If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To compare various suture anchor designs with and without calcium phosphate (CaP) augmentation in an osteoporotic foam block model and decorticated proximal humerus cadaveric model.
This was a controlled biomechanical study, consisting of 2 parts: (1) an osteoporotic foam block model (0.12 g/cc; n = 42) and (2) a matched pair cadaveric humeral model (n = 24). Suture anchors selected were an all-suture anchor, PEEK (polyether ether ketone)-threaded anchor, and a biocomposite-threaded anchor. For each study arm, one half the samples were first filled with injectable CaP and the other half were not augmented with CaP. For the cadaveric portion, the PEEK- and biocomposite-threaded anchors were assessed. Biomechanical testing consisted of a stepwise, increasing load protocol for a total of 40 cycles, followed by ramp to failure.
For the foam block model, the average load to failure for anchors with CaP was significantly greater when compared with anchor fixation augmented without CaP; the all-suture anchor was 135.2 ± 20.2 N versus 83.3 ± 10.3 N (P = .0006); PEEK was 131 ± 34.3 N versus 58.5 ± 16.8 N (P = .001); and biocomposite was 182.2 ± 64.2 N versus 80.8 ± 17.4 N (P = .004). For the cadaveric model, the average load to failure for anchors augmented with CaP was again greater than anchor fixation without CaP; PEEK anchors went from 41.1 ± 21.1 N to 193.6 ± 63.9 N (P = .0034) and biocomposite anchors went from 70.9 ± 26.6 N to 143.2 ± 28.9 N (P = .004).
Augmenting various suture anchors with CaP has shown to significantly increase pull-out strength and stiffness in an osteoporotic foam block and time zero cadaveric bone model.
Rotator cuff tears are common in the elderly patients, in whom poor bone quality jeopardizes treatment success. Exploring methods that increase the strength of fixation in osteoporotic bone to improve outcomes in this patient population is important.
Rotator cuff injuries are one of the most common causes of shoulder disability and are associated with pain and decreased function.
Unfortunately, the development of rotator cuff pathology is considered to be a natural consequence of the aging process, where rotator cuff tears in older individuals are usually atraumatic, resulting from chronic degeneration.
For operative treatment, arthroscopic cuff repair is commonly used to treat painful shoulders caused by rotator cuff tears after patients are unresponsive to conservative management, with good results reported in both short- and long-term studies.
Despite satisfying outcomes of different SA systems, rotator cuff repair in elderly patients presents several unique challenges as the result of increased medical comorbidities, impaired soft-tissue healing, greater tuberosity remodeling, enthesopathy with cystic change, and greater incidence of osteopenia.
Moreover, there are certain risk factors that can lead to more frequent anchor pullout scenarios, including poor bone quality in the greater tuberosity as seen in osteoporosis, bone cysts, previous fractures, or previous failed rotator cuff repair.
demonstrated frequent failure at the bone–anchor interface and that minimal force was needed for proximal anchor migration in osteoporotic bone. Bone cysts also can affect failure rates at the bone–anchor interface since SA fixation depends on bone stock and bone quality.
These risk factors present a difficult situation for establishing adequate fixation of SA devices and commonly present themselves as soft tissue or bony fixation failures, often leading to significant effects on patient outcomes such as persistent pain and low functional scores.
Given the high incidence of rotator cuff tears in elderly patients and increased risk factors for anchor pullout, anchor fixation must be able to account for and mitigate against complications related to osteoporotic bone.
Consequently, various methods have been implemented to address poor bone quality, such as augmentation with cement or calcium phosphate (CaP), curettage with bone grafting, improved implant design (diameter, thread pitch, material, length, etc.), adjustment in number of anchors, and implantation of collagen.
However, there are currently limited data on whether similar augmentation can enhance the performance of an all-SA.
Tactoset (Anika Therapeutics, Bedford, MA) is a synthetic, biocompatible, hyaluronic acid-enhanced CaP bone graft substitute material intended for filling bone voids or defects of the skeletal system that are not intrinsic to the stability of bony structure.
This material has been indicated to augment hardware during the surgical process as an injectable, self-setting, osteoconductive bone graft substitute that hardens postdeployment to reinforce weaknesses and mimic the properties of trabecular bone. The senior author has had clinical experience using Tactoset in rotator cuff repair; however, to date this unique, proprietary, injectable CaP material has not been biomechanically tested for its ability to improve SA pullout strength.
