available were reported to consolidate the data in a single manuscript and avoid future issues with duplicate data publication. A power analysis was not performed because the number of anchors required to achieve a power of 0.8 comparing this many anchors and this many different sizes would be prohibitively large. This is a bench test and the data cannot be applied directly to a clinical setting. Clinical performance cannot be extrapolated from this information. Testing protocols that use little or no cyclic loading may demonstrate different data. The study was focused solely on all-suture based anchors, and other anchor types were not included. The testing was not carried out in age appropriate human glenoid or human greater tuberosity bone. The angle of load application was in line with the insertion angle to be consistent with past anchor tests and to attempt to produce a worst-case scenario. It does not replicate the “transosseous equivalent” loading.31,32 The weakest link in any rotator cuff tendon-suture-anchor construct is the tendon-suture interface,33-35 and the UHMWPEcontaining sutures increase the difference between the tissue strength and suture strength making tissue failure more likely if healing has not occurred.
Conclusions The ultimate failure load of an all-suture anchor is correlated directly with its number of sutures. With cyclic loading, the Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix but not the Iconix and Draw Tight. JuggerKnot (81%) and QFix (97%) anchors failed by suture breaking whereas the Draw Tight anchor failed by anchor pullout (76%).
Acknowledgment The authors wish to acknowledge the help of Kathy Berry and Eric Sanders with specimen preparation.
References 1. Barber FA, Herbert MA, Hapa O, et al. Biomechanical analysis of pullout strengths of rotator cuff and glenoid anchors: 2011 update. Arthroscopy 2011;27:895-905. 2. Mazzocca AD, Chowaniec D, Cote MP, et al. Biomechanical evaluation of classic solid and novel all-soft suture anchors for glenoid labral repair. Arthroscopy 2012;28: 642-648. 3. Barber FA, Herbert MA. Cyclic loading biomechanical analysis of the pullout strengths of rotator cuff and glenoid anchors: 2013 update. Arthroscopy 2013;29:832-844. 4. Pfeiffer FM, Smith MJ, Cook JL, Kuroki K. The histologic and biomechanical response of two commercially available small glenoid anchors for use in labral repairs. J Shoulder Elbow Surg 2014;23:1156-1161.
5. Burkhart SS, Diaz Pagan JL, Wirth MA, Athanasiou KA. Cyclic loading of anchor-based rotator cuff repairs: confirmation of the tension overload phenomenon and comparison of suture anchor fixation with transosseous fixation. Arthroscopy 1997;13:720-724. 6. Burkhart SS, Johnson TC, Wirth MA, Athanasiou KA. Cyclic loading of transosseous rotator cuff repairs: Tension overload as a possible cause of failure. Arthroscopy 1997;13:172-176. 7. Barber FA. Biodegradable shoulder anchors have unique modes of failure. Arthroscopy 2007;23:316-320. 8. Aziz MS, Tsuji MR, Nicayenzi B, et al. Biomechanical measurements of stopping and stripping torques during screw insertion in five types of human and artificial humeri. Proc Inst Mech Eng H 2014;228:446-455. 9. Bhimji S, Meneghini RM. Micromotion of cementless tibial baseplates under physiological loading conditions. J Arthroplasty 2012;27:648-654. 10. Zdero R, Elfallah K, Olsen M, Schemitsch EH. Cortical screw purchase in synthetic and human femurs. J Biomech Eng 2009;131:094503. 11. Nicayenzi B, Crookshank M, Olsen M, et al. Biomechanical measurements of cortical screw stripping torque in human versus artificial femurs. Proc Inst Mech Eng H 2012;226:645-651. 12. Dwyer T, Willett TL, Dold AP, et al. Maximum load to failure and tensile displacement of an all-suture glenoid anchor compared with a screw-in glenoid anchor. Knee Surg Sports Traumatol Arthrosc 2016;24:357-364. 13. Barber FA. Pullout strength of bone-patellar tendon-bone allograft bone plugs: A comparison of cadaver tibia and rigid polyurethane foam. Arthroscopy 2013;29:1546-1551. 14. Poukalova M, Yakacki CM, Guldberg RE, et al. Pullout strength of suture anchors: Effect of mechanical properties of trabecular bone. J Biomech 2010;43:1138-1145. 15. Yakacki CM, Griffis J, Poukalova M, Gall K. Bearing area: A new indication for suture anchor pullout strength? J Orthop Res 2009;27:1048-1054. 16. Bardana DD, Burks RT, West JR, Greis PE. The effect of suture anchor design and orientation on suture abrasion: An in vitro study. Arthroscopy 2003;19:274-281. 17. Mazzocca AD, Millett PJ, Guanche CA, Santangelo SA, Arciero RA. Arthroscopic single-row versus double-row suture anchor rotator cuff repair. Am J Sports Med 2005;33:1861-1868. 18. Schneeberger AG, von Roll A, Kalberer F, Jacob HA, Gerber C. Mechanical strength of arthroscopic rotator cuff repair techniques: An in vitro study. J Bone Joint Surg Am 2002;84:2152-2160. 19. Morgan RT, Henn RF 3rd, Paryavi E, Dreese J. Injury to the suprascapular nerve during superior labrum anterior and posterior repair: Is a rotator interval portal safer than an anterosuperior portal? Arthroscopy 2014;30: 1418-1423. 20. Mahar AT, Tucker BS, Upasani VV, Oka RS, Pedowitz RA. Increasing the insertion depth of suture anchors for rotator cuff repair does not improve biomechanical stability. J Shoulder Elbow Surg 2005;14:626-630. 21. 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.
