Does Press-Fit Technique Reduce Tunnel Volume Enlargement
After Anterior Cruciate Ligament Reconstruction With
Autologous Hamstring Tendons? A Prospective Randomized
Computed Tomography Study
Dae-Hee Hwang, M.D., Gautam M. Shetty, M.S, Orth, Jong In Kim, M.D.,
Jae Ho Kwon, M.D., Jae-Kwang Song, M.D., Michael Muñoz, M.D.,
Jun Seop Lee, M.D., and Kyung-Wook Nha, M.D., Ph.D.
Purpose: The purpose of this prospective, randomized, computed tomographyebased study was to investigate whether
the press-fit technique reduces tunnel volume enlargement (TVE) and improves the clinical outcome after anterior
cruciate ligament reconstruction at a minimum follow-up of 1 year compared with conventional technique.
Methods: Sixty-nine patients undergoing primary ACL reconstruction using hamstring autografts were randomly allocated to either the press-fit technique group (group A) or conventional technique group (group B). All patients were
evaluated for TVE and tunnel widening using computed tomography scanning, for functional outcome using International
Knee Documentation Committee and Lysholm scores, for rotational stability using the pivot-shift test, and for anterior
laxity using the KT-2000 arthrometer at a minimum of 1-year follow-up. Results: There were no significant differences
in TVE between the 2 groups. In group A, in which the press-fit technique was used, mean volume enlargement in the
femoral tunnel was 65% compared with 71.5% in group B (P ¼ .84). In group A, 57% (20 of 35) of patients developed
femoral TVE compared with 67% (23 of 34) of patients in group B (P ¼ .27). Both groups showed no significant difference
for functional outcome (mean Lysholm score P ¼ .73, International Knee Documentation Committee score P ¼ .15), or
knee laxity (anterior P ¼ .78, rotational P ¼ .22) at a minimum follow-up of 1 year. Conclusions: In a comparison of
press-fit and conventional techniques, there were no significant differences in TVE and clinical outcome at short-term
follow-up. Level of Evidence: Level II, therapeutic study, prospective randomized clinical trial.
T
unnel widening (TW) after anterior cruciate ligament (ACL) reconstruction using hamstring tendon is a well-described phenomenon.1 TW has been
reported to be more common with hamstring tendon
autografts than with boneepatellar tendon autografts.2
From the Department of Orthopaedic Surgery (D-H.H., J.I.K., J.H.K.,
J-K.S., M.M., K-W.N.), Inje University, Ilsanpaik Hospital, Ilsan; Department of Orthopaedic Surgery (J.S.L.), Gumdan Top General Hospital,
Incheon, Korea; and Department of Orthopaedic Surgery (G.M.S.), Breach
Candy Hospital, Mumbai, India.
Supported by a 2011 Inje University Research Grant. The authors report that
they have no conflicts of interest in the authorship and publication of this article.
Received November 24, 2011; accepted July 3, 2012.
Address correspondence to Kyung-Wook Nha, M.D., PhD., Department of
Orthopaedic Surgery, Inje University, Ilsanpaik Hospital, 2240 Daehwadong,
Ilsanseogu, Goyangsi, Gyeonggido, 411-706, South Korea. E-mail:
kwnhamj@hotmail.com
Ó 2013 by the Arthroscopy Association of North America
0749-8063/11781/$36.00
http://dx.doi.org/10.1016/j.arthro.2012.07.007
This has been attributed to the longer time required for
soft-tissueetoebone healing and the lesser stiffness of
hamstring fixation techniques.3 Excessive TW after ACL
reconstruction may jeopardize stability of graft fixation
and compromise revision surgery because of poor bone
stock, and sometimes a staged procedure is required.1
The causes of TW are unclear and are presumed to be
multifactorial with mechanical and biological factors.
Mechanical factors include excessive motion of graft
that can result in a “bungee jump effect,” involving
longitudinal motion of the graft, or “windshield-wiper
effect,” involving transverse motion of the graft, tunnel positioning, and graft fixation method.1,3,4 Biological
factors include access of joint fluid within the graftebone
interface that contains osteolytic cytokines.5 Rigid initial
graft fixation has been cited to be critical to the success of
ACL reconstruction, and TW is more frequent when
indirect fixation techniques such as the cortical button
fixation are used compared with direct fixation methods
such as double cross-pin or transfixation screw.1,6,7
Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 29, No 1 (January), 2013: pp 83-88
83
84
D-H. HWANG ET AL.
