Preparation of the Intact Knee
Ten cadaveric knees (average age of donors, sixty-five years; range,
fifty-six to seventy-four years) were harvested and stored at -20°C.
Radiographs and inspection of the knee at the time of anterior cruciate
ligament reconstruction did not reveal moderate or severe degenerative
arthritis, chondrocalcinosis, or torn menisci. The intact knee was thawed
overnight. All soft tissue was removed 7 cm distal to and proximal to the knee
joint line. The diaphysis of the femur was cut 20 cm proximal to the joint
line, and a 12.7-mm-diameter, 28-cm-long steel rod was cemented inside the
medullary canal of the femur to within 7.5 cm of the joint line. The diaphysis
of the tibia was cut 18 cm distal to the joint line, and a 12.7-mm-diameter,
43-cm-long steel rod was cemented inside the medullary canal to within 7.5 cm
of the joint line. The diaphysis of the femur was cemented in a
6.4-cm-diameter, 20-cm-long aluminum cylinder to a distance 6.4 cm proximal to
the joint line.
The knee was placed supine in a custom-built six-degree-of-freedom testing
apparatus that permitted unconstrained knee motion for flexion angles ranging
from 30° to hyperextension. The testing apparatus was designed and built
in our laboratory and was used for tensioning the graft, inserting each tibial
fixation device, measuring the resultant force on the proximal end of the
graft, measuring the maximal anterior translation of the knee, and applying
cyclic loading treatments to the knee (Fig.
1). The aluminum cylinder containing the femur was clamped in the
femoral fixture, with the flexion-extension axis of the knee perpendicular to
the sagittal plane. Motion of the tibia was unconstrained by attaching a
low-friction bearing to the end of the steel rod extending from the tibia and
resting the bearing on a low-friction Delrin plate (DuPont, Wilmington,
Delaware). The length, ankle height, weight, and center of gravity of the
tibia were set to that of the shank-foot complex of an 81-kg, 180-cm-tall man
on the basis of anthropometric
measurements11-13.
The length of the tibia-steel rod-low-friction bearing was set to 51.5 cm, and
the height of the tibia anterior to the low-friction bearing was set to 2.9
cm, which represents the height of the ankle anterior to the heel. With use of
an iterative protocol, a weight was attached to the tibia to set the weight of
the shank and foot to 49 N and the center of gravity at 27.1 cm distal from
the knee joint line. Applying blocks of different heights under the
low-friction bearing set the flexion angle of the knee. Manual extension of
the knee until resistance was felt defined full
extension14. Knee
flexion was measured with a goniometer (Stryker Howmedica, Mahwah, New Jersey)
with an accuracy to within ±1°.
Measurement of the Maximal Anterior Translation of the Intact
Knee
The maximal anterior translation of the intact knee at 25° of flexion
in response to a 134-N anterior load applied perpendicular to the longitudinal
axis of the tibia was determined with a custom-made arthrometer with use of
the loading protocol of a commercial arthrometer (KT-1000; MedMetric, San
Diego, California) (Fig. 1).
Three 89-N posterior loads were applied to the tibia. The removal of the third
load defined the neutral position of the tibia. The maximal anterior
translation was the difference in the position of the tibia in response to a
134-N anterior load compared with the neutral position.
Technique of Anterior Cruciate Ligament Reconstruction
The tibial metaphysis was reinforced with polyurethane foam to provide
fixation properties in cadaveric tibiae from elderly individuals that were
similar to those in tibiae from young
individuals15. The
anterior cruciate ligament was excised, and tibial and femoral drill-holes
were made with use of a previously described transtibial technique that
positions the graft without roof impingement, without impingement of the
posterior cruciate ligament, and with a tension pattern that matches that of
the intact anterior cruciate
ligament16. The
tibial tunnel was drilled to 8 mm in diameter and was serially dilated in
0.5-mm increments to 9 mm following the recommendation of the manufacturer of
the interference screw. A femoral guide-pin was placed with use of a femoral
aimer with a 5.5-mm offset that was inserted through the tibial tunnel and was
hooked in the over-the-top position. An openend femoral tunnel was drilled to
16 mm in diameter. The blow-out of the posterior wall of the femoral tunnel
was later closed with bone cement. A low-friction femoral bushing with an
outer diameter of 16 mm was machined from Delrin
(Fig. 2). A 9-mm-diameter
tunnel was drilled from distal to proximal through the central axis of the
bushing to a depth of 10 mm. From the opposite end of the bushing, a
12.5-mm-diameter tunnel that stopped at the 9-mm-diameter was drilled. The
bushing was inserted into the femoral tunnel until it was flush with the
intercondylar roof and was fixed there with bone cement.