ANALYSIS OF ALL-SUTURE ANCHORS 22. Fritsch BA, Arciero RA, Taylor DC. Glenoid rim fracture after anchor repair: A report of 4 cases. Am J Sports Med 2010;38:1682-1686. 23. Barber FA, Herbert MA, Click JN. Internal fixation strength of suture anchorsdUpdate 1997. Arthroscopy 1997;13:355-362. 24. Sherman SL, Copeland ME, Milles JL, Flood DA, Pfeiffer FM. Biomechanical evaluation of suture anchor versus transosseous tunnel quadriceps tendon repair techniques. Arthroscopy 2016;32:1117-1124. 25. Deranlot J, Maurel N, Diop A, et al. Abrasive properties of braided polyblend sutures in cuff tendon repair: An in vitro biomechanical study exploring regular and tape sutures. Arthroscopy 2014;30:1569-1573. 26. Lambrechts M, Nazari B, Dini A, et al. Comparison of the cheese-wiring effects among three sutures used in rotator cuff repair. Int J Shoulder Surg 2014;8:81-85. 27. Williams JF, Patel SS, Baker DK, et al. Abrasiveness of high-strength sutures used in rotator cuff surgery: are they all the same? J Shoulder Elbow Surg 2016;25: 142-148. 28. Barber FA, Hapa O, Bynum JA. Comparative testing by cyclic loading of rotator cuff suture anchors containing multiple high-strength sutures. Arthroscopy 2010;26: S134-S141 (suppl 9). 29. Coons DA, Barber FA, Herbert MA. Triple-loaded singleanchor stitch configurations: An analysis of cyclically
30.
31.
32.
33.
34.
35.
9
loaded suture-tendon interface security. Arthroscopy 2006;22:1154-1158. Chiang FL, Hong CK, Chang CH, et al. Biomechanical comparison of all-suture anchor fixation and interference screw technique for subpectoral biceps tenodesis. Arthroscopy 2016;32:1247-1252. Dierckman BD, Goldstein JL, Hammond KE, Karas SG. A biomechanical analysis of point of failure during lateralrow tensioning in transosseous-equivalent rotator cuff repair. Arthroscopy 2012;28:52-58. Park MC, Tibone JE, ElAttrache NS, et al. Part II: Biomechanical assessment for a footprint-restoring transosseous-equivalent rotator cuff repair technique compared with a double-row repair technique. J Shoulder Elbow Surg 2007;16:469-476. 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. Tashjian RZ, Levanthal E, Spenciner DB, Green A, Fleming BC. Initial fixation strength of massive rotator cuff tears: In vitro comparison of single-row suture anchor and transosseous tunnel constructs. Arthroscopy 2007;23:710-716. Strauss E, Frank D, Kubiak E, Kummer F, Rokito A. The effect of the angle of suture anchor insertion on fixation failure at the tendon-suture interface after rotator cuff repair: Deadman’s angle revisited. Arthroscopy 2009;25:597-602.
Purpose: To evaluate the biomechanical and design characteristics of all-suture anchors. Methods: All-suture anchors were tested in fresh porcine cortical bone and biphasic polyurethane foam blocks by cyclic loading (10-100 N for 200 cycles), followed by destructive testing parallel to the insertion axis at 12.5 mm/s. Endpoints included ultimate failure load, displacement at 100 and 200 cycles, stiffness, and failure mode. Anchors tested included JuggerKnot (1.4, 1.5, and 2.8), Iconix (1, 2, and 3), Y-knot (1.3, 1.8, and 2.8), Q-Fix (1.8 and 2.8), and Draw Tight (1.8 and 3.2). Results: The mean ultimate failure strength of the triple-loaded anchors (564 42 N) was significantly greater than the mean ultimate failure strength of the double-loaded anchors (465 33 N) (P ¼ .017), and the double-loaded anchors were stronger than the single-loaded anchors (256 35 N) (P < .0001). No difference was found between the results in porcine bone and biphasic polyurethane foam. None of these anchors demonstrated 5 mm or 10 mm of displacement during cyclic loading. The Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix (P ¼ .025) but not the Iconix and Draw Tight (P > .05). The most common failure mode varied and was suture breaking for the Q-Fix (97%), JuggerKnot (81%), and Iconix anchors (58%), anchor pullout with the Draw Tight (76%), whereas the Y-Knot was 50% suture breaking and 50% anchor pullout. Conclusions: The ultimate failure load of an all-suture anchor is correlated directly with its number of sutures. With cyclic loading, the Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix but not the Iconix and Draw Tight. JuggerKnot (81%) and Q-Fix (97%) anchors failed by suture breaking, whereas the Draw Tight anchor failed by anchor pullout (76%). Clinical Relevance: All-suture anchors vary in strength and performance, and these factors may influence clinical success. Biphasic polyurethane foam is a validated model for suture anchor testing.
T
he importance of suture anchors to assist in the attachment of tendons, ligaments, and other soft tissue to bone is readily apparent, and these devices are used widely for most minimally invasive techniques, especially in the upper extremity. Over time, suture anchors have been improved with fully threaded designs, distally placed eyelets, the ability to
From Plano Orthopedic Sports Medicine and Spine Center (F.A.B.), Plano; and Advanced Surgical Institutes, Medical City Dallas Hospital (M.A.H.), Dallas, Texas, U.S.A. The authors report the following potential conflicts of interest or sources of funding: F.A.B. and M.A.H. receive support from Stryker Endoscopy, Smith & Nephew Endoscopy, ConMed-Linvatec, Biomet Sports Medicine, and Cayenne Medical (funding for Instron costs from the anchor makers). This study was funded by support from the F. Alan Barber, M.D., F.A.C.S., Research Foundation, and the donation of anchors by Biomet Sports Medicine, Stryker Endoscopy, ConMed Linvatec, Smith & Nephew, and Parcus Medical. Received June 6, 2016; accepted September 26, 2016. Address correspondence to F. Alan Barber, M.D., Plano Orthopedic Sports Medicine and Spine Center, 5228 West Plano Parkway, Plano, TX 75093, U.S.A. Ó 2016 by the Arthroscopy Association of North America. Published by Elsevier. All rights reserved. 0749-8063/16518/$36.00 http://dx.doi.org/10.1016/j.arthro.2016.09.031
accommodate multiple sutures, the use of first biodegradable and later biocomposite materials, and knotless designs. Suture anchors made completely of suture material are a recent development.1,2 These all-suture anchors are based on one of more ultra-highmolecular-weight polyethylene (UHMWPE)-containing sutures. The anchor portion of the device typically consists of a sleeve or tape also made from suture material through which the UHMWPE containing suture is woven. When the all-suture anchor is inserted into bone and the main suture pulled, the sleeve or tape is cinched up to compress against the overlying cortical bone creating a “ball,” which serves as the anchor. As with conventional anchors, these all-suture anchors are configured for glenoid-labral and tuberosity-tendon applications. The glenoid anchors are smaller and usually have 1 or at most 2 sutures.1,3 The rotator cuff anchors are larger and are either double or triple loaded. All-suture anchors are radiolucent, nondegradable, and concern has been raised about the development of cyst formation at the anchor site.4 An understanding of the failure mode and strength characteristics of these all-suture anchors is necessary for the surgeon considering their use. The purpose of this
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol
-,
No
-
(Month), 2016: pp 1-9
1
2
F. A. BARBER AND M. A. HERBERT
study was to evaluate the biomechanical and design characteristics of all-suture anchors. Our hypothesis was that these anchors do not have time zero failure profiles or biomechanical properties that are materially different from conventional suture anchors. A secondary hypothesis was that the all-suture anchor performance could be reliably assessed in either porcine bone or in biphasic polyurethane foam blocks.
