Cyclic Load Testing and Ultimate Failure Strength ofBiodegradable Glenoid Anchors

F. Alan Barber, M.D., David A. Coons, D.O., and Michell Ruiz-Suarez, M.D.

Purpose: The purpose of this study was to compare biodegradable glenoid suture anchors by cyclicloading and load to failure testing. Methods: Seven different suture anchors (BioKnotless and LupineLoop [DePuy-Mitek, Norwood, MA]; BioPushLock, BioSutureTak, and BioFasTak [Arthrex Corp,Naples, FL]; BioAnchor [Conmed Linvatec, Largo FL]; and BioRaptor [Smith & Nephew, Andover,MA]) were tested in 8 matched pairs of human cadaver fresh-frozen glenoids. The anchors wereinserted in rotation into different glenoid rim positions. Sutures attached to the anchors were fixed toan Instron 8871 machine (Instron, Canton, MA) and cyclic loading and destructive testing wereperformed. The cyclic displacement at 100 and 500 cycles, stiffness, ultimate failure strength, andmode of failure were determined. Results: No statistical difference was found in the ultimate failureload for any of these anchors. The Lupine Loop and BioAnchor had greater 100 cycle and 500 cyclemean displacements than the BioPushLock and BioSutureTak. The Lupine Loop also had greater 100cycle and 500 cycle mean displacement than the BioFasTak (P � .05). The BioAnchor had greatermean 500 cyclic displacement than the BioFasTak (P � .05). Mean BioSutureTak stiffness wasgreater than the Lupine Loop, BioAnchor, BioKnotless, and BioRaptor (P � .05). Conclusions: Nodifferences in ultimate failure strength after cyclic loading were found in these seven biodegradableglenoid anchors (BioKnotless, Lupine Loop, BioPushLock, BioSutureTak, BioFasTak, BioAnchor,and BioRaptor). Most displacement occurred in the first 100 cycles. Displacement at 500 cycles wasgreater for the Lupine Loop and the BioAnchor than the BioPushLock, BioSutureTak, and Bio-FasTak. Failure was principally by the anchor pulling out of bone except for the BioSutureTak, whichalso failed by the suture loop eyelet pulling out of the anchor body, and the BioPushLock which failedby the suture slipping past anchor. Clinical Relevance: Biodegradable glenoid anchors did not showstatistical difference in ultimate failure load after cyclic loading. Key Words: Arthroscopy—Bankart—Biodegradable—Biomechanics—Cyclic loading—Suture anchor.

Suture anchors are commonly used to repair shoul-der capsulolabral lesions associated with shoulder

instability and superior labrum from anterior to pos-terior (SLAP) tears1-4 and are superior to older suture

fixation methods.5 The newer biodegradable anchorshave gained considerable popularity in recent yearsand provide comparable results to metal anchors.6

Biodegradable suture anchors are easy to use, appli-cable to arthroscopic techniques, and avoid problemswith postoperative imaging and revision procedures.Biodegradable anchors loaded with high-strength su-tures are now commonly used for cuff repairs aswell.7,8 These high-strength sutures are not only stron-ger than conventional sutures but maintain their supe-rior properties even when partially cut.9 The suture–suture anchor interface is usually an eyelet in a solidbioabsorbable post or an eyelet made of another suturewhich is contained within the biodegradable anchorbody.10,11 The possibility of anchor loosening andmigration exists for any anchor. Consequently, the

From the Plano Orthopedic and Sports Medicine Center(F.A.B.), Plano, Texas; Pacific Sports Medicine (D.A.C.), Tacoma,Washington, U.S.A.; and the National Institute of Rehabilitation(M.R.-S.), Mexico City, Mexico.Supported by Arthrex, Smith & Nephew Endoscopy, Conmed

Linvatec, and DePuy-Mitek. Research was performed at Arthrex,Naples, Florida.Address correspondence and reprint requests to F. Alan Barber,

M.D., Plano Orthopedic and Sports Medicine Center, 5228 WestPlano Pkwy, Plano, TX, 75093, U.S.A.© 2008 by the Arthroscopy Association of North America0749-8063/08/2402-7280$34.00/0doi:10.1016/j.arthro.2007.08.011

224 Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 24, No 2 (February), 2008: pp 224-228

security, holding strength, and mode of failure of thesenew biodegradable glenoid anchors are important toconsider.

