Abstract
Background:
Graft loosening/rupture or iatrogenic patellar fractures after medial patellofemoral ligament reconstruction (MPFLR) can result in continued anterior knee pain and recurrent patellar instability that may necessitate revision surgery. Interference screw fixation at the patellar-graft interface has been associated with higher rates of iatrogenic patellar fractures than suture anchor fixation.
Purpose:
To biomechanically compare interference screw versus suture anchor fixation of a semitendinosus graft to the patella.
Study Design:
Controlled laboratory study.
Methods:
Eight pairs of human cadaveric patellae (16 total specimens) were harvested along with their corresponding semitendinosus autografts. Matched pair specimens were assigned to 2 groups: 8 specimens underwent graft fixation to the patella using 2 parallel 3-mm BioComposite SutureTak suture anchors, and 8 specimens underwent graft fixation to the patella using 2 parallel 3.9-mm BioComposite Swivelock interference screws. Constructs underwent biomechanical testing to compare displacement after cycling, stiffness, ultimate load to failure, and mode of failure.
Results:
There was no significant difference in mean displacement after cycling between the interference screw (5.44 ± 1.22 mm) and suture anchor groups (6 ± 1.64 mm) (P = .47). The interference screw had a significantly lower mean ultimate failure load (153.4 ± 81.5 N) than the suture anchor group (281.56 ± 81.5 N) (P = .008). There was no significant difference in mean stiffness between the suture anchor (48.22 ± 20.75 N/mm) and interference screw groups (52.18 ± 13.24 N/mm) (P = .67).
Conclusion:
Suture anchor fixation (3 mm) of a semitendinosus graft to the patella failed at significantly higher ultimate tensile loads than interference screw fixation (3.9 mm). There was no difference in displacement or stiffness after cyclic loading. This information should be helpful to surgeons seeking to avoid drilling tunnels in the patella.
Clinical Relevance:
Suture anchor fixation at the patella-graft interface in MPFLR provides a biomechanically stronger construct with less patella violation compared with interference screws. This may make it the preferred option for MPFL graft fixation on the patella.
Patellar dislocation is a common knee injury sustained in young and active patients, with an estimated incidence of 43 per 100,000, and >70% of these injuries occur during sports. 7 Traditionally, nonoperative intervention was considered the standard for patellar dislocations. Previous studies have shown that nonoperative management has long-term consequences, such as recurrent dislocations, which have been reported to be as high 9 as 71%, and an increased risk of patellofemoral arthritis. 3
The medial patellofemoral ligament (MPFL) is the primary static stabilizer of the patella in the first 30° of knee flexion, and it is damaged in as many as 94% to 100% of patellar dislocations.1,10 Thus, reconstruction of the MPFL has become a widely accepted tool in treating patellar instability. Additional procedures, such as a tibial tubercle osteotomy, are commonly performed either alone or in addition to an MPFL reconstruction (MPFLR) for treatment of recurrent lateral patellar instability, especially if there is an increased tibial tubercle-trochlear groove distance or trochlear dysplasia. 15 Erickson et al 3 demonstrated that isolated MPFLR is effective, with a low instability rate, regardless of these bony pathologies at early 1- and 2-year follow-up.
One common method of MPFLR is the utilization of a hamstring tendon autograft for double-bundle reconstruction with fixation using anchors, interference screws, or transosseous fixation on the insertion at the superomedial aspect of the patella and an interference screw in the femoral origin between the adductor tubercle and the medial epicondyle.6,11 Previous studies 5 have reported the overall incidence of recurrent patellar instability after MPFLR to be between 1.2% and 8.3%. The most commonly cited reasons for failure of MFPLR have been patellar fracture, graft rupture, elongation, or other factors not previously addressed, such as trochlear dysplasia or valgus limb alignment. One of the theoretical benefits of suture anchor fixation over interference screw or transosseous fixation in the patella is the reduction in the risk of significant complications such as patellar fractures.14,15 Although patellar fractures have been described as occurring rarely in suture anchor fixation cases in the literature, these suture anchors were significantly larger in diameter (4.75 mm) compared with modern suture anchors. 2 This is consistent with recent studies that showed nearly all patellar fractures are associated with tunnels ≥4.5 mm. 5
A direct comparison of the biomechanics of suture anchor versus interference screw fixation at the graft attachment site to the patella in MPFLR would help determine whether modern diameter suture anchor fixation techniques are biomechanically similar to interference screw fixation. Therefore, we biomechanically compared the patellar graft fixation site of MPFLRs using semitendinosus autograft fixed using either interference screws or suture anchors. We measured displacement after cyclic loading, stiffness, and ultimate load to failure. Our null hypothesis was that there would be no significant difference between fixation methods for any of these parameters. Finally, we recorded the modes of failure of the fixation techniques.