Methods All-suture anchors of each type to be evaluated were obtained from their respective manufacturer for testing and implanted into 2 different substrates. The goal was to have 10 samples of each anchor type and size for placement in the diaphyseal cortex of fresh adult porcine femurs obtained from a local abattoir.1,3 Each anchor tested was rotated among at least 5 different femurs, and the insertion sites also were rotated among different positions on the diaphysis to minimize the potential effect of variations in bone thickness. No more than 2 anchors of each type were placed in any one femur. Later, as the literature began to demonstrate the validity of testing implants in polyurethane foam, whenever possible, an additional 10 samples of each anchor type and size were placed in biphasic polyurethane foam blocks (Pacific Research Laboratories, Vashon, WA). The biphasic polyurethane foam material was selected to replicate the environment of the cortical bone in the shoulder. The greater tuberosity and glenoid may offer different cortical thickness, especially if abrasion of the anchor insertion site is done in the greater tuberosity. For an all-suture anchor to deploy, however, it does require some cortex against which to be pulled and compressed. A robust biphasic “cortex” was selected to eliminate this feature as a point of failure in this test system. This biphasic material consisted of a solid rigid polyurethane foam block with a density of 12 pcf with a fiber filled epoxy coating similar in density, hardness and strength to cortical bone laminated on top. Appropriately sized drill holes were placed through the dense simulated cortex to allow access by the anchor into the polyurethane foam beneath. Every all-suture anchor tested was based on UHMWPE-containing suture. The all-suture anchors were separated in either the porcine bone or biphasic polyurethane foam block by at least 1 cm from any adjacent anchor to prevent crack propagation between drill holes during testing. All anchors were inserted according to the manufacturer’s instructions with the appropriate instruments by the senior author (F.A.B.) or by a company product manager or engineer if available to assure adequate familiarity with the anchor and instrumentation. Both anchor insertion and pullout testing were conducted with the bones or biphasic
polyurethane foam blocks at room temperature in a non-aqueous environment. The biomechanical testing was performed (M.A.H.) by securing the femurs or foam blocks holding the anchors to a platform directly under the actuator arm of a mechanical materials testing machine (model 3345; Instron, Canton, MA). The femurs were placed in a specially prepared aluminum box that supported the bone and automatically aligned the anchors’ sutures directly under the actuator arm of the mechanical materials testing machine. The blocks were secured on the sides by clamps secured to the base plate. Thus, the load applied was always in line with the axis of anchor insertion. The sutures were secured in the upper hydraulic fixture with a constant gauge length. The upper arm of the Instron machine was positioned so that there was no load on the device and then under program control, a preload of 10 N was applied. After the preload, a cyclic load alternating between 10 N and 100 N was applied at 0.5 Hz for 200 cycles or until failure occurred. A gauge length of 30 mm was used for all tests to start and the gauge length after the initial cycle considered the baseline. After completion of 200 cycles, destructive testing was performed at a displacement rate of 12.5 mm/s. Data sampling of load and displacement was obtained at 100 samples per second. The number of cycles needed to reach both 5 mm and 10 mm of displacement was recorded if it occurred.1,3,5,6 Endpoints of this study included anchor dimensions and characteristics, the ultimate load at failure, displacement at 100 and 200 cycles (initial displacement was calculated at 10 cycles), stiffness, and mode of failure (M.A.H.) (anchor pull out, eyelet/suture cut out, or suture breakage). The anchors tested varied in size, suture material and number, and characteristics, which are listed in Table 1. All companies producing anchors were invited to provide their anchors for this test. Those who were willing donated an appropriate number of anchors along with the associated insertion equipment and personnel to assist in the anchor insertion. Some anchors were limited in their supply and not available for testing in the polyurethane foam blocks. The anchors tested are listed in the paragraphs to follow. JuggerKnot 1.4 mm, 1.5 mm, and 2.8 mm all-suture anchors (Biomet Sports Medicine, Warsaw, IN) (Fig 1) are composed of 1 or 2 (blue/white) size No. 1 or 2 braided UHMWPE sutures that pass through a flexible tube of No. 6 braided polyester material measuring 2 mm wide and between 20 and 24 mm in length. The size of the polyester tube varies depending on the anchor. This sleeve bunches up under a cortical surface, creating a ball that provides the anchoring effect. The JuggerKnot 1.4 has a single No. 1 braided UHMWPE suture, whereas the JuggerKnot 1.5 has a single No. 2
3
ANALYSIS OF ALL-SUTURE ANCHORS Table 1. All-Suture Anchor Physical Characteristics Anchor JuggerKnot 1.4 JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-knot 1.3 Y-knot 1.8 Y-knot 2.8 Q-Fix 1.8 Q-Fix 2.8 Draw Tight 1.8 Draw Tight 3.2
Anchor Material Polyester Polyester Polyester Polyester Polyester Polyester UHMWPE UHMWPE UHMWPE Polyester Polyester UHMWPE UHMWPE
Suture Size #1 #2 #2 #2 #2 #2 #2 #2 #2 #2 #2 #2 #2
Suture Number Single Single Double Single Double Triple Single Double Double/triple Single Double Single Double
Drill Diameter, mm 1.4 1.5 2.9 1.4 2.3 2.3 1.3 1.8 2.8 2.0 3.0 1.8 3.2
Sleeve Length, mm 20.3 24.4 25.4 25 25 25 25 25 32 15 20 17.8 28.6
Drill Length, mm 24 24 20.8 20 20 20 21 21 23 22.3 26.8 19 19
UHMWPE, ultra-high-molecular-weight polyethylene.