Despite proper fixation, bone TW has been observed
following ACL reconstruction, especially when using
soft-tissue grafts. Any gap between the graft and the
tunnel wall can lead to ingress of synovial fluid and
delay in tendon-to-bone healing and subsequently TW.8
A press-fit technique for hamstring tendon fixation
might obviate the gap between the graft and tunnel wall
and induce more rapid tendonebone consolidation by
preventing ingress of joint fluids and maximizing the
contact surface of the bone and tendon.8 However, there
are few studies in the literature that have reported on
tunnel volume enlargement (TVE) with the press-fit
technique compared with the conventional technique.
The timeline of TW has also been studied. Fink et al.9
noted that the greatest amount of widening occurs
within the first 6 weeks. Webster et al.2 showed that the
radiographic tunnel width did not significantly change
between 4 months and 2 years. Therefore, the computed
tomography (CT) examination in the current study was
performed at 12 months after surgery. Hence, the
purpose of this study was to investigate whether the
press-fit technique reduces the TVE and improves the
clinical outcome after ACL reconstruction at a minimum
follow-up of 1 year compared with the conventional
technique. We hypothesized that the press-fit technique
would reduce the TVE and improve clinical scores
compared with the conventional technique.
Methods
All patients who underwent arthroscopic primary ACL
reconstruction using auto-hamstring tendon graft for an
ACL rupture at a single institution between October 2007
and January 2009 were eligible for enrollment in this
prospective randomized study. The inclusion criteria were
a unilateral isolated ACL rupture, age 18 to 40 years,
a normal contralateral knee, and informed consent by the
patient for participation in the study. Patients with an
associated meniscal injury were also included in the study.
There were 27 meniscal injuries in group A (press-fit
technique) and 28 meniscal injuries in group B (conventional technique). The exclusion criteria were concomitant ligament injury, cartilage degeneration greater than
Outerbridge grade 2 on arthroscopic assessment, ACL
reconstruction using an allograft, and a follow-up of less
than 1 year postoperatively. The study was approved by
the institutional review board of the hospital.
A total of 85 patients underwent the procedure during
the study period, and 79 of the 85 patients agreed to
participate in the study. Based on the exclusion criteria, 4
patients with an associated ligament injury and 3 patients
with cartilage degeneration greater than grade 2 were
excluded from the study. Finally, 72 patients were
enrolled in the study, in which the patients were
randomly allotted to a press-fit technique group or
a conventional technique group based on a random
number table.
Three patients were lost to follow-up for reasons not
known, and the final analysis involved data from 35
patients (35 knees) in the press-fit technique group
(group A) and 34 patients (34 knees) in the conventional
technique group (group B). The surgical procedure was
performed on all patients in both groups by a single
senior surgeon (K.W.N.), and the same postoperative
rehabilitation protocol was followed. After ACL reconstruction, all patients underwent CT scanning to determine femoral TW at 1 week postoperatively and at
a minimum of 1-year postoperative follow-up. There
were 31 male patients and 4 female patients in group A
with a mean age of 28.5 years (range, 19 to 40 years) who
were followed for a mean period of 13.5 months (range,
12 to 17 months), and group B had 27 male patients and 7
female patients with a mean age of 27.4 years (range, 18
to 40 years) who were followed for a mean period of 14.5
months (range, 12 to 18 months).
Surgical Technique
In all cases, the semitendinosus and gracilis tendons,
which had been harvested using a tendon striper, were
prepared as a 4-stranded double-looped hamstring
autograft. The ACL tibial guide (Linvatec, Largo, FL) was
set to 45 to prepare the tibial tunnel. The femoral tunnel
was drilled using the transtibial technique. Tunnel
diameter was determined according to the harvested
tendon diameter, which varied from 7 to 8 mm. For the
femoral tunnel, a 5.5-mm offset drill guide was placed on
the posterior aspect of the notch, which had been positioned as determined with the transtibial guide at
approximately 10:30 o’clock orientation for the right
knee or at approximately 1:30 o’clock orientation for the
left knee and drilled to a depth of 30 mm. In group A
(press-fit technique group), tunnel preparation was performed by underdrilling by 0.5 mm after measuring the
diameter of the prepared auto-hamstring tendon graft to
achieve press-fit of the graft in both the tibial and femoral
tunnel. In group B (conventional technique group),
tunnel preparation was performed using the same-sized
drill after measuring the diameter of the prepared graft to
allow easy passage of the graft in both tunnels.
