Keywords biomechanical phenomena - joint ligaments - animal models - patellofemoral joint -
suture technique - orthopedic fixation devices
Introduction
The anatomy and biomechanical properties of the medial patellofemoral ligament (MPFL)
are described in several studies.[1 ]
[2 ]
[3 ]
[4 ]
[5 ]
[6 ]
[7 ]
[8 ] The MPFL is located in the second of the three layers of the medial region, together
with the medial collateral ligament. It runs transversally from the patella to the
femur. At the femur, it attaches itself posteriorly and proximally to the medial and
distal epicondyle and anteriorly to the adductor tubercle. Its average length ranges
from 53 to 55 mm, and its width varies from 3 to 30 mm.[4 ] Since MPFL is one of the main medial stabilizers of the patella, its injuries (most
often at the femoral attachment site) are associated with patellar dislocations.[1 ]
[2 ]
[3 ] It is frequently reconstructed in patients with recurrent instability, with good
outcomes in clinical studies.[1 ]
[3 ]
[5 ]
[7 ] The MPFL is the main limiting factor for patellar lateral displacement, direction
in which most dislocations occur, contributing with 60% of the restraining force during
lateralization in 20° flexed knee.[4 ] The MPFL reconstruction was first described by Gomes et al,[1 ] and it is performed to restore patellar stability with favorable results.[1 ]
[3 ]
[5 ]
[7 ]
The reconstruction of the MPFL is often performed alone when bone morphology is normal.
In cases with trochlear dysplasia and/or high patella, the MPFL can play an even greater
role in biomechanical restriction compared to cases with normal trochlear fossa and
patellar height.[6 ]
Many surgical techniques have been developed for the treatment of patellar instability,
including free graft fixation in bone canals, free graft fixation with anchors, and
free graft suture to the periosteum[.7 ] Today, there is no gold standard option for femoral fixation; the technique most
frequently used is the tunnel with interference screw, followed by anchors with suture.[9 ]
Fixation by tenodesis is indicated for MPFL reconstruction in immature skeletons and
to avoid transtendinous suture.[10 ] The technique of graft adductor magnus tenodesis performed in young people has the
advantage of not causing any damage to the femoral open growth plate, preventing its
premature closure, which may cause angular deformity.[11 ]
Anchors allow the direct insertion of implants (with no drilling, threading or pre-drilling)
with a self-inclusion tip as well as perfecting and tensioning individual sutures.[12 ] Good functional scores at the Kujala scale were demonstrated in patients submitted
to femoral fixation with titanium anchors. Since only one titanium anchor is applied
to the femur, this procedure is relatively cost-effective; in addition, the use of
many implants for graft stabilization increases the risk of local pain and inflammation.[11 ]
The prominent fixation material in the medial part of the medial femoral condyle is
known to cause local irritation and potentially restrict movement. Even in the absence
of prominent material, a medial femoral condyle tunnel can be the source of refractory
pain.[7 ]
[8 ]
[9 ] Pain and stiffness can also be related to an underlying lesion to the joint surfaces
of the patellofemoral compartment or to the poor tunnel positioning.[6 ]
[7 ]
[8 ]
[9 ]
It is worth noting that any synthesis material on the patellar edge or the medial
femoral condyle may become prominent following surgical swelling resolution. Patients
may be less able to tolerate discomfort in such areas, requiring material removal.[9 ]
The present work aims to test and measure the biomechanical properties of three different
methods of graft fixation in the medial femoral condyle during MPFL reconstruction
in porcine knees—interference screws, titanium anchors with sutures, and adductor
tenodesis—to evaluate the linear resistance under lateral traction at the same patellar
inclination until failure at the graft femoral site.
Materials and Methods
An experimental study was carried out with 30 fresh pigs knees aged 8 to 9 months
and with approximately 110 kg of live weight. The tests were performed at room temperature
and the samples were kept in saline solution for 300 minutes to maintain adequate
hydration of parts before testing.
Porcine knee joints were chosen because they are anatomically similar and have comparable
femoral bone density to human knee joints. Porcine knees are used as a comparative
model for femur fixation,[13 ] and previous studies reported their similar biomechanical properties.[14 ]
[15 ]
[16 ]
[17 ]
The knees were dissected evenly. After skin and subcutaneous removal, patellar and
quadriceps tenotomy, in addition to all extensor tendons, was performed, except in
group 3, and the graft was fixed through an adductor tenodesis. All peripheral structures
were sectioned and removed. Only the femur used for MPFL fixation biomechanical tests
remained.
Thirty femurs were divided into 3 groups of 10, according to the fixation types to
be tested. The reconstruction of the MPFL was performed with grafts dissected from
porcine feet extensor tendons. The medial femoral condyle lengths in the sagittal
plane were measured for sample standardization.