The purpose of this study was to compare various SA designs with and without CaP augmentation in an osteoporotic foam block model and decorticated proximal humerus cadaveric model. It was hypothesized that all anchor types would have improved biomechanical properties seen as increased pullout strength when augmented with CaP and would have equivalent performance.
Foam block testing
A total of 42 rigid polyurethane foam blocks (7.5 PCF, SKU#1522-09; Pacific Research Laboratories, Vashon, WA) were used as testing medium due to their similarities in osteoporotic bone biomechanical properties.
Upon receipt, all stock blocks (size: 13 × 18 × 4 cm) were cut into 6 squares (size: 5 × 5 × 4 cm) and divided among the following testing groups: group A (SA only, n = 21 foam blocks) only received the SA into foam blocks where each SA type had 7 samples, and group B (SA+ CaP, n = 21 foam blocks) received SAs inserted into foam blocks augmented with Tactoset calcium phosphate bone graft substitute (CaP) and tested at time-zero (immediate after material cures). The SAs investigated were the 3.2-mm DrawTight (soft anchor composed of ultra-high-molecular-weight polyethylene; Anika Therapeutics, Bedford, MA), the 4.5-mm Twist PEEK (threaded anchor composed of polyether ether ketone [PEEK]; Anika Therapeutics), and the 4.5-mm Twist AP (threaded anchor composed of 70% polylactic-co-glycolic acid and 30% β-tricalcium phosphate; Anika Therapeutics) (Table 1).
For all groups, the center point of each test sample relative to designated top surface was identified and marked. Samples in the control group were only instrumented with each SA type. To standardize trajectory of the SA, a 2.0-mm pilot hole was created using a drill press to a depth of 1 cm. Thereafter, a 3.2-mm awl was used to appropriately size the pilot hole. After the use of the awl, the DrawTight SA was deployed. For Twist PEEK and Twist AP anchors, a 4.5-mm tap was used before inserting the anchor.
Samples in the experimental group were warmed up to body temperature with a therapeutic heating pad before being injected with 2.5 cc of CaP and inserting appropriate SA. The CaP was mixed for 1 minute and allowed to set for 7 minutes before injecting directly into the location where the anchor would be placed. The depth of the cannula was approximately 24 mm. After implantation of CaP, a therapeutic heating pad was again used to keep the foam blocks at or above body temperature (98.6°F) for the curing process. Foam blocks with CaP were maintained at or above 98.6°F for 10 minutes (total elapsed time of 18 minutes) before the SAs were instrumented as stated previously (Fig 1).
This study group used 3 different SA types: PEEK, biocomposite, and all-suture. All 3 were designed to attach soft tissue to bone. PEEK was specifically designed for bone ingrowth into the SA.
However, the all-SA often requires to be seated to the cortical bone as much as possible at the time of insertion; therefore, it was not selected to test in the decorticated cadaveric portion described to follow.
A total of 12 matched pairs (n = 24) of cadaveric humeri were obtained from an accredited tissue bank (MedCure, Orlando, FL) for this study (4 male and 8 female, average age of 79 years). Upon receipt, the specimens were assessed for any abnormalities and stripped of all soft tissue. Specimens were then divided into the following testing groups, assuring each matched pair was in differing group. Group C (SA only, 12 humeri) received the SAs inserted into decorticated footprint, 6 humeri were instrumented with Twist AP anchor, and the remaining 6 were instrumented with Twist PEEK anchor. For Group D (SA+ CaP, 12 humeri), these samples received the SA inserted into bone augmented with CaP, 6 humeri were instrumented with Twist AP anchor, and the remaining 6 were instrumented with Twist PEEK anchor (Table 2).
All specimens were shortened to 20 cm in length as measured from the apex of the humeral head. The distal diaphysis was potted in high strength resin (Bondo Body Filler; 3M Collision Repair Solutions, St. Paul, MN). For both groups, a 1-cm × 2-cm window was outlined along the superior facet of the greater tubercle with a skin marker to serve as decortication guideline for each sample. Decortication was performed on all samples based on previous studies as worst-case scenario by using a highspeed 3-mm oval burr within outline for an average depth of 3 mm.
To standardize trajectory of the SA, the samples were clamped in a specialized vice at approximately 45° (measured by digital goniometer) such that the anchors were placed perpendicular to the surface. All anchors were placed centrally within the decorticated footprint, 5 mm from anatomical neckline and 1 cm from edge (Fig 2).
