Prospective randomized clinical comparison of femoral transfixation versus bioscrew fixation in hamstring tendon ACL reconstruction--a preliminary report.

Abstract

The purpose of this study is to clinically evaluate hamstring tendon anterior cruciate ligament (ACL)-reconstruction using femoral fixation with bioresorbable interference screws and with a bioresorbable transfixation device.
The ACL-reconstruction using the transfixation device at the femoral side leads to less knee laxity and therefore to a better clinical outcome for the patient.
Prospective randomized clinical outcome study.
From February 2002 to December 2002, a total of 68 patients with hamstring ACL reconstruction using a femoral fixation once with TransFix (n=38; m:22 and f:16; median age=28.5 range 15-47) and the second with bioscrew (BS) (n=30; m:20, f:10; median age=25.5 range 13-61) completed the follow-up period. Patients in each group got a clinical assessment at 3, 6, and 12 months after surgery. The measurement of anterior translation of the tibia has been performed using the Rolimeter device.
No significant differences in the knee laxity testing using the Rolimeter device were seen between both groups and over time within these groups. Ninety percent of all patients had functionally normal or near normal International Knee Documentation Committee (IKDC) knee ligament ratings. The TF-group included 17 grade A, 19 grade B, and 2 grade C knees, and the BS-group had 12 grade A, 13 grade B, and 5 grade C knees. The IKDC rating, the OAK-score, the Tegner-activity-score, and the Lysholm-score did not show significant differences between the TF-group and the BS-group.
We disproved our hypothesis that the transfixation technique leads to less laxity and therefore to a better clinical outcome when compared to the use of BS. The clinical results in this study clarified that this technique is an effective and safe method for femoral hamstring fixation in ACL-reconstruction. However, this technique revealed no advantage compared to the bioscrew fixation technique within the short-term follow-up.

Full text

Page 1

Introduction
The reconstruction of the anterior cruciate ligament
(ACL) with hamstring tendons can be performed using
different
analysis of various fixation devices has shown that the
use of extracortical fixation techniques results in a high
ultimate failure strength but also clarified the low degree
femoralfixationmethods.Biomechanical
Tim Rose
Pierre Hepp
Julia Venus
Christoph Stockmar
Christoph Josten
Helmut Lill
Prospective randomized clinical comparison
of femoral transfixation versus bioscrew
fixation in hamstring tendon ACL
reconstruction—a preliminary report
Received: 22 March 2005
Accepted: 5 July 2005
? Springer-Verlag 2006
Abstract Purpose: The purpose of
this study is to clinically evaluate
hamstring tendon anterior cruciate
ligament (ACL)-reconstruction
using femoral fixation with biore-
sorbable interference screws and
with a bioresorbable transfixation
device. Hypothesis: The ACL-
reconstruction using the transfix-
ation device at the femoral side leads
to less knee laxity and therefore to a
better clinical outcome for the
patient. Type of study: Prospective
randomized clinical outcome study.
Methods: From February 2002 to
December 2002, a total of 68
patients with hamstring ACL
reconstruction using a femoral fixa-
tion once with TransFix (n=38;
m:22 and f:16; median age=28.5
range 15–47) and the second with
bioscrew (BS) (n=30; m:20, f:10;
median age=25.5 range 13–61)
completed the follow-up period.
Patients in each group got a clinical
assessment at 3, 6, and 12 months
after surgery. The measurement of
anterior translation of the tibia has
been performed using the Rolime-
ter?device. Results: No significant
differences in the knee laxity testing
using the Rolimeter device were seen
between both groups and over time
within these groups. Ninety percent
of all patients had functionally nor-
mal or near normal International
Knee Documentation Committee
(IKDC) knee ligament ratings. The
TF-group included 17 grade A, 19
grade B, and 2 grade C knees, and
the BS-group had 12 grade A, 13
grade B, and 5 grade C knees. The
IKDC rating, the OAK-score, the
Tegner-activity-score, and the Lys-
holm-score did not show significant
differences between the TF-group
and the BS-group. Conclusion: We
disproved our hypothesis that the
transfixation technique leads to less
laxity and therefore to a better clin-
ical outcome when compared to the
use of BS. The clinical results in this
study clarified that this technique is
an effective and safe method for
femoral hamstring fixation in ACL-
reconstruction. However, this
technique revealed no advantage
compared to the bioscrew fixation
technique within the short-term
follow-up.
Keywords Anterior cruciate
ligament Æ Hamstrings Æ Transfix Æ
Interference screw Æ
Clinical outcome
DOI 10.1007/s00167-006-0034-9
Knee Surg Sports Traumatol Arthrosc
(2006)
KNEE
T. Rose (&) Æ P. Hepp Æ J. Venus
C. Stockmar Æ C. Josten Æ H. Lill
Department of Trauma and Reconstructive
Surgery, University of Leipzig,
Leipzig, Germany
T. Rose
Abteilung und Poliklinik fu ¨ r
Sportorthopa ¨ die, Technische Universita ¨ t
Mu ¨ nchen, Connollystrasse 32,
80809Mu ¨ nchen, Germany
E-mail: tim.rose@lrz.tu-muenchen.de

