J Shoulder Elbow Surg (2014) -, 1-9

www.elsevier.com/locate/ymse

Biomechanical comparison of two techniques for arthroscopic suprapectoral biceps tenodesis: interference screw versus implant-free intraosseous tendon fixation Nels Sampatacos, MDa, Mark H. Getelman, MDa, Heath B. Henninger, PhDb,c,* a

Southern California Orthopaedics Institute, Van Nuys, CA, USA Department of Orthopaedics, ‘‘H. K. Dunn’’ Orthopaedic Research Laboratory c Department of Bioengineering, University of Utah, Salt Lake City, UT, USA b

Background: A novel arthroscopic technique allows for intraosseous tendon placement in biceps tenodesis using bone tunnels and suture while avoiding the expense of an implant. No biomechanical characterization exists for this construct. Methods: Tensile tests were used to compare a suture-only biceps tenodesis technique (arthroscopic biceps intraosseous tenodesis [ABIT]) with interference screws in 7 pairs of cadaveric shoulders. The ABIT used a modified finger-trap suture method to secure the tendon to itself through an intraosseous bone tunnel. Interference screw placement followed the manufacturer’s protocol for implantation. An open technique was used to provide consistency during laboratory preparation. Results: During cyclic loading, the screws were significantly stiffer (P ¼ .040) but dissipated more energy (P ¼ .002). During failure loading, suture-only specimens showed significantly greater failure loads (P < .001) and deformation (P ¼ .046). The failure mechanism for the ABIT method was tendon elongation with progressive tensioning and slippage of the tendon through the suture mass. No complete tendon failure occurred for the ABIT. Gross tendon failure occurred in all interference screw tests at the bonetendon-screw interface. No screw or suture failed in any biceps tendon test. Conclusion: The ABIT construct showed significantly higher failure loads and deformation compared with interference screws. The comparable stiffness after cycling of both constructs suggests that micromotion at the bone-tendon interface is similar, whichdin addition to the intraosseous fixationdmay be important in promoting healing. The ABIT construct was found to absorb and restore more energy (hysteresis), suggesting potential for greater tendon preservation, which may translate into improved construct longevity. The suture-only method can eliminate the expense of an implant. Level of evidence: Basic Science Study, Biomechanics. Ó 2014 Journal of Shoulder and Elbow Surgery Board of Trustees. Keywords: Biceps tenodesis; interference screw; soft suture anchor; biomechanics

Institutional review board approval: exempt via University of Utah IRB No. 11755, ‘‘Biomechanical Testing of Orthopaedic Devices Using Descendent Tissue Models.’’ *Reprint requests: Heath B. Henninger, PhD, University of Utah Orthopaedic Center, 590 Wakara Way, Salt Lake City, UT 84108, USA. E-mail address: [email protected] (H.B. Henninger).

Tenodesis of the long head of the biceps (LHB) is a common treatment for the management of LHB tears, complex biceps-labral injuries, and biceps instability.8 Surgical decision making regarding biceps tenodeses includes the method of visualization (open vs arthroscopic),13,14,35,36

1058-2746/$ - see front matter Ó 2014 Journal of Shoulder and Elbow Surgery Board of Trustees. http://dx.doi.org/10.1016/j.jse.2014.02.027

2 the fixation construct, and the location of tenodesis placement. Early biceps tenodesis required an open incision to provide adequate visualization, raising the risk of neurovascular injury9 and issues of cosmesis. These concerns led to the development of arthroscopically accessible procedures. Biceps tenodesis techniques may include intraosseous and extraosseous fixation methods such as soft tissue tenodesis, suture anchors, interference screws, cortical button fixation, and bone tunnel techniques, as well as combinations thereof.2,13,36,38 Intraosseous tendon length, the tendon-bone diametrical mismatch, and tendon micromotion within a bone tunnel may affect the healing capacity25 and biomechanical characteristics of the tendonbone interface.16 Because non-biologic constructs may fail during physiological cyclic loading,15,21,31 augmenting the procedure with intraosseous fixation may improve construct longevity. Tenodesis locations include the regions above (suprapectoral) and below (subpectoral) the insertion of the pectoralis major tendon1-4,13,15,22,35 and a transfer of the LHB to the conjoined tendon.40 The suprapectoral region may be further classified into a proximal area above or within the bicipital groove and the area distal to the groove. Within the suprapectoral region, proximal biceps tenodeses may result in localized pain directly over the bicipital groove (ie, ‘‘groove pain’’).20,24,32 Distal suprapectoral tenodeses are of concern as the cortical bone thins during the transition to the soft, cancellous bone of the metaphysis. Interference screws and push-in anchors rely on substantial bone quality for secure fixation39 and may not be ideal for this location. Interference screws have shown greater construct stiffness and ultimate load to failure in a majority of biomechanics studies,15,26,28,31,34 but clinical complications exist. The tendon can wrap around the screw during insertion, and postoperative tendon ruptures at the screw-tendon interface may result from the screw threads damaging the tendon as it is compressed against the cortical bone.18 Humeral shaft fractures related to the residual osseous defect12,17,30,33 and cyst formation related to interference screw materials6,7,11,27 have also been reported. These issues have encouraged the development of alternative fixation constructs. A widely used intraosseous technique, the classic miniopen subpectoral Casperi-Weber (CW) technique,36 uses bone tunnels without an implant and is a modification of the keyhole technique first described by Froimson and O.13 A further refinement of the CW technique has been advanced by the senior author (M.H.G.), allowing for an arthroscopic approach to the distal suprapectoral region. This technique has been coined arthroscopic biceps intraosseous tenodesis (ABIT). The ABIT incorporates a modified finger-trap suture technique, given its superior biomechanical profile compared with other suturing methods.37 Suture-based constructs typically fail at the suture-tendon interface as the