Methods
Specimen Preparation
An exception to the requirement for human subjects review was obtained from the Institutional Review Board at the University at Buffalo. Eight matched pairs of fresh-frozen cadaveric patellae were harvested from fresh-frozen cadaveric specimens, along with their native semitendinosus tendons. The cadaveric specimens were frozen for 48 hours, then thawed for 1 cycle, and testing was conducted on a single day. The paired patellae were randomized into groups of fixation method, such that each graft fixation method was performed on 1 patella from the same cadaveric specimen. Eight specimens underwent MPFLR patellar fixation using 2 parallel 3-mm BioComposite SutureTak (Arthrex) suture anchors, and 8 specimens underwent MPFLR patellar fixation performed with 2 parallel 3.9-mm BioComposite Swivelock (Arthrex) interference screws. Suture anchors or graft tunnels were placed approximately 12 mm apart. A grasping suture technique for suture anchor attachment was used, with each strand having grasped two-thirds of the tendon, with an overlap of one-third of the tendon diameter (Figure 1).
Schematic of the suturing technique used in reconstruction with suture anchors, demonstrating each strand grasping two-thirds of the tendon, with an overlap of one-third of the tendon diameter.
Specimen Testing
All testing was performed by a single author (S.M.). The bone tunnels were standardized by drilling both tunnels parallel to each other in the same coronal plane, 5 mm from the midline of the patella, for a distance of 1 cm between the drill tunnels. Tension was then applied manually. Each graft was preconditioned through a series of 10 cycles, each consisting of 5 to 15 N at a frequency of 1 Hz. This was followed by 1000 cycles of loading between 10 and 50 N at a frequency of 1 Hz. The specimens were then loaded to failure at a rate of 20 mm/s. The line of pull was set to be parallel to the interference screws or suture anchors (Figure 2).
Experimental setup utilizing an MTS machine with the line of pull parallel to the patellar fixation. MTS, materials testing system.
These protocols were selected to be consistent with a previous biomechanical study by Johnston et al, 6 which tested the biomechanics of the MPFLR. Cyclical displacement (measured as peak-to-peak displacement in mm), construct stiffness (N/mm), ultimate load to failure (N), and mode of failure were recorded. The maximum and minimum displacement values corresponding to when the materials testing system was applying 50 and 10 N, respectively, were utilized to calculate the cyclical displacement by comparing the mean length of the graft after 1000 cycles of loading at 10 to 50 N at 1 Hz compared with the preconditioned length after the initial 10 cycles of 5 to 15 N at 1 Hz. Constructs were loaded at 20 mm/sec, and the ultimate load to failure (N) as well as the mode of failure were recorded. The construct stiffness (N/mm) was subsequently calculated as the slope of the linear portion of the load versus displacement curve.
Statistical Analysis
Previous studies have shown that, according to an a priori power analysis, 5 specimens are required per testing group to achieve a power of 80% and detect a significant difference (α < .05). 7 The sample size requirement was determined to be 8 for each group. 8 Data met assumptions for a student's t test, and thus, statistical analysis was performed with a paired student's t test. Data were analyzed in Excel (Microsoft Corp). The level of significance was set at P < .05.
Results
The mean gauge length of the graft specimens was 79 ± 7.5 mm (63.5-82.6 mm). There was no significant difference in mean displacement after cycling between the interference screw (5.44 ± 1.22 mm) and suture anchor groups (6 ± 1.64 mm; P = .47). Construct displacement at 50 N from cycle 1 to 1000 was found to be 3.97 ± 0.75 mm in the suture anchor group and 3.96 ± 0.77 mm in the interference screw group (P = .47) (Tables 1-4). The interference screw had a significantly lower mean ultimate failure load (153.4 ± 81.5 N) than the suture anchor group (281.56 ± 81.5 N) (P = .008) (Figure 3).
Initial Displacement Values
Final Displacement Values
Mean Displacement at 10 N
Mean Displacement at 50 N
Load to failure. Data are presented as mean (red box) ± standard deviation.
There was no significant difference in mean stiffness between the suture anchor (48.22 ± 20.75 N/mm) and the interference screw group (52.18 ± 13.24 N/mm) (P = .67). In the interference screw group, the most common form of failure was tendon pulling out from the tunnel (N = 7). In contrast, 1 interference screw could not be placed due to poor bone quality and large screw size. The suture anchor group more commonly failed at the graft-suture interface (N = 5) compared with the anchor pulling out of the tunnel (N = 3).
Discussion
The null hypothesis for our study was that there would be no significant difference between fixation methods for measured displacement after cyclic loading, stiffness, and ultimate load to failure using semitendinosus autograft fixed using either 3.9 mm interference screws or 3 mm suture anchors in patellar graft fixation of MPFLR. The null hypothesis was accepted regarding cyclic loading and stiffness, but rejected for ultimate load to failure, as suture anchor fixation was found to have a higher load to failure compared with interference screw fixation.