Samples in the control group were only instrumented with either the Twist AP or Twist PEEK SA. Samples in the experimental group were injected with 2.5 cc of CaP before the SA was instrumented. The injection site was placed centrally within decorticated footprint. The depth of the cannula was approximately 24 mm. Before the implantation of CaP, a therapeutic heating pad was used to bring the samples at or above body temperature (98.6°F). Samples with CaP were then maintained at or above 98.6°F with the same heating pad for 10 minutes for the curing process to complete before the SAs were instrumented. Placement of the SA was directly into the cured CaP. Control specimens were also heated to body temperature before instrumenting with anchors only to standardized any effects of heating.
Biomechanical testing was performed on the same day of instrumentation to test the immediate pull-out strength of the SA. A custom fixture was built and fixed to the base of a servo-hydraulic materials testing machine (MTS Bionix; MTS Inc., Eden Prairie, MN) equipped with a 5-kN load cell (model 661.19F-05 [MTS Inc]; repeatability at 0.03% full scale). The design allows for consistent sample placement and the degree of freedom to align the SA construct line of axis with direction of force (Fig 3). The sutures were hand tied around a 1.27-cm diameter cylinder with 5 alternating half-hitch knots, assuring equal lengths. The loop was then secured through a 6.35 mm (¼”) anchor shackle with a working load limit of 453.6 kg (1000 lbs.) and coupled to the load cell. Once the samples were properly aligned, a small preload (<0.5 N) was applied to remove the slack from the system before testing, which was confirmed through tactile and visual inspection.
Testing consisted of a step-wise and increasing load protocol.
Each specimen was tested under axial loading rate of 1 mm/s and consisted of 4 different testing conditions: (1) level 1: 10 to 50 N for 10 cycles; (2) level 2: 10 to 100 N for 10 cycles, (3) level 3: 10 to 150 N for 10 cycles, and (4) level 4: 10 to 200 N for 10 cycles. If the anchor remained intact (no failure) after cyclic loading, the repair was loaded to failure at 1 mm/s. The number of completed cycles and failure load was compared between groups, as well as failure modes. Failure was defined as the first significant decrease in the monotonically increasing force profile, where pull-out strength was defined as the peak load at failure. Peak load for each specimen was recorded whether failure occurred during or after cyclic loading. Stiffness was defined as the linear portion of the force-displacement curve.
Using mean and variance data from previous literature,
an a priori test was used to compute the required sample size to detect a difference between the 3 study groups in the sawbones portion (G∗Power; V220.127.116.11; Franz Faul, Kiel, Germany). Assuming a power of 0.8 and a type I error rate of 0.05, a total sample size of 18 (6 per group) was necessary for the current investigation. The choice of 21 samples (7 per group) powers the study greater than 0.9. Moreover, when performing the a priori test for the cadaveric portion, the use of 24 samples (12 per group) powers this portion of the study greater than 0.9.
A univariate analysis of variance was performed on demographic data (age, sex, weight, body mass index) between the 2 testing groups. A paired sample t-test was performed to identify differences in pull-out strength and stiffness between the control group and the experimental group. Moreover, a 1-way analysis of variance was performed to compare pull-out strength and stiffness across SA types. Data are presented as mean ± standard deviation. All statistical comparisons were performed with SPSS (version 22; IBM Corp., Armonk, NY) at a significance level of α = 0.05.
In the foam block model, a significant increase in both load to failure and stiffness was found across all anchor types when augmented with CaP (Fig 4). The pull-out strength of the DrawTight (P = .0006), Twist AP (P = .0036), and Twist PEEK (P = .0011) increased by roughly 2 times or greater when augmenting with CaP. The stiffness increase was more variable where the DrawTight saw a 5-time increase (P < .001), where was Twist AP (P = .008) and Twist PEEK (P = .0004) both saw 2 times increase in stiffness.
When we compared across anchor types when no CaP had been introduced, it was found that the Twist PEEK had significantly less pull-out strength when compared with both DrawTight (P =.018) and Twist AP (P =.034). No differences were found between DrawTight and Twist AP. The stiffness across the anchor types were all significantly different from each other (Fig 5). The Twist AP had the greatest stiffness and was significantly greater than both DrawTight (P = 0002) and Twist PEEK (P = 048). The DrawTight had the lowest stiffness compared with Twist PEEK (P = .049)
When we compared across anchor types with CaP, no significant differences were detected regarding the pull-out strength. No differences between stiffness were observed when comparing DrawTight and Twist AP. However, the stiffness of Twist PEEK was found to be significantly less than Twist AP (P = .032) (Fig 6). The mode of failure was anchor pull-out for all samples (Fig 7).