After the experiment was completed, an anteroposterior and lateral
radiograph with the knee in full extension was used to verify that the tibial
tunnel was properly placed in each knee. The tibial tunnel was imaged by
placing a 9-mm-diameter stainless-steel rod (impingement rod; Arthrotek,
Warsaw, Indiana) through the tibial tunnel and into the notch. In the coronal
plane, the rod was centered between the tibial spines at a mean angle (and
standard deviation) of 65° ± 3° from the medial joint line of
the tibia. In the sagittal plane, the rod grazed and was parallel to the
intercondylar roof. The tension pattern of the graft matches the intact
anterior cruciate ligament when the femoral tunnel is drilled through a tibial
tunnel placed with these coronal and sagittal radiographic
guidelines16. The
foam-reinforced knee was then replaced in the testing apparatus. The femoral
fixation transducer was made from stainless steel to glide with low friction
in the 12.5-mm-diameter section of the femoral bushing
(Fig. 2). A crossbar (stainless
steel with a 2.4-mm diameter and 10-mm length) was welded to the distal end of
the femoral fixation transducer to fix the double-looped tendon graft.
Forty double-looped tendon grafts were made from bovine extensor
tendon15, which has
structural properties similar to those of a double-looped semitendinosus and
gracilis graft from a young
human17. The
tendons were trimmed until, when looped over a suture, they passed snugly
through a 9-mm-diameter cylinder and not through an 8-mm-diameter cylinder
(Sizing Sleeve; Arthrotek). Four centimeters of the end of each strand were
whip-stitched with use of a number-1 braided, absorbable suture (Polysorb;
United States Surgical/Syneture, Norwalk,
Connecticut)15. A
fresh graft was selected at random for each test and was looped over the
crossbar. The femoral fixation transducer and the graft were inserted into the
femoral tunnel and were connected to a femoral load cell (225 N, SM-50;
Interface, Scottsdale, Arizona). The femoral load-cell, which measured the
resultant force on the proximal end of the graft, was connected to a
turnbuckle that was connected to the base-plate of the testing apparatus. The
turnbuckle was adjusted so that the crossbar was positioned to glide in the
femoral bushing.
Insertion of the Tibial Fixation Devices
Four tibial fixation devices were tested in the following sequence on each
of the ten knees: a spiked, metal soft-tissue washer (20-mm-diameter
No-Profile Tissue Washer; Arthrotek), double staples (medium [11.11-mm] wide,
25.4-mm-long Richards fixation staples with spikes; Smith and Nephew Richards,
Memphis, Tennessee), a bioabsorbable soft-tissue interference screw
(11-mm-diameter, 35-mm-long Cannulated Delta Tapered Bio-Interference Screw;
Arthrex, Naples, Florida), and the WasherLoc (16-mm diameter; Arthrotek).
Three cyclic loading treatments were applied in succession to
conservatively load the graft and all four fixation devices. Each cyclic
loading treatment consisted of maintaining the knee in full extension for
fifteen minutes and then flexing the knee to 25° and cyclically loading
the knee twenty times between a posterior load of 26 N and an anterior load of
100 N. The anterior load of 100 N generated an intra-articular tension of 170
N in the graft, which is similar to the tension in the anterior cruciate
ligament during level
walking18. The
resultant force on the proximal end of the graft and the maximal anterior
translation were recorded after each cyclic loading treatment. For each tibial
fixation, a fresh double-looped tendon graft was used.