Page 2

of stiffness of this construct [2, 23]. Hoher et al. [16] have
shown that this phenomenon may lead to a graft-tunnel
motion of up to 3 mm during physiological loading. The
so called ‘‘Bungy-cord’’ and ‘‘windshield-wiper’’ effects
degrade the bone-tendon healing within the bone tunnel
[25] and can lead to a possible tunnel enlargement [21].
The use of bioresorbable interference screws has the
advantageofdirecttendon-to-bone
acceptableinitialbiomechanical
[30, 31]. However, the micromotion between the graft
and the interference screw within the tunnel during a
cycle loading may lead to a slipping of the graft and
result in a secondary lengthening and loosening of the
graft [2, 13]. The use of a femoral biodegradable trans-
versal fixation technique (TransFix–Arthrex?, Naples,
USA) combines the characteristics of a high failure load
[2], less loss of tension during repetitive loading cycles
[2], and a fixation closer to the joint line [9].
The goal of this study is to evaluate the clinical out-
comes using a transfixation technique in hamstring ACL
reconstruction. Our hypothesis was that the ACL-
reconstruction using the transfixation device at the
femoral side would lead to less knee laxity and therefore
to a better clinical outcome for the patient.
healingwith
fixation strength
Methods
Anterior cruciate ligament reconstructions with ham-
string tendons using two different fixation methods were
performed from February 2002 to December 2002 in 106
consecutive patients. All operations were done by the
same surgeon who had extensive experience in both pro-
ceduresprior tothe studyperiod. 76(72%)patientscould
be included in this study, fulfilling the inclusion criteria.
The inclusion criteria were acute or chronic anterior
instabilityofthekneejoint.Wealsoincludedpatientswith
additionalmeniscalinjuriesand/orfocalchondrallesions.
The exclusion criteria were signs of infection, osteoar-
thritis, reduced general condition, prior reconstruction,
andconcomitantinjurytotheposteriorcruciateligament.
These patients underwent a non-blinded randomization.
The surgeon pulled a card indicating the fixation tech-
nique. 68 of the patients (89% of all randomized patients:
TransFix: 38; bioscrew: 30) were available for the com-
plete follow-up period of 12 months consisting of post-
operative examination at 3, 6, and 12 months. Patients in
the TransFix (TF)-group consisted of 22 men and 16
women. The median age was 28.5 years (range 15–47).
The median age of the 20 men and ten women patients in
the bioscrew (BS)-group was 25.5 years (range 13–61).
Surgical technique
After separation of the semitendinosus and gracilis
tendons, the tendons were harvested with a tendon
stripper (Arthrex). Both tendons were prepared using
the four-strand method. All tendons were of sufficient
length to make quadruple grafts. The femoral and tibial
stumps of the torn ACL were removed to allow an
anatomic insertion of the new ACL-graft. At the tibial
site, a 2.4 mm K-wire was drilled at the correct site of
the place of tibial attachment of the ACL using a tibial
guide system (Arthrex). The guidewire was overdrilled
with a 6 mm reamer and then dilated according to the
maximum diameter of the graft. The femoral tunnel was
performed as required by the selected fixation method.
TransFix technique
The femoral tunnel was drilled passing the tibial tunnel,
with the knee flexed at a 70–90? angle. The insertion site
was placed as close as possible to the posterior wall and
the 11 o’clock position. After the insertion of a K-wire, a
foot print was made using a burr, which was sized to
match the graft. A 30-mm depth hole, 1–2 mm smaller
than the graft in diameter, was drilled, followed by a
dilatation up to the size of the graft. The TransFix guide
sleeve was inserted through the tibial tunnel and the
lateral pin insertion was prepared step-by-step. We used
a bioresorbable TansFix?pin for the femoral fixation.
After the final fixation at the tibial site, the harvested
cancellous bone plug was inserted into the femoral
tunnel using a specific implantation device. The knee
was in maximum flexion.
Bioscrew technique
The femoral tunnel was drilled from the antero-medial
portal with the knee flexed 130?. The tunnel drilling and
dilatation was comparable to the technique described
above. We used a 28 mm bioscrew (Arthrex) with the
diameter matching the graft size.
The tibial fixation was performed using a bioresorb-
able deltascrew?(Arthrex), with the leg extended at 0?,
and the graft under manual tension. The diameter of the
screw was approximately 2 mm larger than the graft.
Postoperative rehabilitation care
The same postoperative rehabilitation protocol was used
for patients in both treatment groups. Full weight
bearing was allowed from the first day post surgery. In
the phase immediately following surgery (days 1–4),
continuous passive motion (CPM) was performed using
a motorized CPM device. Analgesic treatment was
applied if necessary in combination with local cooling
and anti-inflammatory medication (non steroidal anti-
inflammatory drugs). From the fifth day on, physical

