The American Journal of Sports

Medicine
http://ajs.sagepub.com/

Comparison of Hamstring Tendon and Patellar Tendon Grafts in Anterior Cruciate Ligament
Reconstruction in a Nationwide Population-Based Cohort Study: Results From the Danish Registry of
Knee Ligament Reconstruction
Lene Rahr-Wagner, Theis Muncholm Thillemann, Alma Becic Pedersen and Martin Lind
Am J Sports Med 2014 42: 278 originally published online November 25, 2013
DOI: 10.1177/0363546513509220
The online version of this article can be found at:
http://ajs.sagepub.com/content/42/2/278

Published by:
http://www.sagepublications.com

On behalf of:

American Orthopaedic Society for Sports Medicine

Additional services and information for The American Journal of Sports Medicine can be found at:
Email Alerts: http://ajs.sagepub.com/cgi/alerts
Subscriptions: http://ajs.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Jan 31, 2014

OnlineFirst Version of Record - Nov 25, 2013
What is This?

Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

5-in-5

Comparison of Hamstring Tendon and

Patellar Tendon Grafts in Anterior Cruciate
Ligament Reconstruction in a Nationwide
Population-Based Cohort Study
Results From the Danish Registry of
Knee Ligament Reconstruction
Lene Rahr-Wagner,*yz MD, Theis Muncholm Thillemann,y MD, PhD,
Alma Becic Pedersen,z MD, PhD, and Martin Lind,y MD, PhD
Investigation performed at the Departments of Clinical Epidemiology and Orthopaedic Surgery,
Aarhus University Hospital, Aarhus, Denmark
Background: The choice of graft for anterior cruciate ligament reconstruction (ACLR) remains controversial, and despite numerous studies, there is still an ongoing debate on this topic. The 2 most widely used grafts are the hamstring tendon and patellar
tendon.
Hypothesis: In this study, we hypothesized that the revision rate after primary ACLR is greater when using hamstring tendon
grafts compared with patellar tendon grafts.
Study Design: Cohort study; Level of evidence, 2.
Methods: From the nationwide population-based Danish Knee Ligament Reconstruction Registry, we identified all primary ACLR
procedures (n = 13,647) performed in Denmark between July 2005 and December 2011. The end point was revision ACLR. As
other end points, we used objective measurements and patient-reported outcome scores. Revision rates and relative risk estimates for revision ACLR were calculated using Cox multiple regression.
Results: The use of hamstring tendon grafts increased from 68% in 2005 to 85% in 2011. The cumulative revision rates for hamstring tendon grafts at 1 and 5 years were 0.65% (95% confidence interval [CI], 0.51%-0.82%) and 4.45% (95% CI, 3.94%5.01%), respectively. For patellar tendon grafts, the revision rate was 0.16% (95% CI, 0.05%-0.50%) at 1 year and 3.03%
(95% CI, 2.27%-4.05%) at 5 years. The adjusted overall relative risk of revision surgery in the hamstring tendon group compared
with the patellar tendon group was 1.41 (95% CI, 1.03-1.92), and the adjusted relative risk of undergoing revision surgery performed after 1 and 5 years was 3.82 (95% CI, 1.20-12.2) and 1.90 (95% CI, 0.43-8.40), respectively.
Conclusion: In this population-based study, the use of hamstring tendon grafts in ACLR was associated with an increased risk of
revision compared with patellar tendon grafts, in particular during the first year after surgery. These results demonstrate that both
hamstring and patellar tendon grafts reveal good results after ACLR and suggest that graft selection should be based on an individual evaluation of patient demands and graft morbidity.
Keywords: ACL; revision; patellar tendon; bonepatellar tendonbone; semitendinosus/gracilis; hamstring; graft choice

between countries is seen. In Denmark and Sweden, in

recent years, the HT graft has been used in about 84% of
cases.17 In the United States, based on the Multicenter
Orthopaedic Outcome Network (MOON) database, the most
widely used graft types were doubled HT autografts (44%)
and PT autografts (42%), whereas the Norwegian National
Knee Ligament Registry (NKLR) reported that the HT graft
was used in 60% and the PT graft in 37% of patients.26
There are advantages and disadvantages of each graft,
and usually, the choice of graft depends on the surgeons