There were no significant differences found between study groups for age, height, weight, or body mass index (Table 3). In addition, there were no significant differences in the depth of decortication between group 1 and group 2 (group 1: 3.1 ± 0.9 mm vs group 2: 2.8 ± 0.5 mm; P = .393).
Table 3Donor Demographics
Group 1 (Twist AP)
Group 2 (Twist PEEK)
NOTE. P values corresponding to age (P = .717), height (P = .999), weight (P = .482), and BMI (P = .371).
BMI, body mass index; F, female; M, male; SD, standard deviation.
When augmenting with CaP, both anchor types had a significant increase in load to failure, where Twist AP increased 2-fold (70.9 ± 26.6 N vs 143.2 ± 28.9 N; P = .004) and Twist PEEK increased roughly 5-fold (41.1 ± 21.1 N vs 193.6 ± 63.9 N; P = .003). However, only the stiffness in Twist PEEK significantly improved (26.9 ± 27.5 N/mm vs 140.5 ± 26.9 N/mm; P = .002) with the augmentation of CaP (Fig 8). Comparing across the 2 anchor types with no CaP, no significant differences were detected regarding retention strength (Twist AP: 70.9 ± 26.6 N vs Twist PEEK: 53.0 ± 38.2 N; P = .370). However, the stiffness of Twist AP was significantly greater than Twist PEEK (Twist AP: 78.9 ± 38.0 N/mm vs Twist PEEK: 32.2 ± 30.5 N/mm; P = .045). Similarly, when comparing across the 2 anchor types with CaP, no significant differences were detected regarding pull-out strength (Twist AP: 143.2 ± 28.9 N vs Twist PEEK: 193.6 ± 63.9 N; P = .121). Interestingly, the addition of CaP improved the stiffness of the Twist PEEK and was found to be significantly greater than Twist AP (Twist AP: 103.0 ± 22.0 N/mm vs Twist PEEK: 140.5 ± 26.9 N/mm; P = .025). The mode of failure was anchor pull-out for all samples (Fig 9). Cross sections of samples in experimental group were taken to visualize the penetration of CaP and thread interaction (Fig 10). All anchors were fully embedded into CaP.
The results of this study indicate that preparing an osteoporotic foam block model with 2.5 cc of CaP before instrumentation of all SAs resulted in a significantly greater pull-out strength and increase in stiffness than SA’s without the use of CaP. Similarly, it was concluded that in decorticated cadaveric humeri augmented with the CaP, the PEEK-threaded anchors and biocomposite-threaded anchors also experienced significantly greater pull-out strength and increase in stiffness than those without CaP. These findings support our hypothesis that using CaP before instrumentation of SA would have improved biomechanical properties in both osteoporotic models tested. However, although pull-out strength was equivalent across SA types, stiffness was found to differ between SAs in the foam block and cadaveric testing when augmented with CaP. This was likely due to material properties of the anchors. Interestingly, when augmented, the soft SA was found to have equivalent biomechanical performance in the osteoporotic foam block models to hard body threaded SA constructs. This could open opportunities for the use of this particular soft anchor in cases of bone cyst or poor bone stock, by reinforcing existing cortex with CaP before anchor instrumentation, providing a denser surrounding to be properly deployed.
Bone quality is an important factor that modulates SA resistance to pullout, and osteoporosis of the greater tuberosity has been shown to be one of the many factors that affect the pullout strength negatively.
concluded that anchors placed on the posterior greater tuberosity had a greater pullout strength than those set on the anterior portion of the greater tuberosity, despite no difference in bone mineral density between the 2 locations. However, Tingart et al.
found a positive correlation between greater tuberosity bone mineral density and SA pullout strength. Although these studies highlight the conflicting data in the literature regarding SA fixation in patients with compromised bone quality, the need to improve SA fixation still exists.
To reduce bone loss associated with anchor insertion, smaller SAs such as the all-SA design have been developed.
demonstrated that the pullout strength of all-SAs was dependent on thickness of the humeral cortex, with a direct correlation between thickness and force needed to cause pullout failure of sutures.