Page 3

therapy and the CPM device was used to obtain up to
90? flexion and full extension of the knee. Knee
protection was provided by a knee brace (Artrocare
CTS?, Ormed?, Freiburg, Germany) that was fitted and
worn until 6 weeks post surgery without limitation of
extension or flexion. More intensive physical therapy
aimed at specific muscle development. It was initiated in
the third week using muscle sequences and exercises in a
closed muscle chain. Beginning at week 7, post-surgery
coordination exercises were introduced to improve pro-
prioception. Sports activity (e.g., running and open
muscle chain exercise) was permitted after 12 weeks of
rehabilitation. However, sports involving contact and/or
pivot shifting were not allowed until 6 months post
surgery [20].
Clinical assessment
Patients in each group were examined at 3, 6, and
12 months. At each post-operative examination, the
following assessment instruments and measurements
were used and acquired:
1. Overall
Committee (IKDC) rating,
2. Lysholm knee scoring scale,
3. Tegner activity scale,
4. Lachmann and pivot-shift laxity test,
5. Measurement of anterior translation of the tibia with
manual maximum force using the Rolimeter?device
(Aircast?, Summit, NJ, USA) in comparison to the
healthy knee with the knee joint at the 30? position,
and
6. Range of motion measurement using a protractor.
International Knee Documentation
All questions were answered by the patient without
supervision. Two independent examiners, neither of
them the operating surgeon, conducted the clinical
examinations and patient consultations. Anterior laxity
was measured three times by one examiner to determine
the intra-reliability for this examiner regarding the
current patient. Knee extension was evaluated by range
of motion and was compared to the extension angle of
the healthy knee.
Radiographs
Standardized radiographs of the knee were made in
anterior–posterior views and in lateral views preopera-
tively, postoperatively, and at the follow-up examination
at 12 months. If necessary, MRI of the knee was used
for the diagnosis of associated injuries. Furthermore, if
possible, an MRI was performed after the 12 months
follow-up examination for control of the graft incorpo-
ration and tunnel placement.
The placement of the bone tunnels at the femoral site
and at the tibial site was evaluated using different
assessment techniques. At the femoral site (anterior–
posterior view), we looked for the o’clock position
according to the description by Sommer et al. [28].
Because of the difference between the left knee and the
right knee, we used the right knee for the standard
position, whereas the position of the left knee was flip-
ped (i.e., 1:30 o’clock at the left knee approximately
10:30 o’clock for our analysis). The lateral view at the
femoral site was evaluated by the femur 4-zone method
described by Harner et al. [15]. A similar method was
used for the lateral view at the tibial site. According to
the suggestions in the literature, we tried to place the
tunnels at the 10:30 o’clock position at the femoral
anterior–posterior view, and for the lateral view in the
fourth (most posterior) zone at the femoral site, and in
the third zone at the tibial site. The impingement quo-
tient was calculated using the method described by
Staubli and Rauschning [29].
Statistical analysis
The data are presented by descriptive statistics (i.e.,
mean, standard deviation, median, and range). The
ordinal variables (i.e., join laxity, range of motion,
and pain level) were examined using the chi-square
test. The nonparametric Mann–Whitney U test was
used to assess variance in the performance scores used
to identify possible differences between the two surgi-
cal procedures (OAK score, Lysholm score, Tegner
score), and the nonparametric Wilcoxon test was used
to evaluate differences over time. All significance tests
were two-tailed, and statistical significance was set at
0.05.
Results
The demographic data in both groups did not show any
difference. The concomitant injuries are listed in
Table 1. We had no revision ACL reconstruction in this
series. Two superficial infections but no deep infections
occurred postoperatively. We have never seen a migra-
tion of the implants (bioscrews and the transfix-device),
and therefore no hardware re-movement was necessary
within the follow up.
The time for the ACL-reconstruction procedures, the
TF (80 min.; range 38–150) and BS (80 min; range
43–125) did not differ (p=0.8). Also, the postoperative
hospitalization time did not show significant differences
(p=0.5) between both groups (TF-group: 8 days, range
6–13; BS-group: 7.5 days, range 5–11).