Despite numerous studies comparing grafts in anterior

cruciate ligament reconstruction (ACLR), the choice of
graft remains controversial. The most frequently used
grafts for ACLR are hamstring tendon (HT) and patellar
tendon (PT) autografts, although a great difference

The American Journal of Sports Medicine, Vol. 42, No. 2

DOI: 10.1177/0363546513509220
2013 The Author(s)

278
Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

Vol. 42, No. 2, 2014

Registry-Based Comparison of Graft Types in ACLR

personal preferences and an assessment in relation to

patient characteristics. An optimal graft is fast healing,
is strong, and restores knee joint stability with low morbidity. An animal study suggested that PT grafts heal more
rapidly because of bone-to-bone healing compared with
bone-to-tendon healing.31 Rapid healing is essential to
ACLR, as it allows for earlier and more accelerated rehabilitation and thus a more rapid return to previous sports
performance. Therefore, PT grafts have been suggested
for younger and more active patients younger than 20
years of age because failure rates seem to be higher in
this age group.24,27 Also, use of the PT graft has increased
because of easy-to-use interference screw fixation.14,31
However, many studies have shown that ACLR using PT
autografts is associated with well-documented donor site
morbidity, such as anterior knee pain, pain at kneeling,
extensor strength deficit, and patellofemoral osteoarthritis.5,12,21,30,36 Therefore, in recent years, use of the HT
graft has been popularized because of its lower rate of
donor site morbidity.30 Also, this is supported by randomized controlled trials indicating equal stability and subjective outcomes when comparing HT to PT grafts.5,10,22,28,30
However, the HT graft is associated with deficits in knee
flexion strength and internal tibial rotation strength.11,28
An abundance of randomized controlled trials, including
a Cochrane review, have reported slight or no differences
in postoperative stability and the revision rate after primary ACLR between HT and PT grafts; thus, no consensus
has been reached yet.5,10,11,22,28,30 Hence, the objective of
this nationwide population-based cohort study was to compare the revision rate and clinical outcomes between HT
and PT grafts used in primary ACLR. We hypothesized
a lower revision rate for PT grafts compared with HT
grafts because of better graft incorporation in PT grafts.

MATERIALS AND METHODS

Study Setting
Denmark has a population of 5.5 million people. The
National Health Service provides tax-supported health
care to all Danish residents, allowing free access to hospital care at medical, surgical, and psychiatric departments
as well as general practitioner visits. Patients with acute
medical conditions are admitted for specialist treatment
at public hospitals. Private hospitals are also accessible
in Denmark, and they also have reimbursement agreements with the Danish State.

Data Sources
The Danish Knee Ligament Reconstruction Register
(DKRR). The DKRR is a nationwide population-based

279

clinical database that was established on July 1, 2005

with the purpose of improving the monitoring and quality
of both primary and revision ACLR in Denmark.23,34 All
private (n = 27) and public (n = 24) hospitals report to
this register, and registration is compulsory according to
Declaration Number 459 of June 2006.9 The rate of registrations in the DKRR has been more than 85% in the
past 3 years.32,35
Detailed preoperative, intraoperative, and 1-year
follow-up data are recorded by the operating surgeon using
a standardized form and a secure Internet portal.23 Furthermore, patients independently report subjective scores
on knee function using the Knee injury and Osteoarthritis
Outcome Score (KOOS)37 and the Tegner functional
score.41 These data are web-recorded by the patient before
surgery and 1 year after surgery. The KOOS ranges from
0 to 100, and the Tegner score ranges from 1 to 10, with
higher scores representing better results. The KOOS4
can also be calculated as a validated average of 4 KOOS
subscales: quality of life, sport, pain, and symptoms.13
The Civil Registration System (CRS). A unique 10-digit
personal identification number is given to all Danish citizens
at the date of birth. The CRS records information on changes
in the vital status of all Danish citizens including changes in
address, date of emigration, and date of death since 1968.33
Because this personal identification number is consistent
through all Danish registries, precise individual-level data
linkage between all Danish registries is possible. The CRS
was also used to obtain complete follow-up information on
all patients.