The fixation technique for the femoral and the tibial
side was the same in both groups. After passing the graft
through the femoral tunnel, it was secured using the
RigidFix system (DePuy Mitek, Raynham, MA). Once
the graft had been secured on the femoral side, the knee
was positioned in 10 flexion and both loops were placed
on the IntraFix tie tensioner (DePuy Mitek) with 25-lb
force. The tibial end of the graft was then fixed using an
IntraFix biodegradable screw (DePuy Mitek) supplemented by a screw and spiked washer.
Postoperative Rehabilitation
Patients in both groups followed an identical accelerated postoperative rehabilitation regimen. Closed
PRESS-FIT TECHNIQUE AFTER ACL RECONSTRUCTION
Fig 1. The method of measurement of tunnel volumes in the
axial CT scan.
chain kinetic exercises were started on the third postoperative day. Full weight bearing was allowed 7 days
after surgery. Partial weight bearing with crutches for 6
weeks was mandated for patients who also underwent
a concomitant meniscal repair. The goal for the patients
was to gain a full range of motion at 2 to 6 weeks after
surgery. A perturbation training program was started at
6 weeks after surgery. Running and side-cutting activities were allowed at 3 months, with a return to sports
activities at 6 months after surgery.
CT Evaluation
A noncontrast CT scan of the operated knee in the
supine position was performed using a spiral CT scan
system (Somatom Emotion 6; Siemens GmbH, Berlin,
Germany). The CT images were acquired at 120 kV and
75 mA, with a slice thickness of 2 mm with a 0.5-mm
collimation. Scans were performed at 1 week and
a minimum follow-up of 1 year postoperatively. The
volume and diameter were measured by 2 orthopaedic
surgeons blinded to the groups to which the patients
were allotted. To measure the tunnel volume, a series
of CT cross-sectional images were obtained for each
femur condyle. The femoral tunnel was identified at the
axial viewing image. Using a digitized picture archiving
and communication system (STARPACS; Infinite
Healthcare, Seoul, South Korea), we estimated the
cross-sectional area of each tunnel in each image
(Fig 1). The cross-sectional area on each picture was
totaled and multiplied to calculate the total volume for
the femoral tunnels. The volume of the tunnel was
calculated using the formula pr2h. TVE was classified
into 2 categories based on the presence or absence of
enlargement. Tunnel enlargement was present if the
85
TVE was 65% or greater and was absent if it was less
than 65%. As per this classification, an increase in 65%
of the volume in a tunnel that was reamed to 7 mm
indicates a 2-mm increase in its diameter.
On the CT scans, femoral tunnels were divided into 3
assessment zones: zone I, which made up one-third of
the tunnel closest to the joint; zone II, which made up
the middle third of each tunnel; and zone III, which
made up the outermost third of the tunnel with respect
to the joint. Tunnel width was measured as the distance
between the widest marginal rim of each zone of the
femoral and tibial tunnel on both the coronal and
sagittal sections of CT scan using STARPACS. TVE was
defined as an increase in volume of greater than 65%
when femoral tunnel volume measured at a minimal
1-year follow-up was compared with the baseline
volume measured at 1 week postoperatively (Fig 2).
The tunnel volume measured at 1 week postoperatively
was used as the baseline measurement, which was
compared with the volume measured at a minimum of
1-year follow-up postoperatively. The intraobserver and
interobserver agreements, for tunnel diameter measurements as assessed with use of the k coefficient, were
excellent in every case (0.92 [95% confidence interval
(CI), 0.9 to 0.94] and 0.9 [95% CI, 0.89 to 0.91],
respectively).