Considering native MPFL size variations,[18 ] we used grafts with a thickness of 4 mm and length of 126 mm (30 mm for femoral
fixation, 40 mm for fixation to the traction equipment, and 56 mm of free tendon)
for the screw and tenodesis groups, and tendons 4-mm thick and 192-mm long for the
anchor group (0 mm for femoral fixation, two 56 mm free arms, and two 40 mm arms for
fixation to the traction equipment). All grafts were measured with a mechanical caliper.
In group 1, fixation was performed with a 7 × 25 mm titanium interference screw, and
a 1.5-mm Kirschner wire was used to find the best point for the femoral tunnel. It
was placed at the anatomical site for MPFL attachment, 10 mm proximally and 2 mm posteriorly
to the medial femoral epicondyle, or 4 mm distally and 2 mm anteriorly to the femoral
adductor tubercle.[6 ]
[17 ]
A 7-mm cannulated drill was passed over the wire up to 30 mm in depth to accommodate
the free ends of the tendon and preventing the 25 mm screw of going beyond the free
end of the tendon in the tunnel to ensure better fixation.[19 ] Using a 2-mm Beath pin, the ends of the tendon were passed through the femoral tunnel,
with the stems coming out through the lateral condyle. Applying traction to the stem,
the graft was fixed with a 7 × 25-mm titanium interference screw and, as in the preconized
technique, the screw head was buried subcortically.[6 ]
In group 2, fixation was performed with 5-mm titanium anchors in the femur, at the
same point described for the previous group; the graft was folded into two arms of
equal length, 96 mm, and fixed to the anchor with Ethibond 5.0 (Ethicon Inc., Somerville,
NJ, USA) in the center of the fold over the tendon with two knots using the Pauchet
technique.[20 ] The traction exerted during testing will be at a 90° angle, optimizing the force
against the anchor pull.[21 ]
In group 3, adductor magnus tenodesis was performed; Ethibond 5.0 was used through
4 transfixations at the proximal region of the graft and 1 Pauchet knot[20 ] after each transfixation, totaling 4 knots ([Figure 1 ]).
Fig. 1 Anterior view of a left porcine femur with a perpendicular graft under traction on
the universal traction testing machine. Authors, 2017.
Biomechanical Testing
Biomechanical tests were performed for different types of graft fixation using a universal
tensile testing machine at a speed of 20 mm/minute. These tests predict the vulnerability
of a special fixation to failure during postoperative rehabilitation and provide an
environment for direct comparison of different fixation techniques and devices. The
force parameters were recorded using a Spider data acquisition system with 8 channels.
The data processing software used was Catman Easy 3.1. Both manufactured by HBM Headquarters
Germany / Darmstadt. The femurs were connected to the base of the machine by a tuning
fork and kept parallel to the ground, with the medial epicondyle facing downwards;
the bones were fixed by their diaphysis through a screw anchored to the two ends of
the tuning fork. A clip was used to fix the free graft extremity to the traction machine.
The graft was maintained in the femoral fixation region at a 90° angle in relation
to the axis of the machine ([Figure 1 ]), keeping the traction visually in a straight line with the patellar lateral displacement
vector.
The parameters evaluated were the following: force at the maximum resistance limit
for different ligament fixation types, expressed in Newtons, and failure modes. Sample
failure was defined by a sudden drop in graph curve (N) during the test. The test
was then stopped, and the graph was analyzed.
Statistical Analysis
Numerical variables were expressed as means, standard deviations, and coefficients
of variation, whereas the categorical variable was expressed as absolute frequencies.
The sample size was determined by using the comparison methods of two mean values
(Student t-test) from previous studies data.[13 ]
[17 ]
The Kruskal-Wallis test was used to compare the maximum traction force between the
groups. Data normality and variances equality hypotheses were verified using the Shapiro-Wilk
and Levene tests; size comparisons were used due to the rejection of variance equality
between the groups.[22 ]
The margin of error for statistical tests was 5.0%. Data were entered into an Excel
spreadsheet, and statistical calculations were performed with SPSS version 23 (IMB
Corp., Armonk, NY, USA).
This study was approved by the Animal Ethics Committee (CEUA, in the Portuguese acronym)
of the institution.
Results
The highest average linear resistance under lateral traction occurred in group 1 (screw
fixation; 185.45 ± 41.22 N), followed by group 2 (anchor fixation; 152.97 ± 49, 43 N);
the lower average was observed in group 3 (tenodesis fixation; 76.69 ± 18.90 N) ([Table 1 ]). According to the fixed error margin (5%), there was a statistically significant
difference between the groups (p < 0.001); in addition, multiple comparison tests (between group pairs) also showed
significant differences. Variability, expressed by the variance coefficient, was small,
lower than 33.3%. [Figures 2 ], [3 ] and [4 ] show the maximum force to failure in each sample per group
Table 1
Group
Mean (N)
Standard deviation (N)
Variation coefficient (%)
1 Screw
185.45(A)
41.22
22.23
2 Anchor
152.97(B)
49.43
32.31
3 Tenodesis
76.69(C)
18.90
24.64
P -value
p[a ] < 0.001*
Fig. 2 Maximum femoral traction force in group 1: Screw fixation. Authors, 2017.