Study Population
In total, we identified 13,760 primary ACLR procedures
using either the HT or the PT graft in 13,565 patients in
the period from July 2005 to December 2011. We excluded
22 patients who had a status date before the operation
date, which is the end of the study period (December 31,
2011), and 1 patient who had the revision surgery date registered before the primary surgery date. In 40 operated
knees, the patients lived in Greenland, and in 50 operated
knees, the patients were from other countries who had visited Denmark; these 2 groups could not be properly followed up and were therefore excluded. Thus, 13,647
primary ACLR procedures were included in the final
analysis.

Exposure
In this study, we investigated the results of the use of
either 4-stranded semitendinosus/gracilis grafts, defined
as HT grafts, or PT grafts on our outcome measures. In
total, we identified 14,755 operated knees from the

*Address correspondence to Lene Rahr-Wagner, MD, Department of Clinical Epidemiology, Aarhus University Hospital, Olof Palmes Alle 43-45, 8200
Aarhus N, Denmark (e-mail: lrw@dce.au.dk).
y
Department of Orthopaedic Surgery, Aarhus University Hospital, Aarhus, Denmark.
z
Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark.
One or more of the authors has declared the following potential conflict of interest or source of funding: This study received financial support from the
Danish Rheumatism Association, the Elisabeth and Karl Ejnar Nis-Hanssens Scholarship, and the Aase and Ejnar Danielsens Foundation.

Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

280

Rahr-Wagner et al

The American Journal of Sports Medicine

DKRR: 11,676 HT grafts and 1971 PT grafts. The remaining 1108 grafts used were 4-stranded semitendinosus
grafts alone (1.7%), iliotibial tract grafts (1.6%), doublebundle HT grafts (1.4%), 2-stranded semitendinosus grafts
alone (1.2%), quadriceps tendon grafts (0.9%), allografts
(0.2%), and others (0.5%). In this study we only included
ACLR procedures using the HT or the PT graft; hence,
our final exposed group was 13,647 reconstructed knees.

Outcomes
The primary outcome was revision ACLR, defined as
another ACLR performed in the same knee as the primary
ACLR. The follow-up period started on the date of the primary ACLR and ended on the date of revision ACLR if revision occurred, at the time of death, or on the status date,
which is the end of the study period (December 31, 2011),
whichever came first. Furthermore, we described the cause
of revision surgery.
The secondary outcomes were parameters of objective
knee stability in terms of instrumented side-to-side differences (eg, Rolimeter [Aircast, Boca Raton, Florida] or KT-1000
arthrometer [MEDmetric Corp, San Diego, California]) and
pivot-shift scores. The pivot-shift test is a dynamic and passive test of the knee that measures rotational stability of the
ACL. The pivot-shift test is graded by a 4-point scale from
normal (0), glide (1), clunk (2), and gross (3).19 The pivotshift data were divided into one group with negative pivotshift results (normal) (n = 4674) and the other group with
positive pivot-shift results (glide, clunk, and gross) (n =
918). The instrumented side-to-side difference measures
the difference in sagittal stability between the operated
knee and the healthy knee at the 1-year control visit. The
side-to-side difference was measured manually as the maximal translation at 25 of flexion using either the KT-1000
arthrometer or Rollimeter. Patients were categorized as
having a difference of \2 mm (n = 4096) or .2 mm (n =
790). Only patients with no prior ACLR of the contralateral
knee were included in this analysis. Hence, 428 knees were
excluded from this analysis.
Finally, we used patient-reported outcomes, the KOOS
and the Tegner score, at 1 year postoperatively, if reported.
The KOOS and Tegner score are validated, subjective
patient-reported outcomes, calculated according to published standards.37,41 The KOOS4 is a validated patientreported outcome computed from the 4 most responsive
KOOS subscores13: symptoms, pain, sport, and quality of
life. Preoperative and postoperative KOOS and Tegner
scores were available in 4516 of 13,647 patients (33%)
and in 3614 of 13,647 patients (26%), respectively.