Clinical Assessment
All patients were assessed for function using the International Knee Documentation Committee (IKDC) and
Lysholm scores both preoperatively (on admission)
and postoperatively (at 1 week and minimum 1-year
follow-up). Similarly, anterior knee laxity was quantitatively measured both preoperatively and postoperatively
using the KT-2000 arthrometer (MEDmetric, San Diego,
CA) with a standard manual force of 134 N and 3 mm of
side-to-side difference as the cut-off point. The rotational
stability was confirmed by the pivot-shift test.
Statistical Analysis
A power analysis was performed before commencing
the study to determine the sample size. In a pilot study of
10 cases, femoral tunnel width was 9.9 1.3 mm, 9.1 1.2 mm at 1 year follow-up and 1 week postoperatively.
For a power set at .8 and the a value at .05, the sample size
required was 34 patients in each group. The current study,
with 35 patients in group A and 34 patients in group B, had
a power of .81. The presence or absence of tunnel
enlargement (i.e., enlargement of >65% or 65%) in the
2 groups was compared using the c-square test for independence. Independent-samples t-test was used to
compare the amount of tunnel volume increase at 1 week
and minimum 1-year follow-up within the 2 groups.
Clinical scores of patients between preoperative and
postoperative evaluations were compared with use of the
Mann-Whitney U test. Statistical analysis was performed
86
D-H. HWANG ET AL.
Fig 2. CT scans in sagittal plane after 1-year
follow-up showing (A) femoral tunnel
enlargement in group A (press-fit technique
group) and (B) femoral tunnel enlargement
in group B (conventional technique group).
using the SPSS 18.0 (SPSS, Chicago, IL) software, and P <
.05 was considered to be significant.
Results
There was no significant difference between the 2
groups for mean age at the time of operation (P ¼ .47),
mean follow-up period (P ¼ .24), and sex ratio (P ¼ .46).
CT Measurements
The initial CT mean tunnel volume was 1,386 86
mm2 in group A and 1,325 99 mm2 in group B. In
group A (press-fit technique), the increment of the
mean femoral tunnel volume was 65%, and it was
71.5% in group B at 1-year follow-up. There was no
statistical difference for mean TVE in the femur
between the 2 groups (P ¼ .84).
The number of cases with an increase in femoral
tunnel volume in group A (20/35 cases, or 57%) was
not significantly different from the number of cases in
group B (23/34 cases, or 67%) that showed femoral
tunnel enlargement (Table 1).
Table 1. TVE of the Femoral Tunnel at Minimum Follow-up
of 1 Year in Both Groups
Volume increase
Yes (>65%)
No (65%)
Mean volume
increasing (%)
Mean volume
Group A (n ¼ 35)
Press-Fit
Technique
Group B (n ¼ 34)
Conventional
Technique
P (<.05)
20
15
64.9
23
11
71.5
.27
.27
.33
2,309 588
2,272 723
.84
Clinical Results
The median Lysholm score in group A increased from
63 (range, 59 to 68) preoperatively to 94 (range, 61 to
100) at last follow-up, which was not significantly
different from group B, in which the mean preoperative
score of 59 (range, 36 to 71) increased to the median
score of 92.5 (range, 52 to 100) postoperatively (P ¼
.73). Preoperative IKDC scores were abnormal in 21
patients and severely abnormal in 14 patients in group
A and abnormal in 22 patients and severely abnormal
in 12 patients in group B. This improved to normal in
16 patients and nearly normal in 19 patients in group A
and normal in 12 patients, nearly normal in 20 patients,
and abnormal in 2 patients in group B. There was no
significant differences in IKDC scores between the 2
groups (P ¼ .15). On KT-2000, 32 patients in group A
had less than 3 mm of side-to-side difference (average
1.5 1.3 mm), and in group B, 31 patients had less
than 3 mm (average, 1.8 1.5 mm) difference. In
group A, there were 3 cases of grade 1 pivot-shift test.
There were 4 cases of grade 1 pivot-shift test in group B.
No significant difference for pivot-shift test and anterior
laxity as measured on KT-2000 was found between the
2 groups (P ¼ .78 and .22).
Discussion
We hypothesized that the press-fit technique might
reduce the volume increase in femoral tunnel and
improve clinical scores by reducing synovial fluid
ingress within the bone tunnel. However, the results of
our CT-based study did not show any significant
difference in TVE when the press-fit technique was
compared with the conventional technique during
PRESS-FIT TECHNIQUE AFTER ACL RECONSTRUCTION
primary ACL reconstruction with hamstring autograft
at a minimum postoperative follow-up of 1 year.