Page 4

Knee joint laxity
During
12 months after the surgery, four patients in each group
showed a positive result for the pivot-shifting phenom-
enon. At this time, 90% of all patients had functionally
normal or near normal IKDC knee ligament ratings.
The TF-group included 17 grade A, 19 grade B, and 2
grade C knees, and the BS-group had 12 grade A, 13
grade B, and 5 grade C knees. No significant difference
in the knee laxity testing using the Rolimeter device was
seen between the two groups (Table 2).
the clinicallaxityexaminationconducted
Tunnel placement
The femoral tunnel placement following ACL-recon-
struction revealed no significant differences between the
TF-group and the BS-group (Table 3). However, the
femoral tunnels of the TF-group tend to be closer to the
12 o’clock position (in a.p.—view) when compared to
the femoral tunnels of the BS-group. The tunnels in the
BS-group are closer to the 11 (or 13) o’clock position.
The evaluation of the postoperative radiographs with a
followingcalculation regarding
quotient for the ACL-graft [17] was positive for four
knees in the TF-group and for seven knees in the
BS-group. But the chi-square test revealed no significant
(p=0.14) differences between these groups. A correla-
tion between the impingement quotient and knee laxity
or between impingement quotient and extension deficit
was not detected.
theimpingement
Thigh muscle and range of motion
Between 3 and 12-months postoperative, there was a
clear improvement in knee joint mobility in both groups
(Table 4). No significant differences were measured
between these groups during the 3, 6, and 12 months
postoperative examination.
TF-group had an extension deficit in the repaired knee
(relative to the healthy knee) at the end of the follow up
and two patients in the BS-group. A deficit in flexion
(<120?) was seen in 13 patients in the TF-group and in
six patients in the BS-group after 12 months. The side-
to-side difference of the thigh muscle circumference was
in the TF-group: 1.5 cm (0–6) at 3 months, 1 cm (0–4) at
6 months, and 1 cm (0–4) at 12 months. The difference
in the BS-group was: 2 cm (0–5) at 3 months, 1 cm (0–3)
at 6 months, and 1 cm (0–2.5) at 12 months. No sig-
nificant differences were calculated at 3 (p=0.80), 6
(p=0.69), and 12 months (p=0.45) between these two
Fourpatients inthe
Table 2 Values of the laxity test using the Rolimeter
Follow up Rolimeter TF groupBS group
3 months Contralateral side
Operated side
Difference
p
Contralateral side
Operated side
Difference
p
Contralateral side
Operated side
Difference
p
6.7±1.9 mm
7,1±1.7 mm
0.4±2.2 mm
0.36
6.4±2.1 mm
7.3±1.7 mm
0.9±2.5 mm
0.76
6.5±2.1 mm
7.3±1.6 mm
0.8±2.1
0.14
6.8±2.3 mm
7.8±2.5 mm
0.9±2.3 mm
6 months7.1±2.3 mm
7.6±1.7 mm
0.5±2.4 mm
12 months7.1±2.3 mm
7.6±1.7 mm
1.2±2.8 mm
TransFixBioscrew
Age 28,5 years (15-47)25,5 years (13-61)
Female (n)1610
Male (n) 2220
Meniscal injuries (n)16 (3 x refixed, 13 x resected) 13 (2 x refixed, 11 x resected)
grade:12341234 Chondral lesions
n:04222213
Graft size 7,7 mm (sd=0,6)7,7 mm (sd=0,7)
Pre-injury Sport level
Professional:6
Amateur: 16
No organization: 16
Professional:6
Amateur: 10
No organization: 14
Table 1 Demographic data of both treatment groups