Because single-pull, ultimate failure loads do not re-flect the clinical condition,12,13 a more clinically relevantcomparison would involve cyclic loading such as thatwhich occurs with normal arm motion in which a modestforce is repeatedly applied to the repaired tissue. Labralrepair displacement of 5 mm after cyclic loading wouldin reality be a clinical failure.8,14,15 Tests reflecting sutureanchor performance during and after cyclic loading canprovide clinically relevant information. The purpose ofthis study was to compare biodegradable glenoid sutureanchors by cyclic loading followed by load to failuretesting. The hypothesis was that some biodegradableanchors would show greater displacement under loadthan others.

METHODS

Eight pairs of glenoids (average age, 53.6 years;range, 46 to 65) were stripped of all soft tissue. Thescapula was cut approximately 3 inches medial ofthe glenoid, three dry wall screws were inserted intothe scapula, and the scapula was potted in fiberglassepoxy resin.

Seven suture anchors were inserted in rotation onthe articular surface adjacent to the glenoid rim ofeach specimen (Fig 1). The anchors tested were theMitek BioKnotless and Mitek Lupine Loop (DePuy-Mitek, Norwood, MA), Arthrex BioPushLock, Ar-threx BioSutureTak, and Arthrex BioFasTak (ArthrexCorp, Naples, FL), the BioAnchor (Conmed Linvatec,Largo FL), and the BioRaptor (Smith & Nephew,Andover, MA). All anchors were inserted using themanufacturers’ instrumentation and recommendedtechnique. These anchors were inserted in rotation atthe 12, 1:30, 3, 4:30, 6, 7:30, and 9 o’clock positionsfor right glenoids and 12, 3, 4:30, 6, 7:30, 9, and 10:30o’clock positions for left glenoids. The anchors wereinserted with a 30° tilt relative to the surface of theglenoid as previously described,16 and far enoughapart to leave sufficient spacing between anchors toavoid crack propagation during destructive testing.

Roth et al.17 report decreased cortical layer thick-ness from 1.3 mm at the 2 o’clock position to 0.7 mmat the 5:30 position (right glenoid). Wetzler et al.18

also report increased suture anchor pullout strength inanchors placed in the superior portion of the glenoid.To decrease the effect of decreased bone quality anddensity in the inferior portion of the glenoid, all an-chors were rotated to place them an equal number of

times in all seven different insertion positions. In addi-tion, each anchor was used twice per position to decreasethe effect of variations in glenoid bone quality.

The potted scapulae were clamped to an adjustableangle fixture secured to the base of the Instron 8871machine (Instron, Canton, MA), and the fixture posi-tioned to create a pull for testing that was in line withthe angle of anchor insertion (Fig 1). A 1-kN load cellwas attached to the Instron machine.

All knotless anchors were inserted until the sutureloop was secure around a 2.08-mm diameter dowel (#45drill bit),19 and a high-strength suture (FiberWire;Arthrex, Naples, FL) passed thorough the loop createdonce the dowel was removed. The free strands of thesutures in the anchors were grasped by a pneumaticsuture clamp attached to the upward actuator arm fortesting. The clamp was positioned to create a suturegauge length of 30 mm. Preliminary testing showedno slippage of the sutures during loading.

The constructs were preloaded to 10 N at 1 N/s. Thepreload was held for 5 seconds, and then the con-structs were cycled from 10 to 60 N at 1 Hz for 500cycles. Post-cycling, those anchors that survived thecycling were subjected to a single pull to failure

FIGURE 1. Each potted scapula was clamped to an adjustableangle fixture secured to the base of the Instron, and the fixturepositioned to pull in line with the angle of anchor insertion.