Confounding Factors
We obtained data at the time of surgery from the DKRR on
sex, age (20 and .20 years), cartilage damage .1 cm2
(no/yes or missing), operated meniscal damage (yes/no or
missing), prior surgery of the knee (yes or no), and activity
leading to the primary ACL rupture (sport vs nonsport).
All these variables were used as confounders in, or estimates of, the relative risk (RR).

Statistics
By the Kaplan-Meier method, we estimated graft survival
and the cumulative revision probability at 1 and 5 years
follow-up in the HT and PT groups. We used Cox regression
analyses to compare the revision risk after primary ACLR
among patients with HT and PT grafts. We computed the
hazard ratios as a measure for RR with a 95% confidence
interval (95% CI) for patients with HT grafts compared
with PT grafts, while adjusting for potentially confounding
factors as mentioned above. The assumption of the Cox
regression model was assessed with the use of log-log plots
and Schoenfeld residuals and was found suitable.
Further, using the logistic regression analysis by
adjusting for potential confounders, we calculated the
odds ratio of having a positive pivot-shift test result in
the HT group compared with the PT group and the risk
of having more than a 2-mm side-to-side difference
between the operated and healthy knee in the HT group
compared with the PT group. For the side-to-side difference, we excluded 41 patients (82 knees) who had both
knees operated on as well as registered side-to-side differences and 347 knees that had a previous ACLR performed
on the contralateral knee.
The mean values of the KOOS and Tegner score preoperatively and 1 year postoperatively for the patients in the
HT and PT groups were compared using the Student t test
if data were normally distributed based on Q-Q plots; otherwise, the Wilcoxon rank-sum test was used. All reported
P values have a significant value of .05.
All statistical analyses were computed using Stata version 12 (StataCorp, College Station, Texas). The study was
approved by the Danish Data Protection Agency.

RESULTS
Patient characteristics are outlined in Table 1. The most
commonly used graft types in the DKRR besides HT and
PT autografts are 4-stranded semitendinosus grafts (1.7%)
and iliotibial tract autografts (1.6%). Allografts were only
used in 0.20% of the reconstructions. The inclusion criterion
for this study was the use of either the HT graft or the PT
graft. The use of HT grafts increased from 68% of all graft
types in 2005 to 85% of all graft types in 2011. The mean
follow-up was 3.01 years (95% CI, 2.99-3.09).
In the HT group, 312 of 11,676 knees underwent revision, whereas 47 of 1971 knees underwent revision in the
PT group. The Kaplan-Meier cumulative revision rates of
primary ACLR for HT grafts were 0.65% (95% CI, 0.51%0.82%) at 1 year and 4.45% (95% CI, 3.94%-5.01%) at 5
years. For PT grafts, the cumulative revision rates were
0.16% (95% CI, 0.05%-0.50%) at 1 year and 3.03% (95%
CI, 2.27%-4.05%) at 5 years (Figure 1).
The crude overall RR for revision in the HT group compared with the PT group was 1.50 (95% CI, 1.11-2.04). The
overall adjusted RR for revision was 1.41 (95% CI, 1.031.92), and the adjusted RRs for revision after 1 and 5 years
were 3.82 (95% CI, 1.20-12.20) and 1.90 (95% CI, 0.438.40), respectively.

Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

Graft Choice, n (%)

Male sex
Age 20 y
Prior surgery of the knee
Sports activity leading
to the tear
No cartilage lesion
Meniscal treatment

1309
416
588
1590

(66)
(21)
(30)
(81)

1626 (83)
766 (39)

Hamstring Tendon
(n = 11,676)
6840
3149
3018
9472

(58)
(27)
(26)
(81)

9244 (79)
4469 (38)

New trauma was ascribed as the most frequent reason

for revision surgery and accounted for 41% of revisions in
the HT group and 47% in the PT group. This was followed
by suboptimal placement of the graft in the femur, which
accounted for 16% in the HT group and 19% in the PT
group (Table 2).
In the HT group, 39% had pivot-shift test results recorded
1 year postoperatively, and in the PT group, 52% had pivotshift test results recorded 1 year postoperatively. Sixteen percent of the knees in the HT group had positive pivot-shift test
results, and 19% of the knees in the PT group had positive
pivot-shift test results. The HT group was associated with
a slightly but significantly lower risk of positive pivot-shift
test results compared with the PT group, with an adjusted
odds ratio of 0.81 (95% CI, 0.68-0.96).
In the HT group, 36% had side-to-side differences
recorded 1 year postoperatively, and in the PT group,
44% had side-to-side differences recorded 1 year postoperatively. One year postoperatively, 16% of the knees in the
HT group had side-to-side differences of .2 mm, and 19%
of the knees in the PT group had side-to-side differences of
.2 mm. The odds ratio of having a side-to-side difference of
.2 mm in the HT group compared with the PT group was
0.82 (95% CI, 0.68-1.01).
Preoperatively, the KOOS and Tegner scores were comparable in the 2 groups (Table 3). One year postoperatively, the Tegner scores were 4.9 (95% CI, 4.9-5.0) for
the HT graft and 4.7 (95% CI, 4.6-4.9) for the PT graft
(P \ .05). Further, 1 year postoperatively, the KOOS
sports subscores were 62.6 (95% CI, 61.7-63.5) for HT
grafts and 58.0 (95% CI, 56.2-59.7) for PT grafts (P \ .05).

DISCUSSION
This is the first nationwide registry-based cohort study
presenting the results comparing the use of HT grafts
with PT grafts in primary ACLR. Prior studies relied on
small sample sizes and are often based on 1 surgeon or
patients in 1 department, which may not present the
everyday clinical praxis.2,10 Because the risk for revision
after ACLR is relatively low, a large sample size is necessary for a valid estimation of revision risks. A nationwide
registry-based study would fulfill these criteria. Until
now, such a study has not been conducted.

Hamstring Tendon

95% CI

Patellar Tendon

Patellar Tendon
(n = 1971)

95% CI

281

.04

Cumulative ACL revisions

TABLE 1
Patient Characteristics

.06

Registry-Based Comparison of Graft Types in ACLR

.02

Vol. 42, No. 2, 2014

0
Number at risk:
PT graft (n = 1971)
HT graft (n = 11,676)

2
4
Time of follow-up (years)
1622
7214

1072
3434

6
181
552

Figure 1. Kaplan-Meier cumulative revision curve of primary

anterior cruciate ligament reconstruction using either a hamstring tendon graft or a patellar tendon graft.
TABLE 2
Causes of Revision Surgery Recorded in the Danish Knee
Ligament Reconstruction Registry
Graft Choice, %
Cause
New trauma
Tunnel widening
Suboptimal placement of the graft
in the tibia
Suboptimal placement of the graft
in the femur
Infection
Unknown reason for instability
Other ligament failure
Other
Totala

Hamstring
Tendon

Patellar
Tendon

41.0
2.3
8.3

46.8
0
8.5

15.6

19.2

3.6
22.2
4.3
2.7
100

0
12.7
8.5
4.3
100

Ten knees had no explanation as to the cause of revision.