braided UHMWPE suture. The JuggerKnot 2.8 is double loaded with No. 2 braided UHMWPE sutures. Iconix (1, 2, and 3) all-suture anchors (Stryker Endoscopy, San Jose, CA) (Fig 2) are provided in 3 versions, the Iconix 1, 2, and 3, reflecting the number of strands of No. 2 braided UHMWPE suture they possess. These sutures are woven 3 times through a flat, flexible tube of braided polyester and pushed into a predrilled hole. When tensioned, the braided tube collapses into a “clover leaf” configuration, creating the anchor. Y-Knot (1.3 mm 1.8 mm, and 2.8 mm) all-suture anchors (ConMed Linvatec, Largo, FL) (Fig 3) are of different sizes, containing 1, 2, or 3 No. 2 braided UHMWPE sutures. In contrast to the JuggerKnot and Iconix anchors, the Y-Knot’s anchor portion is created from a flat braided UHMWPE tube threaded with the No. 2 UHMWPE suture. When pulled securely against
Fig 1. The JuggerKnot anchors shown are, from left to right, 2.9, 1.5, and 1.4. The associated No. 1 or 2 ultra-highmolecular-weight polyethylene (UHMWPE) sutures pass through a 2-mm wide flexible No. 6 braided polyester tube. This tube bunches into a ball, creating the subcortical anchor. The JuggerKnot 1.4 has one No. 1 UHMWPE suture, the JuggerKnot 1.5 has one No. 2 UHMWPE suture, and the 2.8 has two No. 2 UHMWPE sutures. (Ó 2016 F. Alan Barber. All Rights Reserved.)
the overlying cortical bone, this device creates a ball of UHMWPE material. The Y-Knot 2.8 is designed for rotator cuff applications and is provided as a doubleloaded or triple-loaded anchor. The Q-Fix (1.8 mm and 2.8 mm) all-suture anchor (Smith & Nephew, Andover, MA) (Fig 4) is based on a single (the 1.8 version) or double (the 2.8 version) No. 2 braided UHMWPE suture that is woven through a braided polyester tape. As with the others, the ball of material serves as the anchor when pulled against cortical bone. The 1.8-mm Q-Fix is designed for glenoid and acetabular rim applications, whereas the 2.8-mm Q-Fix is designed for rotator cuff repairs.
Fig 2. The Iconix anchors shown include, from left to right, the single-loaded Iconix 1 compressed into the “clover leaf” pattern, the Iconix 1 uncompressed, the double-loaded Iconix 2, and the triple-loaded Iconix 3. At the bottom is a double-loaded Iconix 2 demonstrating how the sutures are woven 3 times through a flat, flexible braided polyester tube. (Ó 2016 F. Alan Barber. All Rights Reserved.)
4
F. A. BARBER AND M. A. HERBERT
Fig 3. The Y-knot anchors shown are, from left to right, the triple-loaded Y-Knot 2.8 demonstrating how the sutures are woven through the flat braided ultra-high-molecularweight polyethylene tube, the single-threaded Y-knot 1.3, dual-threaded Y-knot 1.8, and again the triple-loaded Y-knot 2.8 on its inserter. At the bottom is a dual-threaded Y-knot 1.8 compressed, demonstrating the compressed configuration, which as a “W” appearance. (Ó 2016 F. Alan Barber. All Rights Reserved.)
anchor. The mean ultimate failure strength of the triple loaded anchors (564N 42; 95% confidence interval [CI] 509.5-618.8) was significantly greater than the mean ultimate failure strength of the double-loaded anchors (465 33; 95% CI 417.5-512.2) (P ¼ .017), and the double-loaded anchors were stronger than the single-loaded anchors (256 35; 95% CI 219.6-291.8) (P < .0001). Because the triple-loaded anchors are larger and designed for rotator cuff applications, it follows that the glenoid (smaller and single-loaded) anchors have lower mean ultimate failure strengths than the rotator cuff anchors. A comparison of the data generated in the porcine cortical bone and biphasic polyurethane foam blocks demonstrated no difference in the means between ultimate failure loads for the individual anchors but much narrower ranges and standard deviations. The displacement of the constructs was recorded during cyclic loading. The mean cyclic displacement for each anchor is reported in Table 3. Displacement of 5 mm is considered a significant threshold in a biomechanical test.5,6 None of these anchors reached either 5-mm or 10-mm displacement. The Y-Knot demonstrated greater
The Draw Tight (1.8 mm and 3.2 mm) device (Parcus Medical, Sarasota, FL) (Fig 5) is based on a single or double No. 2 braided UHMWPE suture. In contrast to the other all-suture based anchors, the Draw Tight has a small polyether ether ketone tip through which the suture and the sleeve pass, which facilitates insertion. The sleeve through which the suture is woven is braided UHMWPE. In the single-suture version the polyether ether ketone tip measures 1.9 mm 4.9 mm, whereas in the double-suture version it measures 3.0 mm 6.8 mm. Statistical Analysis Statistical analysis was carried out with SAS software (SAS Institute, Cary, NC). Data were read directly from the Excel spreadsheets (Microsoft, Redmond, WA) with the use of dynamic data exchange and then analyzed with SAS. Test data were analyzed for means and standard deviations. A Student t-test was used to determine differences between groups. An analysis of variance was used to test for effects from bone and failure mode differences before combining the data. Statistical significance was set at P < .05.