225CYCLIC LOAD TESTING OF GLENOID ANCHORS

conducted at 33 mm/sec. Failure during cycling wasdefined at a displacement of 5 mm, which was felt tobe a better end-point for capsulolabral failure thansmaller displacements used associated with SLAP re-pair testing.20 The data were plotted in Origin Scien-tific Graphing and Analysis Software (OriginLabCorp, Northampton, MA) and cyclic displacement,yield load (start point at failure), ultimate load, andmode of failure were recorded. The anchor drill holesize was also recorded.

Statistics

The sample size was chosen based on a poweranalysis using expected standard deviations for gle-noid anchors.11 This sample size was sufficient toshow significant difference of 50 N given a standarddeviation of 30 N. A Kruskal–Wallis 1-way analysisof variance on ranks (� � 0.05) was performed tocompare the performance of each anchor. A pair-wisemultiple comparison procedure was conducted usingDunn’s method to compare individual anchors. P �.05 was considered significant.

RESULTS

Some anchors failed during cycling and did notreach the ultimate load to failure stage. Others weredamaged during insertion, reducing the initial numberbelow 14 for some anchors. The number of anchorsthat failed during cycling is listed in Table 1.

Most of the displacement observed during cyclicloading occurred in the first 100 cycles. The meancyclic displacement at 100 cycles and 500 cycles wasrecorded (Table 1). The mean cyclic displacement at100 cycles of the Lupine Loop anchor was significantlygreater than that of the BioPushLock, BioSutureTak, andBioFasTak (P � .05). The mean cyclic displacementat 100 cycles of the BioAnchor was significantly greaterthan that of the BioPushLock and BioSutureTak (P �

.05). All other comparisons were not significantlydifferent. None of these mean displacements reachedthe 5-mm “failure” level.

The cyclic displacement at 500 cycles of the LupineLoop anchor and the BioAnchor was significantlygreater than that of the BioPushLock, BioSutureTak,and BioFasTak (P� .05). All other comparisons werenot significantly different.

The stiffness of the Lupine Loop anchor was signifi-cantly less than that of the BioFasTak and BioPushLock(P � .05). The stiffness of the BioSutureTak wassignificantly greater than that of the Lupine Loop,BioAnchor, BioKnotless, and BioRaptor (P � .05).All other comparisons were not significantly different.No statistical difference existed between the yield loadof any anchor (P � .473).

Ultimate loads to failure were obtained after 500 cy-cles. The results of these tests are reported in Table 2. Nostatistical difference existed between the ultimate loadof any anchor (P � .251).

The modes of failure are reported in Table 3.

DISCUSSION

This comparison of biodegradable glenoid sutureanchors using cyclic loading and load to failure testingsupported the hypothesis that some biodegradable an-chors show greater displacement under load than oth-ers, but not that larger anchors would have higherloads to failure. None had clinical failure during cy-cling, which was defined as a 5-mm displacement.

Arthroscopic repairs for shoulder instability andglenoid labral tearing are commonly performed usingsuture anchors,1-4 which are less likely to cause addi-tional damage in the event of redislocation than tackdevices.21,22 In contrast to rotator cuff tendon repairs,the patients undergoing glenoid procedures are youngerand the bone and soft tissue is more robust. Immobi-lization is usually for a shorter interval, and the reha-

TABLE 1. Results of Cyclic Loading

Anchor Cycling Failures100 Cycle

Displacement (mm)500 Cycle

Displacement (mm) Stiffness (N/mm) Yield (N)