Case reports of excellent 1-year outcomes using CaP bone void filler for rotator cuff and labral repair have been published, where follow-up imagining at 1 year indicated tendon healing, radiographic resolution of previous existing cyst, and improvement in patient clinical scores.
who evaluated SA pullout strength in cement versus noncemented cadaveric humeri. They found that augmented anchors had a significantly (P < .05) greater mean pullout strength (540 N; 95% confidence interval, 389-690 N) than the standard control group (202 N; 95% confidence interval 100-305 N). Direct comparison between our data and that of Aziz et al.
is difficult to do, as they used polymethyl methacrylate; however, the principles of augmentation increasing fixation still apply. One of the disadvantages of polymethyl methacrylate is that it is inert and will not allow for bone remodeling, whereas the use of an injectable biologically active CaP can undergo cell mediated remodeling and can be considered ideal in this use. The rotator cuff tendon heals by formation of granulation tissue at the tendon–bone interface.
performed an animal study in which they studied the use of a CaP matrix to augment rotator cuff repair and demonstrated new bone formation through histomorphometry analysis with significantly greater fibrocartilage and increased collagen organization at tendon-bone insertion site when compared to control (no CaP) at 2 weeks (P = .04). Similarly, Zhao et al.
demonstrated through histological observation the use of a CaP bioceramic can aid in cell attachment and proliferation, accelerating new bone formation and biomechanically improve ultimate load to failure and stiffness compared to control group (no CaP).
In addition, these findings support a previous study with a similar methodology but different device manufacturers and larger anchors (5.5-mm anchors).
In that study, it was concluded that pullout strength with the calcium phosphate-bone substitute material (CP-BSM) in osteoporotic foam block was significantly higher at 332.7 N ± 47.6 compared with the average pullout strength in the non-augmented foam block at 144.4 N ± 14.6 (P = .005). Similarly, it was found that cadaveric humeri with the CP-BSM had significantly greater average pullout strength than those without CP-BSM (274.1 N ± 102.1 vs 138.5 N ± 109.8; P = .029).
The forces presented across these studies have a wide range and many factors need to be considered when interpreting, such as the testing medium (bone/material density), anchor design and size, and loading protocol. The forces experienced in vivo during cuff repair rehabilitation are estimated based on supraspinatus activity during active-assisted forward elevation to be approximately 13% of maximal actitivity
evaluated how the structural design of different SAs influences pullout strength. They concluded that increased contact area between the anchor threads and surrounding bone is one of the main structural designs that increase pullout strength.
Limitations in regard to using all-SAs to treat cuff tears include difficulty of setting the anchor with thin cortical wall, managing revisions when a tear occurs, the expense of new SAs for the patient, dislodgement of the anchor from the bone, and occasional knot impingment.
Percutaneous delivery and curation of CaP may allow the surgeon to achieve better purchase of bone, resulting in increased surface area between the SA threads and the surrounding environment.
The results of this time-zero biomechanical study provide evidence based on osteoporotic foam block model and decorticated cadaveric matched-pairs model that various SA devices can achieve higher pull-out strength and increased stiffness when augmented with CaP bone graft substitute.
As with most biomechanical studies, one of the limitations is the inability to model the healing process. However, this study design elected to evaluate the performance of SAs at time-zero to recreate when anchors are most prone to pull-out. Moreover, the testing medium selected were considered osteoporotic for worst-case scenario. The foam blocks allowed for a uniform material to equally test each SA type with and without CaP augmentation. Another limitation was that bone mineral density testing was not available; however, we verified comparability between cadavers by measuring the mean depth of decortication (3 mm). This value fell between 1.7 and 5.6 mm, which is the range of decortication published by Ruder et al.
respectively. Further limitations regarding the use of cadavers are the age of the population, which averaged at 79 years, whose results may not be applicable to younger patients. Lastly, the study was designed to exclude the use of all-SA in the decortication cadaver model. While the foam block outcomes appear promising, future studies are still needed to test these conditions.
Augmenting various SAs with CaP has shown to significantly increase pull-out strength and stiffness in an osteoporotic foam block and time zero cadaveric bone model.
The authors report the following potential conflicts of interest or sources of funding: funded by Anika Therapeutics. M.A.D. reports grants from Anika Therapeutics, during the conduct of the study; and grants from Zimmer Biomet, outside the submitted work. C.E.B. reports grants from Anika Therapeutics, during the conduct of the study; personal fees from Zimmer Biomet, grants and personal fees from Anika Therapeutics, and speaking and consultation fees from Zimmer Biomet and Anika Therapeutics, outside the submitted work. Full ICMJE author disclosure forms are available for this article online, as supplementary material.
Research Performed at the Phillip Spiegel Orthopaedic Research Laboratory at the Foundation for Orthopaedic Research and Education, Tampa, Florida, U.S.A.