Page 5

groups. However, only nine patients in the TF-group
and five patients in the BS-group reached the same size
of muscle on the operated side compared to the
non-operated side (Table 5).
Qualitative evaluation
The IKDC rating did not show significant differences
(p=0.3) between the TF-group and the BS-group. 76%
of the knees in the TF-group and 73% of the knees in
the BS-group had normal or near normal function
after 12 months (Table 6). Average OAK-scores were
calculated at each follow-up examination and revealed
no differences between both fixation techniques during
the follow up (Fig. 1). Also the pain profile, deduced
from the OAK-score, was low and did not show differ-
ences at 3 months (p=0.85), 6 months (p=0.30), or
12 months (p=0.89). The Lysholm score demonstrated
similar results (Fig. 2) with no significant differences at
all time points (3 months: p=0.90; 6 months: p=0.20;
and 12 months: p=0.09). The degree of activity was
determined using the Tegner activity scale (maximum
score=10 points). No significant differences between the
two treatment groups were found preoperatively at
3 months (p=0.93), 6 months (p=0.53), or 12 months
(p=0.74) post surgery. However, there was a significant
(p<0.05) decrease in the activity scores within both
groups found during each follow-up examination
(Fig. 3), while the level of sport activity prior to the
accident was similar in both treatment groups (Table 1),
only 63% of the patients in the TF group could regain
similar level of sport activity as compared to the prein-
jury level after 12 months, whereas 83% of the patients
in the BS-group were able to participate at the same
sport level compared to the preinjury level (Table 6).
The p value was 0.06.
Discussion
Our results have shown that the use of the transfixation
technique at the femoral site has not led to significant
differences in clinical outcome when compared to the use
of bioscrews. The parameters for laxity, the IKDC-
score, the OAK-score, the Lysholm-score, and the
Tegner activity score were similar between both groups
and comparable with the results in the literature. The
fixation devices, the bioscrew and the TransFix devices,
as well, have shown a low complication rate.
The interference screw fixation of a quadrupled
hamstring graft results in a low linear stiffness, which
may result at submaximal loads in fixation failure [1].
The micromotion between graft and screw during
repetitive loading in the early postoperative phase may
lead to elongation of the graft, slippage, and secondary
graft loosening [13]. Recent biomechanical analysis [2, 6]
Table 3 Tunnel placement at the femoral site and at the tibial site measured using the x-rays
Femoral zone
(a.p. view) [28]
Femoral zone
(lateral view) [15]
Tibial zone
(lateral view) [15]
Impingement
quotient [29]
O’clock position
11:40 (±30 min)
11:10 (±30 min)
0.22
4
32
24
0.90
3
6
6
2
0
0
1
0
0
4
0
0
0.37
3
36
27
2
2
3
1
0
0
%
43 (SD=3.0)
44 (SD=4.0)
0.26
TF-group
BS-group
p
Table 4 Range of motion during 3, 6, and 12 months postopera-
tive examination
Follow up TF groupBS group p
3 monthsExtension deficit
Flexion deficit
Extension deficit
Flexion deficit
Extension deficit
Flexion deficit
4±2.9
12±8.3
2±3.0
7±7.4
2±2.9
5±6.2
6±3.4
10±7.9
3±3.2
7±6.4
1±2.8
4±4.9
0.35
0.1
0.67
0.59
0.22
0.55
6 months
12 months
Table 5 IKDC-rating score during the 3, 6, and 12 months post-
operative examination
Follow
up
IKDC—
scoring
TF group
(n)
BS group
(n)
Significance
3 monthsA
B
C
D
A
B
C
D
A
B
C
D
00
7
p=0.43
12
21
5
0
26
9
3
7
22
8
1
21
2
0
18
11
1
7
15
8
0
6 monthsp=0.42
12 monthsp=0.70
Table 6 Sport level after 12 months compared to the sport level
before torn the ACL (p=0.06)
TFBS
Similar sport level
Lower sport level
24 (63%)
14 (37%)
25 (83%)
5 (17%)