Results The ultimate failure loads for the tested anchors are reported in Table 2 for the porcine bone and when the anchors were available the biphasic polyurethane foam blocks. Anchor ultimate load at failure directly correlated to the number of sutures associated with the
Fig 4. The Q-Fix anchors shown are, from left to right, the single-loaded Q-Fix 1.8, the double-loaded Q-Fix 2.8, and the double-loaded Q-Fix 2.8 compressed into a ball. All sutures are No. 2 braided ultra-high-molecular-weight polyethylene which is woven through a braided polyester tape. Note how with compression the anchor width expands markedly to create a subcortical anchor. (Ó 2016 F. Alan Barber. All Rights Reserved.)
5
ANALYSIS OF ALL-SUTURE ANCHORS
The failure mode for these anchor tests is reported (Table 4). The attempt to test the Iconix 3 in biphasic polyurethane foam was unsuccessful because the biphasic resin fragmented and delaminated during cyclic loading, and only 3 tests survived the initial cycle. It should be noted that the JuggerKnot (81%) and Q-Fix (97%) anchors failed predominantly by the suture breaking whereas the Draw Tight anchor failed by anchor pullout (76%). A majority of the Iconix anchors failed by suture breaking, and the Y-Knot was evenly distributed in failure mode between suture breaking and anchor pullout.
Discussion
Fig 5. The Draw Tight anchors tested included, from left to right, the Draw Tight 1.8 and Draw Tight 3.2. Both are shown double threaded with No. 2 braided ultra-high-molecularweight polyethylene suture. The Draw Tight anchor has a polyether ether ketone tip through which the suture and the sleeve pass which facilitates insertion. (Ó 2016 F. Alan Barber. All Rights Reserved.)
displacement than the JuggerKnot and Q-Fix (P ¼ .025) but not the Iconix and Draw Tight (P > .05). None of the other anchors demonstrated a statistical difference in displacement. The stiffness was not statistically different for the different anchor groups.
For the all-suture anchors, ultimate load at failure correlated directly with the number of UHMWPEcontaining sutures in the anchor. Triple-loaded anchors (564 42 N) were stronger than double-loaded anchors (465 33 N) (P ¼ .017), which in turn were stronger than single-loaded anchors (256 35 N) (P < .0001). Triple (and possibly double)-loaded all-suture anchors are larger and generally designed for rotator cuff repair whereas the smaller, single (and possibly double)-loaded anchors generally are designed for glenoid applications. None of the all-suture anchors reached a clinically significant 5-mm displacement during cyclic loading. Individual anchor performances were compared, and the Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix (P ¼ .025) but not the Iconix and Draw Tight (P > .05). The anchor dimensions and characteristics are listed in Table 1. What is notable about the all-suture anchors is the greater length and longer associated drills needed for insertion when compared with conventional suture anchors.1,3
Table 2. Data From Load to Failure Testing in Both Porcine Bone and When Available Polyurethane Foam Is Presented Porcine Cortical Bone Anchor JuggerKnot 1.4 JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-knot 1.3 Y-knot 1.8 Y-knot 2.8 Q-Fix 1.8 Q-Fix 2.8 Draw Tight 1.8 Draw Tight 3.2
N 10 14 14 11 12 12 12 10 20 19 20 20 20 19
Failure Load, N 239.1 15.1 (215-263) 290.5 15.3 (263-325) 519.3 51.4 (455-620) 208.7 69 (121-298) 468.3 570.3 249.7 477.4 602.9 346.1 495.1 290.3 418.5
45.2 (412-548) 47 (510-654) 53.5 (161-302) 78.7 (265-590) 159 (288-765) 92 (285-602) 87.9 (288-585) 127.7 (171-514) 111.1 (134-548)
NOTE. Ranges are in parentheses.