Few studies have compared the incidence of TW in
groups where different fixation methods have been
used during ACL reconstruction. In a prospective
comparison of EndoButton (Smith & Nephew Endoscopy, Andover, MA) versus double cross-pins for
femoral fixation, Baumfeld et al.10 reported significantly greater femoral TW with the EndoButton fixation method on knee radiographs. However, Klein
et al.11 reported that significant TW occurs with
quadrupled hamstring autografts using femoral crosspin fixation. Hence, although the double cross-pin
method carries less risk of femoral TW than the
suspensory fixation method in ACL reconstruction, the
double cross-pin fixation method still carries a significant risk of TW. In the current study, although cross-pin
fixation was supplemented with the press-fit technique
during ACL reconstruction with auto hamstring
tendons, we could not find any significant reduction in
TVE in the femur.
The results of our study showed no significant
difference between the 2 techniques in terms of functional scores or rotation and anterior knee laxity when
assessed clinically at a minimum postoperative followup of 1 year. This finding is similar to that reported by
other studies that have compared the clinical outcome
of TW in 2 groups where 2 different femoral fixation
techniques were used in ACL reconstruction.10,12
Despite reporting significant increase in widening in
one fixation group compared with the other, these
studies showed no significant difference in the clinical
outcome between the groups in the short term.10,12
The extent of widening and zone of involvement in
the femoral tunnel may vary based on the type of
fixation used. In their CT evaluation of 34 patients with
ACL reconstruction with hamstring graft, Sabat et al.12
reported that femoral TW was highest at aperture,
medium at midway, and least at proximal and that
tunnel and femoral TW was significantly lesser in the
EndoButton group compared with the TransFix
(Arthrex, Naples, FL) group. In the current study,
femoral TW was greater compared with tibial TW in
both groups where a RigidFix fixation method was used
on the femur. In comparisons of each zone of this study,
however, the zone of involvement was significantly
greater in zone II (mid-zone) compared with zone I
(aperture) in group A and the widening was in the form
of a cavity. The zone of involvement was similar in both
zone I (aperture) and zone II (mid-zone) and the
widening was cylindrical in group B. This finding
implies that the press-fit technique used in group A
resulted in less TW at the tunnel aperture compared
with the mid-zone. This could be the result of a lesser
amount of a synovial fluid entering the tunnel due to
the presence of a snugly fitting graft in group A.
87
Previous reports have described other techniques of
tunnel preparation such as impaction drilling,13 drilling
and dilatation,14 and press-fit bone plug fixation
without hardware to improve fixation minimize TW;
only one study actually analyzed the effect of such
techniques on TW. Jagodzinski et al.,15 in a CT-based
study, reported significantly less widening of the proximal tibial tunnel with press-fit bone plug fixation after
ACL reconstruction at short-term follow-up. Although
their study did not analyze femoral tunnels, the findings concerning tibial TW were similar to those of our
study. However, the modification in operative technique of underdrilling used in the present study is
much simpler and less technically challenging compared with the other techniques described.
In this prospective, randomized study, we sought to
keep the 2 study groups as similar to each other as
possible by using the same surgeon, fixation method,
and postoperative rehabilitation regimen. However,
there are a few limitations to this study. First, this study
has a short-term follow-up for CT scan analysis and
evaluation of clinical outcome in the patients. Although
there was no significant difference in change in tunnel
volume and clinical outcome between the 2 techniques
in the current study, it is possible that this may become
evident if these patients are followed for longer durations. Second, the time period when CT analysis was
done postoperatively could also be an important factor
that can influence the result of the study. Fink et al.9
noted that the greatest amount of widening occurs
within the first 6 weeks, whereas Webster et al.2
showed that the radiographic tunnel width did not
significantly change between 4 months and 2 years. The
time point for the CT evaluation was 12 months in our
study. This could have influenced the results of our
study. Third, tunnel enlargement based on the type of
fixation method used was not part of the study. Fourth,
clinical outcome based on other activity scores, Tegner
score, or pivot shift was not used in this study. Finally,
the changes in incidence and extent of tunnel
enlargement over mid- and long-term follow-up are
not known and were not part of this study.