In this study, we found a minor difference in the overall

risk of revision between HT grafts and PT grafts when
adjusting for relevant confounders. However, we found
almost 4 times an increased risk of early revision at 1-year
follow-up in the HT group compared with the PT group.
Objective stability measures showed a minor decreased risk
of positive pivot-shift test results in the HT group compared
with the PT group. Further, the Tegner scores and KOOS
subscores of sport and activities of daily living at 1 year postoperatively were slightly higher in the HT group.
The 3.8-fold increased RR of revision at 1 year after
ACLR using HT grafts compared with PT grafts is a concern. However, the absolute risk of revision is small: less
than 1% for both groups. This short-term increased risk
of revision could represent failures that occur because
patients return to sport too quickly; hence, a solution
may be to prolong the rehabilitation period before returning to sport. This is also supported by the better KOOS

Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

282

Rahr-Wagner et al

The American Journal of Sports Medicine

TABLE 3
Preoperative and 1-Year Postoperative KOOS and Tegner Scores for Primary ACLR in the PT and HT Groupsa
HT Group
Preoperative KOOS (n = 4516, 33%), n (%)
Symptoms
Pain
ADL
Sport
Quality of life
KOOS4
Preoperative Tegner score (n = 4516, 33%)
Postoperative KOOS (n = 3677, 27%), n (%)
Symptoms
Pain
ADL
Sport
Quality of life
KOOS4
Postoperative Tegner score (n = 3677, 27%)

3726
70.8
71.0
78.2
38.3
39.5
54.9
3.0
2968
77.1
83.6
89.0
62.6
59.6
70.8
4.9

(32)
(70.2-71.3)
(70.5-71.6)
(77.6-78.8)
(37.5-39.2)
(38.9-40.0)
(54.4-55.4)
(3.0-3.1)
(25)
(76.4-77.7)
(83.0-84.1)
(88.6-89.5)
(61.7-63.5)
(58.8-60.3)
(70.1-70.9)
(4.9-5.0)

PT Group
790
71.7
71.7
78.8
38.8
39.3
55.4
3.0
709
78.3
83.4
88.5
58.0
59.0
69.7
4.7

(40)
(70.6-72.8)
(70.5-72.9)
(77.5-80.0)
(37.0-40.6)
(38.2-40.5)
(54.3-56.5)
(2.8-3.1)
(36)
(77.2-79.4)
(82.4-84.4)
(87.6-89.5)
(56.2-59.7)
(57.5-60.5)
(68.5-70.9)
(4.6-4.9)

P Valueb

NS
NS
NS
NS
NS
NS
NS
NS
NS
\.05
\.05
NS
NS
\.05

a
Values are presented as the median (95% confidence interval) unless otherwise indicated. ACLR, anterior cruciate ligament reconstruction; ADL, activities of daily living; HT, hamstring tendon; KOOS, Knee injury and Osteoarthritis Outcome Score; KOOS4, quality of life,
sport, pain, and symptoms of the KOOS; NS, not significant; PT, patellar tendon.
b
P values for the difference in the KOOS and Tegner scores between the 2 groups were calculated using the Wilcoxon rank-sum test and
Student t test.

sport subscore in the HT group, as this may indicate that

patients feel ready for sports activities. Also, it is well
known that PT grafts heal faster because of bone-to-bone
healing.31 This may also explain the higher risk of revision
in the HT group at 1 year after surgery because this soft
tissue graft with bone-to-tendon healing needs more time
to obtain maximal strength.
Zebis et al42 showed a reduced risk for noncontact ACL
injuries in female athletes who had undergone neuromuscular training. They found that it seems essential to have
an adequate function of the HT muscles to protect the
ACL from injuries. These findings could also support our
data, suggesting that patients who undergo ACLR with
HT grafts need more focused rehabilitation and a thorough
evaluation before they return to sport.
Many randomized controlled trials have studied the difference between HT and PT grafts.10,11,22,28 Contrary to our
study, many recent studies have shown comparable revision
rates between HT and PT grafts.2,5,20,25,40 Supporting our
data is a meta-analysis from 2003 in which Freedman
et al12 compared different studies using either HT or PT
grafts in ACLR with a minimum of 2-year follow-up. Similar
to our study, they showed a significantly higher graft failure
rate when using HT grafts (4.9%) at a mean follow-up of 34
months compared with PT grafts (1.9%) at a mean followup of 46 months.12 Most of the studies included in this
meta-analysis were from the early and mid-1990s. Also,
a recent study by Maletis et al29 supports our data of a higher
revision rate in the HT group compared with the PT group.
Despite this abundance of literature, it has not been possible to show any clear advantage of one method over the
other, and it seems that one ideal graft for all patients
does not exist. Hence, recently, a Cochrane review that
included 19 randomized or quasi-randomized controlled