Fig. 3 Maximum femoral traction force in group 2: Anchor fixation. Authors, 2017.
Fig. 4 Maximum femoral traction force in group 3: Tenodesis fixation. Authors, 2017.
The femoral condyle widths on the sagittal axis were homogenous in each group. There
was little variability between parts. Variation coefficients were up to 4.48%, as
shown in [Table 2 ].
Table 2
Group
Mean (mm)
Standard deviation (mm)
Variation coefficient (%)
1 Screw
69.00
3.09
4.48
2 Anchor
68.90
2.77
4.02
3 Tenodesis
68.40
1.78
2.60
Regarding the causes of failure in each technique, group 1 (screw fixation) presented
6 loosening of the graft due to its sliding in the tunnel, and 4 failures resulting
from graft rupture; in group 2 (anchor fixation), 8 failures were due to anchor suture
rupture, and 2 due to graft rupture; finally, in group 3 (tenodesis fixation), all
10 failures resulted from graft rupture ([Table 3 ]).
Table 3
Failure type
Screw
Anchor
Tenodesis
N
N
N
Graft sliding into femoral tunnel
6
–
–
Graft rupture
4
2
10
Anchor wire rupture
–
8
–
TOTAL
10
10
10
Discussion
The present study aims to compare biomechanical linear resistance from three fixation
methods previously described. Fixation with interference screws in a bone tunnel (group
1) resulted in the greatest tensile strength load (185.45 ± 41.22 N) for medial patellofemoral
ligament reconstruction, consistent with the native ligament strength in humans (145 ± 68N)
reported by Criscenti et al.[2 ] The anchor fixation group (group 2) also showed an average strength (152.97 ± 49.43 N)
until failure close to findings from the aforementioned study; however, this was not
observed in the adductor tenodesis group (group 3) (76.69 ± 18.90 N).
Our results are consistent with the published data: fixation with anchors or interference
screws results in good postoperative outcomes.[3 ]
[5 ]
[7 ]
[8 ]
The modes of failure ([Table 3 ]) resulting from maximum strength until fixation failure in each group were different
depending on the fixation method; as such, a given type of failure prevailed in each
group.
Our study showed that group 1 presented greater traction force compared to the other
groups ([Figure 2 ]), and its main type (60%) of failure was tendon sliding into the tunnel.
In the same group 1, graft rupture occurred in 40% of the samples. The simple placement
of an interference screw is known to compromise the biomechanical properties of the
graft[23 ] this effect can be attenuated by avoiding the use of excessively large screws.
In group 2, 8 samples (80%) presented failure due to anchor suture (Ethibond 5.0)
rupture; moreover, traction force variation was reasonably low, as evidenced by the
variation coefficient (32.31%). We assume that the type of anchor wire, in addition
to the anchor/wire interface, can directly influence the total force supported by
the fixture. This finding is corroborated by a work showing that an anchorage using
Ethibond 2.0 (Ethicon) is more fragile than one using Ethibond 5.0.[24 ] Knowing that wires with much higher resistance than Ethibond 5.0 are available,
this technique proved to be quite strong to keep the patella on its trochlear track.
Barber et al concluded that some more recent suture anchors showed significant improvements
in load to failure values when compared to braided polyester sutures. Therefore, suture
material influences failure modes.[23 ]
[25 ]
Another factor that apparently contributed to the anchor not being pulled out was
the graft traction angle, of approximately 90° ([Figure 1 ]), avoiding more acute angles (< 90°) that facilitate anchor pullout[.21 ]
Group 3 showed lower tensile strength ([Figure 4 ]), and all failures resulted from graft rupture after suture (Ethibond 5.0) transfixation
for biotenodesis. Thus, it seems that the use of a transient suture may weaken the
graft, increasing its vulnerability. Although some studies showed a certain advantage
in not transfixing the graft,[26 ] further research on this subject are required.
Direct comparisons between human specimens are complicated, since factors such as
donor age and bone density differences are difficult to control. Therefore, we chose
to use porcine bones, which allowed us to control these factors. Since this study
involved zero-time biomechanical testing in immediate postoperative conditions, no
histological comparison was possible. Therefore, there are potential differences between
in vitro and in vivo results, including because of the contribution of the other static
and dynamic stabilizers at the patellofemoral joint. In addition, there was no measurement
of graft slippage to reduce error or intra/interobserver reliability tests.[11 ]
[27 ]
Conclusion
The use of bone tunnel interference screws in porcine knees is strong enough for femoral
fixation in medial patellofemoral ligament reconstruction, as well as the fixation
with mountable anchors and high resistance wire. Adductor tenodesis was deemed fragile
for this purpose.