Conclusions
Between the press-fit and conventional techniques,
there were no significant differences in TVE and clinical
outcome at short-term follow-up.
References
1. Fauno P, Kaalund S. Tunnel widening after hamstring
anterior cruciate ligament reconstruction is influenced by
the type of graft fixation used: A prospective randomized
study. Arthroscopy 2005;21:1337-1341.
2. Webster KE, Feller JA, Hameister KA. Bone tunnel
enlargement following anterior cruciate ligament reconstruction: A randomized comparison of hamstring and
88
3.
4.
5.
6.
7.
8.
9.
D-H. HWANG ET AL.
patellar tendon grafts with 2-year follow-up. Knee Surg
Sports Traumatol Arthrosc 2001;9:86-91.
Höher J, Möller HD, Fu FH. Bone tunnel enlargement after
anterior cruciate ligament reconstruction: Fact or fiction?
Knee Surg Sports Traumatol Arthrosc 1998;6:231-240.
Buck DC, Smonian PT, Larson RV, Borrow J,
Nathanson DA. Timeline of tibial tunnel expansion after
single-incision hamstring anterior cruciate ligament
reconstruction. Arthroscopy 2004;20:34-36.
Darabos N, Haspl M, Moser C, Darabos A, Bartolek D,
Groenemeyer D. Intraarticular application of autologous
conditioned serum (ACS) reduces bone tunnel widening
after ACL reconstructive surgery in a randomized
controlled trial. Knee Surg Sports Traumatol Arthrosc
2011;(Suppl 1):S36-S46.
Buelow JU, Siebold R, Ellermann A. A prospective evaluation of tunnel enlargement in anterior cruciate ligament reconstruction with hamstrings: Extracortical versus
anatomical fixation. Knee Surg Sports Traumatol Arthrosc
2002;10:80-85.
Milano G, Mulas PD, Ziranu F, Piras S, Manunta A,
Fabbriciani C. Comparison between different femoral
fixation devices for ACL reconstruction with doubled
hamstring tendon graft: A biomechanical. Arthroscopy
2006;22:660-668.
Paessler HH, Mastrokalos DS. Anterior cruciate ligament
reconstruction using semitendinosus and gracilis tendons,
bone patellar tendon, or quadriceps tendon-graft with
press-fit fixation without hardware: A new and innovative procedure. Orthop Clin North Am 2003;34:49-64.
Fink C, Zapp M, Benedetto KP, Hackl W, Hoser C,
Rieger M. Tibial tunnel enlargement following anterior
10.
11.
12.
13.
14.
15.
cruciate ligament reconstruction with patellar tendon
autograft. Arthroscopy 2001;17:138-143.
Baumfeld JA, Diduch DR, Rubino LJ, et al. Tunnel
widening following anterior cruciate ligament reconstruction using hamstring autograft: A comparison
between double cross-pin and suspensory graft fixation. Knee Surg Sports Traumatol Arthrosc 2008;16:
1108-1113.
Klein JP, Lintner DM, Downs D, Vavrenka K. The incidence and significance of femoral tunnel widening after
quadrupled hamstring anterior cruciate ligament reconstruction using femoral cross-pin fixation. Arthroscopy
2003;19:470-476.
Sabat D, Kundu K, Arora S, Kumar V. Tunnel widening
after anterior cruciate ligament reconstruction: A
prospective randomized computed tomographyebased
study comparing 2 different femoral fixation methods for
hamstring graft. Arthroscopy 2011;27:776-783.
Siebold R, Kiss ZS, Morris HG. Effect of compaction drilling during ACL reconstruction with hamstrings on postoperative tunnel widening. Arch Orthop Trauma Surg
2008;128:461-468.
Dargel J, Schmidt-Wiethoff R, Brüggemann GP, Koebke J.
The effect of bone tunnel dilation versus extraction drilling on the initial fixation strength of press-fit anterior
cruciate ligament reconstruction. Arch Orthop Trauma Surg
2007;127:801-807.
Jagodzinski M, Geiges B, von Falck C, Knobloch K,
Haasper C, Brand J, Hankemeier S. Biodegradable screw
versus a press-fit bone plug fixation for hamstring anterior
cruciate ligament reconstruction: A prospective randomized study. Am J Sports Med 2010;38:501-508.