trials with a minimum of 2 years follow-up concluded that

there was no evidence to draw definite conclusions on differences between the 2 graft types for long-term follow-up in
terms of subjective scores and objective measurements.30
In contrast to our study, they did conclude that the PT
group is more likely to have statistically significant stable
knees and is associated with more anterior knee pain.30
Knee stability after ACLR is of major importance to
return to normal physical activity and to avoid the early
onset of osteoarthritis due to an unstable knee.18 Therefore,
postoperative stability, measured by the pivot shift and
side-to-side difference4 of the knee, is an essential parameter
when evaluating outcomes after ACLR. The risk of instability
in terms of positive pivot-shift test results was slightly lower
in the HT group compared with the PT group, whereas sideto-side instrumented stability measures were comparable
between the groups. In our results, we divided pivot shift
into 2 groups only (positive and negative). This was done
because the pivot-shift test is a very subjective evaluation
and because a more specific classification of pivot shift would
result in groups with too small numbers of patients and even
smaller numbers of outcomes, despite our large sample size.
Therefore, we chose this subdivision to have a valid analysis
and conclusions. Several studies have reported comparable
clinical assessment or instrumented stability findings for different graft types,10,15,20,22,28 with no preferences between PT
and HT autografts. Our objective stability results may indicate that the HT group, which also reports higher KOOS
sport subscores, may have a higher activity level, which corresponds with our finding of less positive pivot shift in the
HT group and thus a possible higher risk of reinjury.
Patient-reported outcome measures, the Tegner score
and KOOS subscores of sport and activities of daily living,
were higher for HT grafts at 1 year postoperatively

Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

Vol. 42, No. 2, 2014

Registry-Based Comparison of Graft Types in ACLR

compared with PT grafts; however, the difference is small

and clinically not important. Different aspects influence
the strength of the graft, such as graft size, tunnel position,
bone density, and graft fixation.25,27 As previously mentioned, PT grafts have the advantage of bone-to-bone healing in contrast to HT grafts, which do not have a bone block
and need tendon-to-bone healing and thus prolonged time
for graft incorporation.8,16 Furthermore, most surgeons
standardized screw fixation for PT grafts often uses interference screws,7 which is in contrast to the great variety of
fixation methods used for the HT graft. Hence, a lot of factors are a concern when optimizing graft fixation.
Our results do not necessarily support PT grafts to be
superior to HT grafts but rather that clinicians should continue to select the graft type according to patient-reported
factors including graft morbidity, patient activity level (ie,
high-performance athletes), instability of the knee, and
those with jobs demanding kneeling. Further, a prolonged
rehabilitation period should be considered for athletes
treated with an HT graft.