The spiked washer was screwed partway into a tapped 4.5-mm bicortical
drill-hole positioned 15 mm distal to the tibial tunnel on the anteromedial
cortex. The knee was placed in full extension. Two tibial load-cells (225 N,
SM-50; Interface) were used to apply the tensile force to the graft. Each
tibial load-cell was connected to a pneumatic cylinder (Illinois Pneumatics,
Roscoe, Illinois) mounted on a fixture connected to the base-plate of the
testing apparatus proximal to the knee. For the spiked washer fixation, the
tensile force was applied proximal to the knee. Two strands of one tendon were
wrapped 180° clockwise, and the other two strands were wrapped 180°
counterclockwise around the shank of the screw holding the spiked washer. The
sutures from the two strands of each tendon were tied, and each loop of suture
was hooked on a load-cell. The pneumatic cylinder attached to each tibial
load-cell was adjusted to maintain a tensile force of 110 N, with an accuracy
of ±1 N. The femoral load-cell recorded the resultant force on the
proximal end of the graft. The screw was firmly tightened, and the resultant
force on the proximal end of the graft and the maximal anterior translation
were recorded.
For the double staples, one pair of 1.9-mm-diameter cortical drill-holes
was made 7 mm distal to the tibial tunnel on the anteromedial cortex with use
of a drill-guide (medium width, fixation staple drill jig; Smith and Nephew
Richards), and a second pair of drill-holes was made 6 mm distal to the first.
The knee was placed in full extension. The tibial load-cells were repositioned
distal to the knee, and a 110-N tensile force was applied to the graft at
approximately 20° with respect to the surface of the anteromedial cortex
of the tibia. The resultant force on the proximal end of the graft was
recorded. The first staple was driven in the proximal drill-holes, and the
second staple was driven in the distal drill-holes with use of a staple driver
(fixation staple driver-extractor; Smith and Nephew Richards).
For the interference screw, the knee was placed in full extension. The
tibial load-cells were repositioned distal to the knee, and a 110-N tensile
force was applied to the graft in line with the long axis of the tibial
tunnel. The resultant force on the proximal end of the graft was recorded. A
1.1-mm-diameter guide-pin was placed between the anterior wall of the tibial
tunnel and the graft. The interference screw was inserted until the distal end
of the screw was flush with the distal cortex of the tibial tunnel.
For the WasherLoc, a 17-mm-diameter counterbore was drilled into the distal
end of the tibial tunnel. The tibial load-cells were positioned distal to the
knee, and a 110-N tensile force was applied to the graft in line with the
tibial tunnel. The resultant force on the proximal end of the graft was
recorded. The WasherLoc was threaded on a drill sleeve, and the drill sleeve
was threaded on an awl. The awl was positioned in the hole created by the
counterbore, and one strand from each tendon was placed on opposite sides of
the awl. Striking the awl with a mallet drove the WasherLoc into the bone
within the counterbore. A 6.5-mm-diameter, self-tapping cancellous screw was
inserted through the WasherLoc and was tightened to fix the graft.
Correction for Frictional Loss in the Femoral Tunnel
For each reconstructed knee, the resultant force on the proximal end of the
graft was corrected for frictional loss in the femoral tunnel with use of a
correction ratio, which yielded the intra-articular graft tension. The
correction ratio was determined at the end of each experiment after removing
the tibia from the femur. A double-looped tendon graft was inserted in the
femoral tunnel, and the graft was oriented parallel and adjacent to the
intercondylar roof to position the graft as it was positioned during
measurements of graft tension with the knee in full extension. A tensile force
of 100 N was applied to the distal end of the graft, and the resultant force
on the proximal end of the graft was recorded with the femoral fixation
transducer. The correction ratio was equal to 100 N divided by the resultant
force on the proximal end of the graft. The resultant force on the proximal
end of the graft during each experiment was multiplied by the correction ratio
to yield the intra-articular graft tension.
Statistical Analysis
For each tibial fixation device, the intra-articular tension measured after
the third cyclic loading treatment (i.e., the final measurement) was compared,
with use of a one-sample t test, with the tensile force. To determine whether
friction caused a loss in intra-articular tension, the intra-articular tension
measured after applying the tensile force (i.e., the first measurement) was
compared, with use of a one-sample t test, with the tensile force. To
determine whether inserting the tibial fixation device and each cyclic loading
treatment caused a loss in intra-articular tension, a one-factor
repeated-measures analysis of variance was used, with the independent variable
consisting of five levels (applying the tensile force, inserting the tibial
fixation device, the first cyclic loading treatment, the second cyclic loading
treatment, and the third cyclic loading treatment) and with the dependent
variable being intra-articular tension.