N. Sampatacos et al. sutures cut through the tendon under direct loading. The modified finger-trap involves a series of suture wraps and half-hitches around the tendon that avoid penetrating the tendon and increase the surface area of the suturetendon interface. The modified finger-trap may reduce the tendon-bone motion that occurs during cyclic loading, previously observed in a CW-equivalent bone tunnel construct.21 Additional potential advantages of the distal suprapectoral ABIT include the following: an arthroscopic approach limiting cosmetic concerns, reduced neurovascular risks compared with subpectoral approaches, avoidance of groove pain associated with proximal tenodeses, lack of interference screw–related issues (eg, tendon damage, humeral fracture, or cyst formation), use of a robust and tendon-preserving suturing technique, an intraosseous location that may promote healing, and the cost savings derived from eliminating the need for implant hardware. The purpose of this study was to characterize the biomechanical behavior of the ABIT while comparing its performance with a well-accepted interference screw fixation technique in a distal suprapectoral location. The primary hypotheses were that the cyclic and failure load/ deformation properties of the constructs would not significantly differ when tested in paired cadaveric shoulders.

Materials and methods Seven pairs of fresh-frozen human cadaveric upper extremities were thawed at room temperature for 24 hours before dissection. Each specimen was dissected down to the glenohumeral joint. The LHB attachment to the superior labrum was cut at the supraglenoid tubercle, and the humerus was disarticulated from the glenoid. All soft tissue was removed, leaving the proximal humerus, biceps tendon, and biceps muscle belly intact. Implants and proximal bone tunnels were placed in the bicipital groove 1.5 cm proximal to the superior border of the pectoralis major insertion (Fig. 1, A and B), where a mark was placed on the LHB and the bone. By use of a digital caliper, the width and thickness of the LHB were measured at the level at which the mark on the tendon had been made. A fair coin was flipped to determine which specimen in the pair would receive which construct. For ABIT specimens, the tendon was first prepared. The tendon was cut 2.5 cm proximal to the mark placed on the tendon, and the resected tissue was discarded. Two marks were then made 1.5 cm proximal and distal to the initial mark, defining the proximal and distal extent of the finger-trap suture wraps, respectively. With modification from a previous method,37 the finger-trap suture was then placed (Fig. 2) using a high-strength suture (No. 3-4 Force Fiber; Tornier, Edina, MN, USA). When the modified finger-trap stitch was placed around the 3-cm region of tendon, that segment condensed to a length of approximately 2 cm. The free ends of the suture were passed through and along the central third of the residual 1-cm tendon stub using a free needle, completing the tendon preparation. Bone preparation was then completed. A 6-mm hole was drilled into the humerus at the previously marked location in the distal suprapectoral region. A 2-mm hole was then drilled 1.5 cm