BioAnchor 6 of 12 3.7 � 1.2 4.1 � 1.1 82 � 9.0 89 � 34BioRaptor 3 of 9 3.2 � 1.5 3.4 � 1.4 98 � 21 129 � 78Lupine Loop 6 of 14 4.0 � 1.1 4.1 � 1.1 59 � 1.7 113 � 65BioKnotless 3 of 12 2.4 � 1.4 2.4 � 1.4 97 � 13 119 � 48BioSutureTak (3.7 mm) 1 of 14 1.8 � 1.4 1.9 � .14 150 � 18 127 � 39BioFasTak (3.0 mm) 3 of 14 1.9 � 1.7 2.0 � 1.6 128 � 13 124 � 45BioPushLock (3.5 mm) 2 of 14 1.8 � 1.6 2.0 � 1.6 129 � 55 120 � 51

226 F. A. BARBER ET AL.

bilitation program is typically started earlier. Thisleads to an earlier application of stress on the sutureanchor–based repair. While the suture–tissue interfaceis the greatest area of concern for most soft tissuerepairs, in glenoid capsulolabral repairs the security ofthe suture anchor–bone interface and the attachedsutures may be more important.

Clinical failure may result if a significant gap de-velops between the bone and capsulolabral tissue as aresult of cyclic loads.23 For this in vitro study, adisplacement of 5 mm or more after cyclic loadingwas considered equivalent to clinical failure.8,14 Largeranchors show greater failure loads, but may not be asappropriate in smaller bones because of concernsabout bone loss and the potential need for futurerevision surgery.12 This is especially true when con-trasting rotator cuff anchors to glenoid anchors.

In the cuff model, Burkhart has shown that as fewas 25 cycles can cause a 5-mm gap between a rotatorcuff tendon edge and the bone, and a 10-mm gap canoccur after only 188 cycles.15 In the current study,almost all of the displacement encountered for theseanchors occurred in the initial 100 cycles (Table 1). Itwas noted that the two anchors with the largest 100cycle and 500 cycle displacements (BioAnchor and

Lupine Loop) also had the greatest number of failuresduring cyclic loading (6 each), the lowest stiffness,and were also among the smallest anchors. The Bio-Anchor and Lupine Loop displacement was signifi-cantly greater than that shown by the BioPushLockand BioSutureTak (P � .05).

Variations in stiffness were also observed (Table 1).The Lupine Loop anchor showed significantly lessstiffness than either the BioFasTak or BioPushLock (P� .05). The 3.7-mm BioSutureTak was significantlystiffer than the Lupine Loop, BioAnchor, BioKnot-less, and BioRaptor anchors (P � .05).

The ultimate loads to failure of many suture anchorshave been reported.12 The mean ultimate load to failurefor the anchors tested in this study ranged from 104 to159 N, but these means were not statistically different.

All anchors tested were threaded with new high-strength sutures composed either in part or entirely ofultra high–molecular weight polyethylene. The cyclicloads were applied directly to these sutures, and no caseof suture breakage was observed. The most commonmode of failure was anchor pullout from the bone. Thetwo exceptions to this were the BioSutureTak and theBioPushLock. The BioSutureTak failed by anchor pull-out of bone and eyelet breaking with equal frequency. Itshould be noted that the eyelet in this case is a loop ofsuture imbedded within the anchor body. The eyeletfailure was with this eyelet pulling out of the anchorbody as has been previously observed.11,24 The BioPush-Lock failed predominantly by the suture slipping past theanchor and leaving the anchor body behind in the bone.Cutting of the anchor loop by the anchor during insertionwas not observed in this series.4

These biodegradable suture anchors have been usedclinically for arthroscopic Bankart procedures withgood clinical results.1,3 While the clinical results ofknotless anchors have been reported to be unsatisfac-tory compared with those of standard suture anchors,especially when considering redislocation rates,3 no

TABLE 2. Results of Ultimate Load Testing

Anchor

UltimateLoads

(N)SutureType Drill Size

BioAnchor 104 � 48 Hi-Fi 2.7 mmLupine Loop 132 � 72 Orthocord 2.9 mmBioKnotless 135 � 56 Panacryl 2.9 mmBioRaptor 141 � 80 Ultrabraid 3.0 mmBioFasTak

(3.0 mm)124 � 45 FiberWire 3.0 mm

BioSutureTak(3.7 mm)