For each tibial fixation device, a one-sample t test was used to compare
the maximal anterior translation after the third cyclic loading treatment and
that of the intact knee. A one-sample t test was also used to compare the
maximal anterior translation after inserting the tibial fixation device and
that of the intact knee. To determine whether each cyclic loading treatment
increased the maximal anterior translation, a one-factor repeated-measures
analysis of variance was used, with the independent variable having four
levels (after inserting the tibial fixation device, after the first cyclic
loading treatment, after the second cyclic loading treatment, and after the
third cyclic loading treatment) and with the dependent variable being the
difference in the maximal anterior translation of the reconstructed knee from
that of the intact knee. Significant main effects were further analyzed with
the Tukey test. The level of significance was set at p < 0.05.
Spiked Washer
For the spiked washer, there was a 50% loss in intra-articular
tension from the 110-N tensile force applied to the graft and the third cyclic
loading treatment (p = 0.0004) (Fig.
3). Friction from the tibial tunnel as a result of wrapping the
graft around the shank of the screw caused a loss in intra-articular tension
to a mean (and standard deviation) of 32 ± 6 N (range, 19 to 39 N) (p
< 0.0001). Inserting the spiked washer increased the intra-articular
tension to a mean of 111 ± 30 N (range, 58 to 148 N) (p < 0.05). The
first cyclic loading treatment caused a loss in intra-articular tension to a
mean of 79 ± 32 N (range, 28 to 142 N) (p < 0.05), the second cyclic
loading treatment did not change the intra-articular tension (mean, 65
± 33 N; range, 21 to 137 N), and the third cyclic loading treatment did
not change the intra-articular tension (mean, 55 ± 32 N; range, 16 to
126 N).
The difference in the maximal anterior translation of the reconstructed
knee after insertion of the spiked washer and that of the intact knee was 0.0
± 1.4 mm (range, -2.9 to 1.9 mm) (p = 1.0000), which increased to 2.0
± 1.7 mm (range, -1.3 to 4.3 mm) after the third cyclic loading
treatment (p = 0.005). The maximal anterior translation had not stabilized by
the third cyclic loading treatment (p < 0.05)
(Fig. 3).
Double Staples
For the double staples, there was a 100% loss in intra-articular tension
from the 110-N tensile force applied to the graft and the third cyclic loading
treatment (p < 0.0001) (Fig.
4). Friction from the tibial tunnel caused a loss in
intra-articular tension to a mean of 82 ± 6 N (range, 70 to 88 N) (p
< 0.0001). Inserting the staples caused a further loss in intra-articular
tension to a mean of 52 ± 10 N (range, 39 to 70 N) (p < 0.05). The
first cyclic loading treatment caused a further loss in intra-articular
tension to a mean of 1.6 ± 1.3 N (range, 0 to 4.3 N) (p < 0.05), the
second cyclic loading treatment did not change the intra-articular tension
(mean, 0.8 ± 0.8 N; range, 0 to 2.2 N), and the third cyclic loading
treatment did not change the intra-articular tension (mean, 0.3 ± 0.5
N; range, 0 to 1.1 N).
The difference in the maximal anterior translation of the reconstructed
knee after insertion of the double staples from that of the intact knee was a
mean of 2.8 ± 1.1 mm (range, 1.6 to 4.9 mm) (p < 0.0001), which
further increased to a mean of 7.8 ± 1.9 mm (range, 5.3 to 10.7 mm) (p
< 0.0001) after the third cyclic loading treatment. The maximal anterior
translation stabilized by the second cyclic loading treatment (p = 0.05)
(Fig. 4).
Interference Screw
For the interference screw, there was a 64% loss in intra-articular tension
from the 110-N tensile force applied to the graft and the third cyclic loading
treatment (p = 0.0001) (Fig.