Techniques for arthroscopic biceps tenodesis

3

Figure 1 (A) Screw placed 15 mm from proximal border of pectoralis major insertion. (B) ABIT placed 15 mm from proximal border of pectoralis major insertion. (C) For testing, specimens were inverted in the Instron machine with the humeral head clamped in a fixture. Freeze clamps gripped the muscle belly, and the Instron machine applied displacement. distal to the primary hole. A suture shuttling device (No. 1 PDS; Ethicon, Somerville, NJ, USA) was passed from the distal to the proximal hole using an arthroscopic grasper in preparation for shuttling the tendon. The free ends of the suture were then shuttled from proximal to distal through the bone tunnels, using the previously placed shuttle. With 1 suture limb on either side of the tendon, the sutures were pulled, shuttling the tendon into the proximal tunnel until the distal extent of the finger-trap was at the level of the distal 2-mm bone tunnel. While tension was maintained on the sutures, the biceps tendon/muscle belly was gently pulled distally, simulating the in vivo muscle tension with the elbow in a fully extended position. This lengthened the finger-trap construct by approximately 2 mm. A Revo knot was then tied over the tendon such that the knot was laid directly on top of the distal finger-trap suture half-hitch (Fig. 3). If the tendon condensed by more than 1 cm with placement of the finger-trap suture, the Revo knot was unable to be placed on top of the finger-trap without significant tension being applied. In this case (1 construct in our study), the knot was instead placed just distal to the finger-trap suture bundle. For arthroscopic placement, a racking hitch is applied intraarticularly where the LHB exits the bicipital groove with a clamp placed on the suture adjacent to where it exits the skin to maintain the resting tendon length after transection of the insertion at the superior labrum. In addition, the arthroscopic technique requires an additional anteroinferior portal. The portal is typically created 6 to 7 cm from the anterolateral corner of the acromion and localized with a spinal needle first. This portal can be easily recreated by drawing a line between the anterior and lateral portals and then moving 6 to 7 cm perpendicular to the midpoint of that line. With creation of the anteroinferior portal as described, the tendon can typically be retrieved with appropriate length to apply the suture configuration. If the tendon is short in a muscular patient, then the arm is placed in slight forward elevation and the elbow is flexed 90 , allowing additional tendon length for suture placement. For the interference screw specimens (8  23–mm Milagro BR screw; DePuy Mitek, Warsaw, IN, USA), the tendon was cut 2.5 cm proximal to the mark on the LHB and the resected tissue was discarded. An interlocking Krackow stitch was placed in the LHB with high-strength suture (No. 2 FiberWire; Arthrex, Naples, FL, USA). Starting at the proximal tendon edge, 4 passes of an interlocking Krackow stitch were made, heading from proximal to distal, to the level of the initial tendon mark. The suture was then

passed through the tendon, and 4 more passes were made, returning to the proximal end. The 2 suture limbs were tied together with 3 alternating half-hitches, forming a loop with a diameter of 2 to 3 mm. The loop was passed over the tapered tendon fork for later seating of the tendon within the bone tunnel. Per the manufacturer’s published surgical technique, at the mark on the humerus corresponding to the initial tendon mark, an 8  26–mm bone tunnel was created, and the tendon was secured within the tunnel with an 8  23–mm screw.

Experimental protocol The protocol was adapted from a prior study.38 Each repair construct was inverted and mounted in a servohydraulic materials testing machine (Instron 1331 with model 8800 controller; Instron, Norwood, MA, USA) equipped with a 1-kN tension-compression load cell (Dynacell model 2527130; Instron) (Fig. 1, C). A thermoelectric cryoclamp secured the biceps muscle–tendon unit to the actuator and load cell, and a threaded jig stabilized the humeral head to the machine platform. The humerus and biceps tendon were aligned such that the tensile forces were applied parallel to the longitudinal axis of the humerus, thus approximating the in vivo biceps muscle/tendon force vector.15 To normalize the initial loading between specimens and constructs, displacement was applied to preload the tissue to 0.5 MPa of stress (force/cross-sectional area of native tissue), and then the displacement was held for 2 minutes while the tissue stress relaxed. Cyclic loading with a triangle waveform was then applied for 500 cycles at 1 Hz to 10% clamp-to-construct strain, which kept the tissue below theoretical microstructural failure limits of 5% to 6% tissue strain,29 where tissue strain is approximately 50% of clamp strain.5 Clamp-to-construct strain was measured as the change in length divided by the initial length, where initial length was measured from the cryoclamp to the central position of the construct insertion into bone. After a 2minute recovery period, the constructs were pulled to failure at 1 mm/s. The tissue was regularly moistened with a 0.9% saline solution spray throughout testing.