159 � 37 FiberWire 3.25 mm prox/2.75 mmdistal

BioPushLock(3.5 mm)

144 � 57 FiberWire 3.25 mm prox/3.00 mmdistal

TABLE 3. Modes of Failure

Anchor Pullout Eyelet Break Eyelet Pullout Suture Slipped Past Anchor

BioAnchor 10 2 — —Lupine Loop 10 4 (suture loop) — —BioKnotless 8 4 (suture loop) — —BioRaptor 9 — — —BioFasTak (3.0 mm) 9 4 1 —BioSutureTak (3.7 mm) 7 7 — —BioPushLock (3.5 mm) 3 — — 11

227CYCLIC LOAD TESTING OF GLENOID ANCHORS

performance difference was observed in this study thatwould explain this observation.

Strengths of this study include the use of paired humanglenoid specimens, the number of tests performed, thedistribution of the anchors among seven different testingsites in each of two glenoids from the same individual,and the cyclic loading protocol. These specimens werefresh-frozen glenoids with an intact labrum.

The limitations of this study include the fact that it isan in vitro study and performed at room temperature ina dry (non-arthroscopic) environment.25 Variations inglenoid bone density within a single specimen do exist,as do variations between specimens. Also, a load tofailure in line with the angle of anchor insertion does notreplicate the in vivo loading of anchors used to repair thelabrum and may be a worst-case scenario. This test doesnot have direct application to the clinical setting for severalreasons. The cadaver specimens had an average age of 54years, which is greater than what would be expected clini-cally with capsulolabral pathology. An appropriate postop-erative bracing and rehabilitation program would not stressa capsulolabral repair as the 500 cycles of 60 N applied inthe line of anchor insertion did. Once healing is adequatelyadvanced, the capsulolabral loads should be no longer borneby the suture anchor–suture construct.

CONCLUSIONS

No differences in ultimate failure strength after cyclicloading were found in these seven biodegradable glenoidanchors (BioKnotless, Lupine Loop, BioPushLock, Bio-SutureTak, BioFasTak, BioAnchor, and BioRaptor).Most displacement occurred in the first 100 cycles. Dis-placement at 500 cycles was greater for the Lupine Loopand the BioAnchor than the BioPushLock, BioSuture-Tak, and BioFasTak. Failure was principally by theanchor pulling out of bone except for the BioSutureTak,which also failed by the suture loop eyelet pulling out ofthe anchor body, and the BioPushLock which failed bythe suture slipping past anchor.

REFERENCES

1. Thal R, Nofziger M, Bridges M, Kim JJ. Arthroscopic Bankartrepair using Knotless or BioKnotless suture anchors: 2- to7-year results. Arthroscopy 2007;23:367-375.

2. Hill JD, Lovejoy JF Jr, Kelly RA. Combined posterior Bankartlesion and posterior humeral avulsion of the glenohumeralligaments associated with recurrent posterior shoulder insta-bility. Arthroscopy 2007;23:327.e1-3.

3. Cho NS, Lubis AM, Ha JH, Rhee YG. Clinical results ofarthroscopic Bankart repair with knot-tying and knotless su-ture anchors. Arthroscopy 2006;22:1276-1282.

4. Hayashida K, Yoneda M, Mizuno N, Fukushima S, Nakagawa

S. Arthroscopic Bankart repair with knotless suture anchor fortraumatic anterior shoulder instability: Results of short-termfollow-up. Arthroscopy 2006;22:620-626.

5. van Oostveen DP, Schild FJ, van Haeff MJ, Saris DB. Sutureanchors are superior to transglenoid sutures in arthroscopicshoulder stabilization. Arthroscopy 2006;22:1290-1297.

6. Tan CK, Guisasola I, Machani B, Kemp G, Sinopidis C, Brown-son P, Frostick S. Arthroscopic stabilization of the shoulder: Aprospective randomized study of absorbable versus nonabsorb-able suture anchors. Arthroscopy 2006;22:716-720.