5). Friction from the tibial tunnel caused a loss in
intra-articular tension to a mean (and standard deviation) of 98 ± 4 N
(range, 89 to 104 N) (p < 0.0001). Insertion of the interference screw did
not change the intra-articular tension (mean, 91 ± 28 N; range, 58 to
141 N). The first cyclic loading treatment caused a further loss in
intra-articular tension to a mean of 58 ± 35 N (range, 10 to 117 N) (p
< 0.05), the second cyclic loading treatment did not change the
intra-articular tension (mean, 47 ± 35 N; range, 2 to 108 N), and the
third cyclic loading treatment did not change the intra-articular tension
(mean, 40 ± 34 N; range, 1 to 105 N).
The difference in the maximal anterior translation of the reconstructed
knee after insertion of the interference screw and that of the intact knee was
a mean of 1.1 ± 1.1 mm (range, -0.4 to 3.1 mm) (p = 0.011), which
further increased to a mean of 2.7 ± 2.0 mm (range, 0.2 to 7.4 mm) (p =
0.001) after the third cyclic loading treatment. The maximal anterior
translation stabilized by the second cyclic loading treatment (p = 0.05)
(Fig. 5).
WasherLoc
For the WasherLoc, there was a 56% loss in intra-articular tension from the
110-N tensile force applied to the graft and the third cyclic loading
treatment (p < 0.0001) (Fig.
6). Friction from the tibial tunnel caused a loss in
intra-articular tension to a mean of 100 ± 4 N (range, 93 to 107 N) (p
< 0.0001). Inserting the WasherLoc caused a further loss in intra-articular
tension to a mean of 79 ± 20 N (range, 41 to 112 N) (p < 0.05). The
first cyclic loading treatment caused a further loss in intra-articular
tension to a mean of 62 ± 18 N (range, 23 to 88 N) (p < 0.05), the
second cyclic loading treatment did not change the intra-articular tension
(mean, 54 ± 18 N; range, 17 to 81 N), and the third cyclic loading
treatment did not change the intra-articular tension (mean, 48 ± 16 N;
range, 15 to 71 N).
The difference in the maximal anterior translation of the reconstructed
knee after insertion of the WasherLoc and that of the intact knee was a mean
of 1.3 ± 1.0 mm (range, -0.9 to 2.6 mm) (p = 0.0039), which further
increased to a mean of 2.1 ± 1.0 mm (range, 0.1 to 3.7 mm) (p <
0.0001) after the third cyclic loading treatment. The maximal anterior
translation stabilized by the first cyclic loading treatment (p = 0.05)
(Fig. 6).
In our opinion, there are four implicit assumptions in the selection
of a tensile force at the time of anterior cruciate ligament reconstruction:
(1) the tensile force applied to the graft distal to the tibial tunnel is
transferred without change to the intra-articular portion of the graft, (2)
the intra-articular tension is maintained after insertion of the tibial
fixation device, (3) the intra-articular tension is maintained after cyclic
loading of the knee, and (4) the intra-articular tension restores the maximal
anterior translation to that of the intact knee. The present study shows that
the tensile force applied to a double-looped tendon graft is not transferred
intra-articularly, that the intra-articular tension is not maintained after
the insertion of a tibial fixation device and cyclic loading of the knee, and
that the loss in intra-articular tension increases the maximal anterior
translation. The clinical relevance of these results is that a single value
for a tensile force cannot restore the maximal anterior translation when knees
with a torn anterior cruciate ligament are reconstructed with a double-looped
tendon graft.
The first reason that a single value for a tensile force does not restore
the maximal anterior translation is that the direction in which the tensile
force is applied with respect to the knee and the type of fixation device have
different effects on the transfer of the tensile force to the intra-articular
portion of the graft. The application of the tensile force distal to the knee
with the double staples, interference screw, and WasherLoc caused a 10% to 26%
frictional loss between the graft and the tibial tunnel. The application of
the tensile force proximal to the knee with the spiked washer caused a 71%
frictional loss between the graft and the shank of the screw. Therefore, the
frictional loss between the graft and the tibial tunnel is determined by the
direction of the tensile force and the type of fixation device.