Displacement, mm 0.22 0.22 0.22 0.31 0.23 0.20 0.45 0.33 0.55 0.19 0.23 0.30 0.30
Polyurethane Foam Stiffness, N/mm 198 98 57 3.3 76 8.3 65.2 2.1 82.9 88.9 64.8 74.1 84.1 55 56.8 41.5 49.4
5.7 4.7 4.2 5.6 3.7 3.4 6.8 5.6 10.5
Failure Load, N
Displacement, mm
Stiffness, N/mm
234.7 73.6 (115-351)
0.23
72.5 6.1
519.9 50.8 (434-584)
0.43
86.0 4.9
0.23 0.19 0.11 0.16 0.24 0.20
75.4 81.8 55 56.8 41.5 49.4
151.8 531.1 657.1 291.8 494.7 262.7 190.8
12.6 (135-174) 33.3 (485-590) 128.1 (324-766) 17.6 (258-316) 106 (213-578) 91.5 (148-401) 98.8 (134-451)
7.5 2.5 3.4 6.8 5.6 10.5
6
F. A. BARBER AND M. A. HERBERT
Table 3. Displacement During Cyclic Loading After 100 Cycles and 200 Cycles Anchor JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-Knot 1.3 Y-Knot 1.8 Y-knot 2.9 Draw Tight 1.8 Draw Tight 3.2 Q-Fix 1.8 Q-Fix 2.8
Displacement 100 Cycles, mm 1.17 0.46 1.22 0.47 1.56 0.37 1.32 0.36 1.24 0.33 1.95 0.79 1.67 0.51 2.97 1.55 1.82 0.36 2.32 0.8 1.03 0.24 1.35 0.31
Tests 14 14 11 12 12 19 20 19 20 19 20 19
Displacement 200 Cycles, mm 1.39 0.57 1.44 0.57 1.87 0.48 1.55 0.39 1.44 0.35 2.40 0.96 2.0 0.6 3.52 1.79 2.12 0.42 2.62 0.65 1.22 0.27 1.58 0.34
Failure mode was evaluated, and major differences were found in the common modes of failure for these all-suture anchors. Both JuggerKnot (81%) and Q-Fix (97%) anchors failed predominantly by the suture breaking, whereas the Draw Tight anchor failed by anchor pullout (76%). This finding may have clinical significance. It is accepted that suture breaking or the suture pulling out of the anchor is clinically better than having the anchor itself pull out of the bone.3 This is not to minimize the fact that even a loose suture in the joint can be destructive.7 With an all-suture based anchor, anchor pullout can result in a suture-ball within the joint, which can cause unappreciated damage because these are not readily identified on routine radiographs. It should be pointed out that this destructive potential also exists with polyether ether ketone, biodegradable, and biocomposite anchors. This polyurethane foam model was biphasic with a dense polymer resin to simulate cortical bone against which the all-suture anchor was secured. The polyurethane foam modeled the human cortical bone, providing a test substrate that demonstrated consistent performance with low standard deviations. Differences exist between the anchor insertion sites in the glenoid and greater tuberosity. Patient age plays a role, because Table 4. Mode of Failure for the Anchors Tested Anchor JuggerKnot 1.4 JuggerKnot 1.5 JuggerKnot 2.9 Iconix 1 Iconix 2 Iconix 3 Y-Knot 1.3 Y-Knot 1.8 Y-knot 2.9 Q-Fix 1.8 Q-Fix 2.8 Draw Tight 1.8 Draw Tight 3.2
Tests 20 14 14 23 22 12 19 20 19 20 19 18 19
Anchor Pullout 9
Anchor Break
3 1
9 11
14 7 8 1 12 16
1 2
Suture Break 11 14 14 11 21 1 5 13 11 20 18 5 1
Change, mm 0.22 0.22 0.31 0.23 0.20 0.45 0.33 0.55 0.30 0.33 0.19 0.23
Displaced, 5 mm/10 mm 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0 0/0
typically glenoid anchors are used to treat instability, which occurs in younger individuals, whereas greater tuberosity anchors are used to treat rotator cuff tears occurring in older individuals. Consequently, the glenoid bone has a more robust cortex ideally suited to activate an all-suture anchor, whereas the thinner greater tuberosity cortex may be compromised by surface abrading designed to promote healing. The biphasic polyurethane foam model was selected to simulate a femoral diaphyseal test environment that would efficiently activate the all-suture anchor deployment mechanism. The desire was to remove the substrate as a mode of failure and simulate the biologic bone model against which the foam data were compared. Polyurethane foam has been used many times for this type of testing.8-12 The use of a polyurethane foam block has been validated for tibial boneepatellar tendonebone graft testing.13 In that study, the human cadaver tibia demonstrated far more variability than the foam block and did not secure the bone plugs as effectively as the polyurethane foam altering the mode of failure. Its use for suture anchors has also been documented12,14,15 as well as testing the effects of suture anchor orientation.16 Polyurethane foam is much less expensive than a biologic specimen, readily available, with no infection risk, and no disposal challenges. These all-suture anchors typically have a 1- to 3-mm wide sleeve of suture material (either UHMPWE or braided polyester) through which the base suture is passed. The Iconix anchor and the Y-Knot have their base sutures woven though the sleeve, whereas the JuggerKnot anchors have the suture passed down the sleeve lumen. This influences the shape of the anchor when it is tightened (Figs 2-4). Displacement of 5 mm or more often is considered a significant threshold in a biomechanical test.5,6 None of the anchors tested demonstrated displacement of 5 mm or more during cyclic loading. The minimum load required for a rotator cuff repair has not been established clinically and probably can
ANALYSIS OF ALL-SUTURE ANCHORS
vary with the bone and rotator cuff tendon quality; however, a load of 250 N has been identified by Mazzocca et al.17 and others5,18 as the load that is required for the rehabilitation phase of rotator cuff tendon repair. The anchors designed for rotator cuff applications successfully exceeded this level. Specifically, the JuggerKnot 2.9 (519 N), Iconix 3 (570 N), Y-knot 2.8 (602N), Q-Fix 2.8 (495N), and Draw Tight 3.2 (418 N) all had mean failure loads above the 250 N level. The biologic bone model used was not in metaphyseal bone as might be found at the greater tuberosity. Another distinctive feature of these all-suture anchors is the drill length. As noted in Table 1, the drill lengths range from 19 to 27 mm. This drill length is longer than most conventional suture anchors.1,3 Although probably not a significant concern for rotator cuff applications, a long drill can penetrate the superior glenoid cortex associated with a SLAP repair, reaching the suprascapular nerve, or the inferior glenoid, reaching the axillary nerve.19 The most common conventional suture anchor design is a fully threaded screw-in anchor designed to achieve its greatest purchase in the cortical and subcortical bone. The biomechanical problems associated placing anchors deeper than the cortex are well known.20,21 The drill hole for the all-suture anchors are not filled with rigid polymer or metal, thus creating the potential for these holes to act as stress risers increasing the risk of subsequent glenoid fracture.4 This risk is difficult to calculate, although there are case reports of glenoid