A previous study found that the native MPFL has an ultimate load to failure of up to 208 N. 4 In our study, the suture anchor group had a higher ultimate tensile load compared with the native MPFL, while the interference screw group had slightly lower ultimate tensile strength than the native MPFL, as cited in previous studies.12,13 In our study, the native MPFL was not directly compared with the fixation groups, and thus, different experimental conditions and specimen differences may exist, which is a potential drawback. Our findings suggest, however, that the suture anchor reconstruction technique is sufficient to recreate the ultimate failure load of the MPFL, whereas the interference screw method may not be sufficient. Two other biomechanical studies examined patellar graft fixation in MPFLR, comparing interference screw fixation and suture anchor biomechanics, although these studies both fail to report graft dimensions or gauge length.12,13 Raoulis et al 12 reported the ultimate tensile load of the suture anchor group at 243 ± 41.9 N and the interference screw group at 263.2 ± 9.6 N, while Russ et al 13 reported the ultimate tensile load of the suture anchor group at 201.54 ± 63.14 N and the interference screw group at 299.25 ± 99.87 N. One likely explanation for the difference between our findings and those of these studies is that Raoulis et al 12 used two 6-mm interference screws. In contrast, Russ et al 13 used two 4.5-mm interference screws, as opposed to the two 3.9-mm interference screws utilized in this study. This suggests that increasing the interference screw size may increase the ultimate tensile load; however, increasing the size of the tunnels may also increase patellar fracture rates. 5 Furthermore, the larger tunnels created are undesirable in the event of a revision surgery. Both fixation techniques demonstrated similar stiffness during cyclic loading.
Previous studies evaluating the biomechanics of suture anchor and interference screw fixation did not specifically report displacement during cyclic loading.12,13 Displacement is noteworthy to study in MPFLR because significant displacement may indicate that the graft will no longer function as a checkrein for the patella during the knee range of motion, thus leading to patellar subluxation/dislocation. In our study, both fixation techniques demonstrated similar displacement during cyclic loading with no difference found in displacement at 50 N from cycle 1 to 1000 (suture anchor group = 3.97 ± 0.75 mm; interference screw group = 3.96 ± 0.77 mm (P = .47)). It is unknown, however, whether this finding is of clinical significance and whether this maximum displacement would cause any clinical instability. Future studies may investigate the minimum displacement required to elicit clinical symptoms.
The most common failure method in the suture anchor group was failure at the graft-suture interface. In contrast, the interference screw group’s most common failure method was tendon pullout through the tunnel. The technique of suture anchor failure at the graft-suture interface was similar to findings by Russ et al. 13 Additionally, the interference screw fixation method of failure was comparable to previous studies.12,13
Patellar fixation techniques without hardware typically involve either transosseous or violating the anterior patellar cortex, which has been shown to have an increased rate of patellar fracture. 4 Patellar fracture has also been shown to be more likely with larger drill bit size, with smaller sizes conveying less risk of fracture. 2 While some studies may indicate that smaller bone tunnels may decrease the risk of patellar fracture, our study also suggests that suture anchors may be a better alternative. In a retrospective study with a 2-year follow-up, Yoon et al 16 demonstrated similar clinical outcomes between suture anchor groups and transosseous fixation, suggesting that suture anchors in the patella should be a reasonable option for surgeons seeking strong fixation and wishing to avoid patellar fracture.
Limitations
This study has several limitations. First, it is a cadaveric biomechanical study; therefore, the testing represents the immediate postoperative period and does not reflect any healing or longevity of the implant durability during follow-up. Second, the bone quality of the cadaveric specimens is significantly lower than that of any live patient, especially given that patellar instability is more prevalent in younger populations. Additionally, the specimens were frozen before our study testing, which may have further affected tissue quality. Third, it was conducted with the uniaxial line of pull horizontally, testing the MPFL, while the kinematics of the patellofemoral joint are more complex and likely differ in vivo. Because our study aligns directly with the anchors or screws, it should represent a worst-case scenario. Fourth, we did not compare the fixation techniques to the native MPFL strength of the specimens being tested. Lastly, we had a limited number of specimens, which may limit the generalizability of our study and likely caused increased variability and a greater magnitude of difference between the means reported in the present study. This may have also prevented us from capturing all smaller differences, such as those related to displacement or stiffness.
Conclusion
Suture anchor fixation (3 mm) of a semitendinosus graft to the patella failed at significantly higher ultimate tensile loads than interference screw fixation (3.9 mm). There was no difference in displacement or stiffness after cyclic loading. This information should be helpful to surgeons seeking to avoid drilling tunnels in the patella.
Footnotes
Final revision submitted July 8, 2025; accepted September 8, 2025.
One or more of the authors has declared the following potential conflict of interest or source of funding: Materials for this study were donated by Arthrex (grant No. IIRR-01282). L.J.B. has received hospitality payments from Arthrex. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval was not sought for the present study.