4

N. Sampatacos et al.

Figure 2 Modified tendon finger-trap for ABIT. (A) A single half-hitch is tied over the tendon at the distal mark on the tendon. (B) After the first half-hitch is secured with a hemostat on the hitch to maintain tension, each strand is passed around the tendon in opposite directions heading proximally. The strands are then tied together with a single half-hitch. (C) The surgeon continues this process heading proximally until 4 separate half-hitches have been placed. The distal hemostats are removed sequentially as proximal half-hitches are tightened. The final half-hitch should be placed over the proximal mark on the tendon. (D) Retaining the final hemostat, the surgeon passes the first leg of suture around the tendon in the same direction. (E) The first leg is wrapped around the tendon a second time. (F) The surgeon aligns the second suture leg with the tendon, laying it over the loops of the first suture leg. (G) The first leg is loosely passed around the tendon again. (H) The second leg is secured to the surface of the tendon with a half-hitch made by the first leg. (I) The half-hitch made by the first leg is tightened. (J) While cinching the proximal knot, the surgeon removes the hemostat from the last half-hitch. (K) Using a free Keith needle (Havel, Cincinnati, OH), the surgeon passes both free suture legs through the central third of the residual 1-cm tendon stub. (L) The free ends of suture are pulled through the tendon to complete the finger-trap sequence. The free ends are then shuttled through the holes in the humerus (proximal to distal) to complete the bone tunnel portion of the ABIT construct.

Time, applied force, and actuator displacement data were continuously recorded throughout testing. Structural characteristics (peak cyclic load, stiffness [in newtons per millimeter], and ultimate load) were quantified for cycles 1 and

500, as well as for the failure test. Effective material properties (stress, strain, and modulus) were determined by use of clampto-clamp strain to ensure that the tissue stayed below failure limits, assuming that tissue strain was 50% of applied clamp

Techniques for arthroscopic biceps tenodesis

Figure 3

ABIT construct placed in proximal humerus.

strain. Hysteresis, or energy dissipation, was defined as the area between loading and unloading stress-strain curves. During failure testing, yield was defined as the point on the stress-strain curve that departed from a linear response, signifying the start of non-recoverable plastic deformation. Failure was defined as any point that exhibited an abrupt decrease in load support during the test, where ultimate failure occurred at the peak deformation corresponding to peak load before significant loss of construct integrity.

Data analysis A pair-wise experimental design was used to reduce error due to tissue variability and bone density between individual specimens. An a priori power analysis showed that 7 specimens per group would provide 80% power to detect a significant difference in mean construct stiffness (30 N to 90 N) between groups with an effect size of 0.6 and significance level of P  .05.38 All statistical comparisons were made with paired t tests, and P  .05 was considered significant. Post hoc power calculations were carried out for all detected significant differences using a 2-tailed matched pair-wise comparison, with effect size calculated from the respective means and standard deviations. Power calculations were carried out with G*Power 3.1.10

5 an interference screw. There were no significant differences in any of the specimen preparation metrics (Table I). Peak cyclic load tended to be higher for interference screws, but the difference was not significant (Table II). The screw was nearly twice as stiff as the ABIT (P ¼ .040). When normalized to tissue cross-sectional area and construct length, the stress and moduli were not significantly different between constructs, but again, the screw tended to have higher values. Peak hysteresis was significantly higher for the interference screw constructs (P ¼ .041). After 500 cycles (Table III), peak cyclic load decreased significantly for both constructs, but the difference between constructs did not differ. The interference screw softened considerably more than the ABIT, where the change in stiffness (in newtons per millimeter) was triple that of the ABIT. The changes in stress and moduli were not significantly different between constructs, but the screw tended to decrease more. Hysteresis decreased more for the interference screw constructs compared with the ABIT (P ¼ .002), resulting in less energy storage capacity of the screw construct. Qualitatively, hysteresis rapidly decreased from cycle 1 to cycle 10 and then progressively decreased up to cycle 500 (Fig. 4). The ABIT failed at loads nearly 2.5 times those of the interference screw (P ¼ .0001), whereas the screw maintained higher structural stiffness (P ¼ .043) (Table IV). The ABIT construct tended to have slightly higher yield stress (not significant), but the yield strain was significantly higher for the ABIT (P ¼ .044). The ABIT had significantly higher ultimate failure stress and strain (P  .046), whereas the interference screw showed a higher ultimate failure modulus (P ¼ .031). The mechanism of failure differed between constructs. All interference screw specimens failed the same way, with tearing of the tendon at the bone-screw-tendon interface. Of the 7 specimens, 3 exhibited a first-stage failure with initial gross tissue tearing, followed by a higher ultimate failure. The other 4 specimens had a single ultimate failure resulting in complete disruption of the tendon. The ABIT construct consistently exhibited a multi-stage failure mechanism. The initial failure typically involved suture elongation with the Revo knot ratcheting to the next adjacent half-hitch. Ultimate failure was typically due to a combination of tendon elongation, partial slippage of the tendon through the finger-trap, and partial tearing of the tendon. None of the ABIT specimens failed by complete tendon disruption, and all tendons maintained their close proximity to the bone. No anchor or screw hardware failed in any test. Given the inconsistent first failure in interference screw specimens, no statistical comparison was possible between constructs for that metric.