7. Cummins CA, Murrell GA. Mode of failure for rotator cuffrepair with suture anchors identified at revision surgery.J Shoulder Elbow Surg 2003;12:128-133.

8. Gartsman GM. All arthroscopic rotator cuff repairs. OrthopClin North Am 2001;32:501-510.

9. Wright PB, Budoff JE, Yeh ML, Kelm ZS, Luo ZP. Strengthof damaged suture: An in vitro study. Arthroscopy 2006;22:1270-1275.

10. Wust DM, Meyer DC, Favre P, Gerber C. Mechanical andhandling properties of braided polyblend polyethylene suturesin comparison to braided polyester and monofilament poly-dioxanone sutures. Arthroscopy 2006;22:1146-1153.

11. Barber FA, Coons DA, Ruiz-Suarez M. Cyclic load testing ofbiodegradable suture anchors containing 2 high-strength su-tures. Arthroscopy 2007;23:355-360.

12. Barber FA, Herbert MA, Coons DA, Boothby MH. Sutures andsuture anchors-update 2006. Arthroscopy 2006;22:1063-1069.

13. Hecker AT, Shea M, Hayhurst JO, Myers ER, Meeks LW,Hayes WC. Pull-out strength of suture anchors for rotator cuffand Bankart lesion repairs. Am J Sports Med 1993;21:874-879.

14. Cummins CA, Appleyard RC, Strickland S, Haen PS, Chen S,Murrell GA. Rotator cuff repair: An ex vivo analysis of sutureanchor repair techniques on initial load to failure. Arthroscopy2005;21:1236-1241.

15. Burkhart SS, Johnson TC, Wirth MA, Athanasiou KA. Cyclicloading of transosseous rotator cuff repairs: Tension overloadas a possible cause of failure. Arthroscopy 1997;13:172-176.

16. Zumstein M, Jacob HA, Schneeberger AG. In vitro compari-son of standard and knotless metal suture anchors. Arthroscopy2004;20:517-520.

17. Roth CA, Bartolozzi AR, Ciccotti MG, Wetzler MJ, GillespieMJ, Snyder-Mackler L, Santare MH. Failure properties ofsuture anchors in the glenoid and the effects of cortical thick-ness. Arthroscopy 1998;14:186-191.

18. Wetzler MJ, Bartolozzi AR, Gillespie MJ, Roth CA, CiccottiMG, Snyder-Mackler L, Santare MH. Fatigue properties ofsuture anchors in anterior shoulder reconstructions: Mitek GII.Arthroscopy 1996;12:687-693.

19. Thal R. A knotless suture anchor. Design, function, and bio-mechanical testing. Am J Sports Med 2001;29:646-649.

20. Domb BG, Ehteshami JR, Shindle MK, Gulotta L, Zoghi-Moghadam M, MacGillivray JD, et al. Biomechanical com-parison of 3 suture anchor configurations for repair of type IISLAP lesions. Arthroscopy 2007;23:135-140.

21. Mueller MB, Fredrich HH, Steinhauser E, Schreiber U, AriansA, Imhoff AB. Biomechanical evaluation of different sutureanchors for the stabilization of anterior labrum lesions. Arthros-copy 2005;21:611-619.

22. Sassmannshausen G, Sukay M, Mair SD. Broken or dislodgedpoly-L-lactic acid bioabsorbable tacks in patients after SLAPlesion surgery. Arthroscopy 2006;22:615-619.

23. Tamai K, Higashi A, Tanabe T, Hamada J. Recurrences afterthe open Bankart repair: A potential risk with use of sutureanchors. J Shoulder Elbow Surg 1999;8:37-41.

24. Barber FA. Biodegradable shoulder anchors have uniquemodes of failure. Arthroscopy 2007;23:316-320.

25. Barber FA, Feder SM, Burkhart SS, Ahrens J. The relationshipof suture anchor failure and bone density to proximal humeruslocation: A cadaveric study. Arthroscopy 1997;13:340-345.

228 F. A. BARBER ET AL.