fractures associated with suture anchors.22 The biomechanical superiority of suture anchors in comparison with bone tunnels was established in the 1990s,5,6,23,24 and one of the most significant modes of failure is the suture cutting through the bone with cyclic loading.16 The increased abrasiveness of some of the current UHMWPE-containing sutures and tapes is well known,25-27 and with an anchor located in a subcortical position the pivot point of the suture loading may result in some cutting through of the overlying cortex.16 The force required to adequately “set the anchor” manually was not determined in this study. Manually inserting the all-suture anchor and pulling back on the suture to “set the anchor” did not result in significant anchor displacement during cyclic loading. With the exception of the 2.9 Y-Knot anchor, none of the anchors approached the 5-mm displacement level (Table 3). The Y-Knot and Draw tight anchors did have some anchors with an absolute displacement of 3 mm, but this should not be considered clinically applicable.3,5,6,28,29 What was measured was the mean displacement change and that was less than 0.5 mm for almost all of these anchors (Table 3). Pfeiffer et al.4 performed a biomechanical and histological study in a canine glenoid rim model and compared the JuggerKnot 1.4 with a conventional
7
2.4-mm fully threaded anchor. At 8 weeks, all JuggerKnot 1.4 anchor sites became cyst-like cavities (mean of 6.3 mm in diameter) with dense lamellar bone rims in stark contrast to the conventional 2.4-mm fully threaded anchor sites which contained intact anchors (mean of 2.7 mm in diameter) closely approximated to trabeculae of the adjacent lamellar bone.4 Biomechanical testing demonstrated significantly greater displacement at ultimate failure load for the JuggerKnot anchors (13.7 6.6 mm) compared with the conventional anchors (3.2 0.5 mm) (P ¼ .0009) as well as greater mean cyclic displacement (2.9 mm compared with 1.3 mm, respectively; P ¼ .0001). Mazzocca et al.2 compared these 2 anchors biomechanically using human cadaver glenoids simulating tears of the anterior-inferior and posterior-inferior labrum. Although similar ultimate failure loads were demonstrated (JuggerKnot 146 N and the conventional anchor 172 N) for these anchors, the solid anchor required significantly higher loads to achieve 2 mm of labral displacement (JuggerKnot 39.2 N and the conventional anchor 84.1 N; P < .001). Dwyer et al.12 expanded on the model of Mazzocca et al. and compared the 1.3 Y-Knot with a conventional 3.1-mm screw-in anchor. These were compared in human cadaver glenoids, bovine bone, and polyurethane foam. The Y-Knot also was either hand tensioned or pretensioned by a machine to 60 N. In the human cadaveric glenoid, after cyclic loading the ultimate failure loads for the 1.3 Y-Knot (91 N if hand set or 145 N if pretensioned) were comparable with the conventional anchor (107 N). These load levels were significantly greater in bovine bone and in polyurethane foam, with no difference being demonstrated between the pretensioned 1.3 Y-Knot and the conventional anchor. Chiang et al.30 performed a biomechanical analysis of a biceps tenodesis in a human humerus comparing techniques using either a 1.3 Y-Knot all-suture anchor or an 8 23-mm biocomposite interference screw fixed 15 mm proximal to the inferior border of the pectoralis major tendon insertion. Interestingly the 1.3 Y-Knot demonstrated a mean failure load of 239.1 N which compares favorably with the mean failure load for the 1.3 Y-knot of 249 N in our study. The biocomposite interference screw failure load of 254.4 N was not statistically different (P ¼ .87). Limitations Study limitations include the fact that this was a nonarthroscopically performed test, done in nonaqueous conditions at room temperature with the use of porcine bone or biphasic polyurethane foam. Because of the limited device availability, not all of the anchors tested in porcine bone also were tested in biphasic polyurethane foam. Specifically, the
8
F. A. BARBER AND M. A. HERBERT
Juggerknot anchors were not available for testing in foam and the large Iconix 3 anchors broke the foam substrate; however, all the data available were reported to consolidate the data in a single manuscript and avoid future issues with duplicate data publication. A power analysis was not performed because the number of anchors required to achieve a power of 0.8 comparing this many anchors and this many different sizes would be prohibitively large. This is a bench test and the data cannot be applied directly to a clinical setting. Clinical performance cannot be extrapolated from this information. Testing protocols that use little or no cyclic loading may demonstrate different data. The study was focused solely on all-suture based anchors, and other anchor types were not included. The testing was not carried out in age appropriate human glenoid or human greater tuberosity bone. The angle of load application was in line with the insertion angle to be consistent with past anchor tests and to attempt to produce a worst-case scenario. It does not replicate the “transosseous equivalent” loading.31,32 The weakest link in any rotator cuff tendon-suture-anchor construct is the tendon-suture interface,33-35 and the UHMWPEcontaining sutures increase the difference between the tissue strength and suture strength making tissue failure more likely if healing has not occurred.
Conclusions The ultimate failure load of an all-suture anchor is correlated directly with its number of sutures. With cyclic loading, the Y-Knot demonstrated greater displacement than the JuggerKnot and Q-Fix but not the Iconix and Draw Tight. JuggerKnot (81%) and QFix (97%) anchors failed by suture breaking whereas the Draw Tight anchor failed by anchor pullout (76%).
Acknowledgment The authors wish to acknowledge the help of Kathy Berry and Eric Sanders with specimen preparation.
References 1. Barber FA, Herbert MA, Hapa O, et al. Biomechanical analysis of pullout strengths of rotator cuff and glenoid anchors: 2011 update. Arthroscopy 2011;27:895-905. 2. Mazzocca AD, Chowaniec D, Cote MP, et al. Biomechanical evaluation of classic solid and novel all-soft suture anchors for glenoid labral repair. Arthroscopy 2012;28: 642-648. 3. Barber FA, Herbert MA. Cyclic loading biomechanical analysis of the pullout strengths of rotator cuff and glenoid anchors: 2013 update. Arthroscopy 2013;29:832-844. 4. Pfeiffer FM, Smith MJ, Cook JL, Kuroki K. The histologic and biomechanical response of two commercially available small glenoid anchors for use in labral repairs. J Shoulder Elbow Surg 2014;23:1156-1161.