Results

Discussion

Specimens were obtained from 7 donors (6 male donors and 1 female donor) aged 65  6 years. Four right shoulders received the ABIT construct, and 3 right shoulders received

This study compared the mechanical properties of a bone tunnel/suture construct (ABIT) with a well-accepted interference screw technique in distal suprapectoral biceps

6

N. Sampatacos et al. Table I

Specimen preparation data

Construct

Tendon cross-sectional area

Pre-tension force at 0.5 MPa

Clamp-to-construct length

ABIT (mean  SD) Screw (mean  SD) P value

15.4  3.7 mm2 14.6  4.8 mm2 .693

7.7  1.7 N 7.6  1.9 N .853

29.9  2.0 mm 29.0  1.5 mm .219

Table II

Cyclic test data for cycle 1

Construct

Peak load

Peak stiffness

Peak stress

Peak modulus

Peak hysteresis

ABIT (mean  SD) Screw (mean  SD) P value (between constructs) Power (between constructs)

67.9  17.6 N 87.6  35.6 N .329 d

33.5  6.7 N/mm 63.5  27.3 N/mm .040 0.59

4.6  1.5 MPa 6.6  2.8 MPa .171 d

74.3  27.5 MPa 112.5  49.4 MPa .094 d

63.1%  4.7% 70.8%  5.0% .041 0.57

Table III

Cyclic test data for change between cycle 1 and cycle 500

Construct

Change in load

Change in stiffness

ABIT (mean  SD) Screw (mean  SD) P value (between constructs) Power (between constructs)

44.2  14.4 N 13.9  6.1 N/mm 63.6  23.4 N 42.0  16.9 N/mm .197 .008 d 0.89

Figure 4 Representative data showing loading and unloading curves of cycles 1, 10, 100, and 500 of a single ABIT construct. Hysteresis is calculated as the area between the loading and unloading curves.

tenodesis. During cyclic loading, the screw was initially stiffer but softened over 3 times as much as the ABIT over 500 cycles of loading. Similarly, the change in hysteresis was higher for the screw. During failure loading, ABIT specimens showed significantly greater failure loads and deformation whereas the screw maintained a higher stiffness. This is the first study to demonstrate a bone tunnel construct with a significantly higher ultimate failure load compared with an interference screw (2.5 times higher, P < .001). Interference screws failed with complete tendon tearing at the bone-tendon-screw interface, whereas ABIT

Change in stress

Change in modulus Change in hysteresis

3.0  1.1 MPa 34.2  16.2 MPa 40.2%  6.5% 4.8  2.0 MPa 69.3  44.7 MPa 64.9%  8.0% .119 .086 .002 d d 0.99

constructs preserved tendon integrity with multimodal failure via suture elongation and partial tendon slippage. This has a potentially significant clinical implication. A ‘‘Popeye’’ deformity will not develop with the suture elongation mechanism, whereas complete tendon rupture with inference fixation will be at high risk of showing clinical deformity. The ABIT technique offers several advantages over prior methods. First, as an arthroscopic approach, it reduces cosmetic concerns related to open surgery. Second, the suprapectoral tenodesis generally avoids the neurovascular risks associated with the subpectoral region.9 Placement of fixation in the distal groove is associated with a lower rate of postoperative groove pain compared with tenodesis in a more proximal location.20,32 Third, the overall elongation of the construct under loading versus interference screws could provide a mechanism to limit or avoid the incidence of tendon rupture and subsequent Popeye deformity. Finally, at a time when fiscal responsibility plays a greater role in surgical decision making, the cost savings by eliminating the implant is a significant advantage particularly given the comparable or superior performance of the construct. The intraosseous placement of the LHB tendon may also be advantageous in promoting healing and subsequently improving construct strength. Using a canine model, Greis et al16 showed that the length and fit of tendon in a bone tunnel were associated with superior failure properties at 6 weeks postoperatively. In areas in which the tendon-bone contact was intimate, histologic analyses showed wellorganized collagen fibers that arose from newly woven

Techniques for arthroscopic biceps tenodesis Table IV Construct

7

Failure test data Failure load

Failure stiffness

347.7  64.9 N 32.0  6.5 N/mm ABIT (mean  SD) Screw 142.5  48.9 N 50.0  17.5 N/mm (mean  SD) P value

Biomechanical comparison of two techniques for arthroscopic suprapectoral biceps tenodesis: interference screw versus implant-free intraosseous tendon fixation.

A novel arthroscopic technique allows for intraosseous tendon placement in biceps tenodesis using bone tunnels and suture while avoiding the expense o...
1MB Sizes 1 Downloads 4 Views