The second reason that a single value for a tensile force does not restore
the maximal anterior translation is that the insertion of the tibial fixation
device causes variability in the intra-articular tension. For each fixation
device, the mean intra-articular tension varied widely among the knees (111 N
after fixation with a spiked washer, 52 N after use of double staples, 91 N
with the interference screw, and 79 N with the WasherLoc), which also caused
wide variability in the mean maximal anterior translation (0.0 mm after
fixation with a spiked washer, 2.8 mm after use of double staples, 1.1 mm with
the interference screw, and 1.3 mm with the WasherLoc). The surgeon should be
mindful of the technique used to insert a tibial fixation device because the
technique changes intra-articular
tension19 and
anterior translation.
The final reason that a single value for a tensile force does not restore
the maximal anterior translation is that cyclic loading causes a loss in
intra-articular tension. The loss in intra-articular tension from the three
cyclic loading treatments increased the mean maximal anterior translation (2
mm for the spiked washer, 5 mm for the double staples, 1.6 mm for the
interference screw, and 0.8 mm for the WasherLoc). The increase in anterior
translation from cyclic loading is likely from lengthening at the two sites of
fixation (e.g., slippage and contact
deformation)10.
A limitation of this in vitro study is that the loss in intra-articular
tension and the increase in the maximal anterior translation that could occur
in vivo might have been underestimated. One reason for this is that resumption
of the activities of daily living and aggressive rehabilitation place more
tensile load on the knee than the three brief, conservative cyclic loading
treatments used in the present study. A more sustained or higher tensile load
on the knee, such as those from the activities of daily living and aggressive
rehabilitation, might cause more lengthening at the sites of
fixation10. Six
weeks of performing the activities of daily living corresponds to
approximately 220,000 cycles to the anterior cruciate
ligament20 at a
tensile load of 169
N18. Animal studies
have shown that fixation devices provide a substantial proportion of the
fixation until four to eight weeks after surgery, after which the fixation
transfers to the biologic bond between the graft and the bone
tunnel21-23.
Since patients with a soft-tissue graft begin walking without crutches and a
brace and resume exercise within the first week after
surgery24,25,
the increase in maximal anterior translation from slippage at the site of
fixation might be greater in vivo than in our study.
A second reason for this underestimation of tension loss is that there
might be a greater loss in intra-articular tension in vivo with the use of a
femoral fixation device that allows more lengthening at the femoral site of
fixation than the crossbar used in the present study. Femoral fixation
devices, such as a suture bridge, either attached to a button or tied to a
post (closed-loop and open-loop Endobutton; Acufex Microsurgical, Mansfield,
Massachusetts); interference screws (BioScrew [Linvatec, Largo, Florida] and
RCI [Acufex Microsurgical]); and two cross pins that skewer the graft
(RIGIDfix cross-pin guide; Mitek Products, Norwood, Massachusetts) allow
substantially more lengthening under cyclic load at the site of fixation than
the crossbar used in the present
study26,27.
The observation that intra-articular tension is lost from friction in the
tibial tunnel, insertion of the tibial fixation device, slippage at the site
of tibial and femoral fixation, and from activities of daily living raises the
possibility of compensating for the loss by using a higher tensile force.
Compensating for the loss in tension depends on predicting (1) the loss in
tension from each of these causes and (2) the intra-articular tension required
to restore the maximal anterior translation. Predicting the loss in tension is
a daunting, if not impossible, task because the tension loss from friction
depends on the tightness of fit between the graft and the tunnel, the
coarseness of the bone lining the
tunnel28,29,
and the angle of the wrap with respect to the long axis of the
tunnel28. The
tension loss from inserting the tibial fixation device varies widely even with
careful surgical technique, and the tension loss from activities of daily
living and aggressive rehabilitation is not quantifiable with any current
methods. Predicting the intra-articular tension required to restore the
maximal anterior translation for a given knee is not possible because the
tension is not the same for every fixation device and every knee. A post hoc
analysis revealed that the intra-articular tension that restored the maximal
anterior translation varied from 82 to 230 N for the spiked washer, from 98 to
224 N for the double staples, from 110 to 243 N for the interference screw,
and from 116 to 222 N for the WasherLoc. Considering these complexities, it is
unlikely that a single value of tensile force can be used for every knee.