5. Burkhart SS, Diaz Pagan JL, Wirth MA, Athanasiou KA. Cyclic loading of anchor-based rotator cuff repairs: confirmation of the tension overload phenomenon and comparison of suture anchor fixation with transosseous fixation. Arthroscopy 1997;13:720-724. 6. Burkhart SS, Johnson TC, Wirth MA, Athanasiou KA. Cyclic loading of transosseous rotator cuff repairs: Tension overload as a possible cause of failure. Arthroscopy 1997;13:172-176. 7. Barber FA. Biodegradable shoulder anchors have unique modes of failure. Arthroscopy 2007;23:316-320. 8. Aziz MS, Tsuji MR, Nicayenzi B, et al. Biomechanical measurements of stopping and stripping torques during screw insertion in five types of human and artificial humeri. Proc Inst Mech Eng H 2014;228:446-455. 9. Bhimji S, Meneghini RM. Micromotion of cementless tibial baseplates under physiological loading conditions. J Arthroplasty 2012;27:648-654. 10. Zdero R, Elfallah K, Olsen M, Schemitsch EH. Cortical screw purchase in synthetic and human femurs. J Biomech Eng 2009;131:094503. 11. Nicayenzi B, Crookshank M, Olsen M, et al. Biomechanical measurements of cortical screw stripping torque in human versus artificial femurs. Proc Inst Mech Eng H 2012;226:645-651. 12. Dwyer T, Willett TL, Dold AP, et al. Maximum load to failure and tensile displacement of an all-suture glenoid anchor compared with a screw-in glenoid anchor. Knee Surg Sports Traumatol Arthrosc 2016;24:357-364. 13. Barber FA. Pullout strength of bone-patellar tendon-bone allograft bone plugs: A comparison of cadaver tibia and rigid polyurethane foam. Arthroscopy 2013;29:1546-1551. 14. Poukalova M, Yakacki CM, Guldberg RE, et al. Pullout strength of suture anchors: Effect of mechanical properties of trabecular bone. J Biomech 2010;43:1138-1145. 15. Yakacki CM, Griffis J, Poukalova M, Gall K. Bearing area: A new indication for suture anchor pullout strength? J Orthop Res 2009;27:1048-1054. 16. Bardana DD, Burks RT, West JR, Greis PE. The effect of suture anchor design and orientation on suture abrasion: An in vitro study. Arthroscopy 2003;19:274-281. 17. Mazzocca AD, Millett PJ, Guanche CA, Santangelo SA, Arciero RA. Arthroscopic single-row versus double-row suture anchor rotator cuff repair. Am J Sports Med 2005;33:1861-1868. 18. Schneeberger AG, von Roll A, Kalberer F, Jacob HA, Gerber C. Mechanical strength of arthroscopic rotator cuff repair techniques: An in vitro study. J Bone Joint Surg Am 2002;84:2152-2160. 19. Morgan RT, Henn RF 3rd, Paryavi E, Dreese J. Injury to the suprascapular nerve during superior labrum anterior and posterior repair: Is a rotator interval portal safer than an anterosuperior portal? Arthroscopy 2014;30: 1418-1423. 20. Mahar AT, Tucker BS, Upasani VV, Oka RS, Pedowitz RA. Increasing the insertion depth of suture anchors for rotator cuff repair does not improve biomechanical stability. J Shoulder Elbow Surg 2005;14:626-630. 21. 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.
ANALYSIS OF ALL-SUTURE ANCHORS 22. Fritsch BA, Arciero RA, Taylor DC. Glenoid rim fracture after anchor repair: A report of 4 cases. Am J Sports Med 2010;38:1682-1686. 23. Barber FA, Herbert MA, Click JN. Internal fixation strength of suture anchorsdUpdate 1997. Arthroscopy 1997;13:355-362. 24. Sherman SL, Copeland ME, Milles JL, Flood DA, Pfeiffer FM. Biomechanical evaluation of suture anchor versus transosseous tunnel quadriceps tendon repair techniques. Arthroscopy 2016;32:1117-1124. 25. Deranlot J, Maurel N, Diop A, et al. Abrasive properties of braided polyblend sutures in cuff tendon repair: An in vitro biomechanical study exploring regular and tape sutures. Arthroscopy 2014;30:1569-1573. 26. Lambrechts M, Nazari B, Dini A, et al. Comparison of the cheese-wiring effects among three sutures used in rotator cuff repair. Int J Shoulder Surg 2014;8:81-85. 27. Williams JF, Patel SS, Baker DK, et al. Abrasiveness of high-strength sutures used in rotator cuff surgery: are they all the same? J Shoulder Elbow Surg 2016;25: 142-148. 28. Barber FA, Hapa O, Bynum JA. Comparative testing by cyclic loading of rotator cuff suture anchors containing multiple high-strength sutures. Arthroscopy 2010;26: S134-S141 (suppl 9). 29. Coons DA, Barber FA, Herbert MA. Triple-loaded singleanchor stitch configurations: An analysis of cyclically
30.
31.
32.
33.
34.
35.
9
loaded suture-tendon interface security. Arthroscopy 2006;22:1154-1158. Chiang FL, Hong CK, Chang CH, et al. Biomechanical comparison of all-suture anchor fixation and interference screw technique for subpectoral biceps tenodesis. Arthroscopy 2016;32:1247-1252. Dierckman BD, Goldstein JL, Hammond KE, Karas SG. A biomechanical analysis of point of failure during lateralrow tensioning in transosseous-equivalent rotator cuff repair. Arthroscopy 2012;28:52-58. Park MC, Tibone JE, ElAttrache NS, et al. Part II: Biomechanical assessment for a footprint-restoring transosseous-equivalent rotator cuff repair technique compared with a double-row repair technique. J Shoulder Elbow Surg 2007;16:469-476. 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. Tashjian RZ, Levanthal E, Spenciner DB, Green A, Fleming BC. Initial fixation strength of massive rotator cuff tears: In vitro comparison of single-row suture anchor and transosseous tunnel constructs. Arthroscopy 2007;23:710-716. Strauss E, Frank D, Kubiak E, Kummer F, Rokito A. The effect of the angle of suture anchor insertion on fixation failure at the tendon-suture interface after rotator cuff repair: Deadman’s angle revisited. Arthroscopy 2009;25:597-602.