One assumption of the experiment was that the use of a tensile force of 110
N was sufficient to restore the maximal anterior translation of the intact
knee. The choice of a tensile force of 110 N was considered to be appropriate
for the following reasons. First, the results from a pilot study, involving
three specimens, demonstrated that an intra-articular tension of 110 N with
the knee in full extension was the average tension required to restore the
maximal anterior translation to within ±0.5 mm of that of the intact
knee for the four fixation devices before cyclic loading of the knee. Second,
an intra-articular tension of 110 N with the knee in full extension matched
the anterior laxity of the intact knee with a double-looped tendon graft in an
in vitro
study16.
A second assumption of the experiment was that the nonrandomized,
sequential testing of the four tibial fixation devices did not cause carryover
effects that affected the loss in intra-articular tension. A carryover effect
might have occurred for the double staples, interference screw, and WasherLoc
if the insertion and removal of the preceding fixation device fractured the
bone. A fracture in the bone did not occur because the three cyclic loading
treatments were conservative in that the applied load was well below the yield
load of the fixation device in the foam-reinforced
tibia15, and the
bone was drilled and tapped before inserting the spiked washer and screw and
was drilled before impacting the double staples. The condition of the bone was
visually inspected after removal of each device, and the cortex was observed
to be intact. Considered together, this experimental approach and the visual
observations suggest that sequential, nonrandomized testing of the four
fixation devices did not lead to bone fracture and therefore did not produce
carryover effects that affected the loss in intra-articular tension.
A third assumption of the experiment was that the loss in intra-articular
tension and the increase in the maximal anterior translation were similar to
those that would be found in knees in young individuals. The loss in
intra-articular tension and the increase in the maximal anterior translation
might have been less if knees from young individuals had been used instead of
knees from elderly individuals that had reinforcement of the tibia with foam.
In the present study, the tibiae were reinforced with foam because (1)
lengthening at the site of fixation with tandem screws and washers,
interference screw, and WasherLoc in foam-reinforced tibiae from elderly
individuals is not substantially different from that in tibiae from young
individuals and (2) knees from young individuals are difficult to
obtain15. The
increase in the maximal anterior translation after cyclically loading the knee
with the spiked washer, interference screw, or WasherLoc in the present study
was consistent with lengthening measured in other studies that used either
tibiae from young individuals or more dense porcine
tibia9,15.
While the use of a foam-reinforced knee instead of a knee from a young
individual had little effect on the loss in intra-articular tension and the
increase in maximal anterior translation with the spiked washer, interference
screw, and WasherLoc, the loss in intra-articular tension and the increase in
maximal anterior translation might have been excessive with the double
staples. We followed the manufacturer's recommended technique of predrilling
the holes, which prevented the bone from fracturing during impaction of the
staples. However, the increase in the maximal anterior translation with
predrilling in foam-reinforced tibiae in the present study was greater than
the lengthening with predrilling in porcine
tibiae9, and it was
greater than the increase in anterior laxity in a clinical study that placed
the same staples without predrilling in bone in young
individuals25. This
could be the result of the use of foam-reinforced knees instead of knees from
young individuals and from predrilling the holes, which is a step recommended
by the manufacturer but one that we do not use in clinical practice.
In summary, the results in the present study suggest a biomechanical
explanation for the clinical observations that a single value for a tensile
force does not restore anterior laxity for a knee reconstructed with a
double-looped tendon graft and that there is variability in anterior laxity
after anterior cruciate ligament reconstruction. The present study indicates
that tensile force applied to a soft-tissue anterior cruciate ligament graft
is not fully transferred intra-articularly and is not maintained during cyclic
loading. The transfer of the tensile force into the knee is determined by the
direction that the tensile force is applied to the graft, which is determined
by the type of tibial fixation device. Surgeons should pay close attention to
the technique for inserting the tibial fixation device because the results of
this study support the assumption that this step induces the greatest change
and variability in the intra-articular tension and maximal anterior
translation in the knee reconstructed with a double-looped tendon graft.
Cyclically loading the knee causes a further loss in intra-articular tension
and an increase in the maximal anterior translation. The results of this study
support the assumption that the loss of intra-articular tension can be reduced
by the use of fixation devices that resist lengthening at the site of fixation
and by limiting cyclic loading of the knee. ?