Strength and Limitations

This study was based on data from a national clinical registry. Thus, there are several strengths and limitations.
The fact that the DKRR is a large national database is
a clear strength for the quality of the data. Also, because
of the unrestricted and free access to health care in Denmark, the DKRR provides an unselected study population.
Also, data are readily available, and the DKRR gives
access to a large amount of data that otherwise would
not be possible to obtain, reduces the risk of bias, and minimizes the cost of research.39 In addition, clinical databases
make timely and early dissemination of information on
specific clinical issues possible. Further, the DKRR has
the potential for extensive linkage on an individual level
to other important databases because of the unique personal identification number assigned to all Danish citizens,
which enables the possibility of individual measurements.
Further, the cohort design provides the researcher with
the possibility of adjusting for relevant confounders.
Also, data from a national clinical registry encounter several limitations. One problem is the completeness of the
data and patients compliance of online subjective patient
registrations. The rate of registration in the database was
greater than 85% for the past 3 years, which we consider
as acceptable.35 However, the low number of patients
reporting preoperative and postoperative patient-reported
outcome scores is a concern. This could lead to information
bias if missing data on patient-reported outcomes are associated with both graft choices and later revision. Because
the data collection is prospective and registration of primary
ACLR is independent of registration of later revision, the
risk of information bias is very limited. Also, because there
is no reason to believe that patients completing the subjective scores are fundamentally different than those lost to
follow-up, the risk of bias is considered to be low. Nevertheless, a validity study has been performed recently, showing
no difference in patient-reported outcome scores between
responders and nonresponders in the DKRR.35 We therefore

283

consider our subjective data as valid despite the limited

completeness of data. A substantial number of data were
missing at the 1-year postoperative clinical control visit.
The challenge of obtaining a high percentage of patients
at clinical follow-up is substantial when dealing with
a national cohort. Again, because of the prospective collection of data, it is unlikely that a lack of these data is associated with both graft choice and later revision. Using
revision ACLR as the outcome in this study is looking at
only 1 definition of failure after ACLR. Some patients are
not willing to undergo revision surgery, accepting a reduction in their activity level and chronic knee instability. In
other cases, surgeons do not find patients suitable for revision surgery. Thus, the revision rate may be a conservative
measure for the real number of ACL patients with treatment failure. However, there is no reason to believe that
patient and surgeon willingness for revision differs in the
HT and PT groups. In this multivariate analysis, we
included important confounders that have previously been
associated with the risk of revision surgery.1,3,6,12 We chose
not to include fixation technique as a confounding factor
because graft choice is associated with fixation and not
the other way around. Thus, fixation is part of the causal
path between the graft (exposure) and revision (outcome)
and should not be included in the model as a confounder.38
Although we adjusted for a number of potential confounding
factors, our study, like all observational studies, may suffer
from unmeasured and residual confounding. For example,
data on graft size, smoking habit, alcohol consumption,
medication use, sports activity during the follow-up period,
and occupation were not available in the DKRR.

CONCLUSION
This is the first nationwide registry-based cohort study of
more than 13,000 ACL knees to present results comparing
HT and PT grafts in primary ACLR. This study demonstrated
a minimally higher overall revision rate using HT grafts.
However, HT grafts were associated with a substantially
increased risk of early revision at 1-year follow-up. Thus,
the recovery and training period before returning to sport
may be of special importance for patients treated with HT
autografts in primary ACLR. We conclude that both HT
and PT grafts reveal good results after primary ACLR and
recommend that graft selection should be based on an individual evaluation of patient demands and graft morbidity.

ACKNOWLEDGMENT
The authors acknowledge the help and support of Frank
Mehnert at the Department of Clinical Epidemiology and
the kind secretarial help of Anne Haagen Hjelm at the
Department of Clinical Epidemiology.

REFERENCES
1. Ageberg E, Forssblad M, Herbertsson P, et al. Sex differences in
patient-reported outcomes after anterior cruciate ligament

Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014

284

2.

3.

4.

5.

6.

7.

8.
9.

10.

11.

12.

13.

14.

15.

16.

17.
18.

19.
20.

21.

22.

Rahr-Wagner et al

The American Journal of Sports Medicine

reconstruction: data from the Swedish knee ligament register. Am J

Sports Med. 2010;38(7):1334-1342.
Aglietti P, Giron F, Buzzi R, et al. Anterior cruciate ligament reconstruction: bone-patellar tendon-bone compared with double
semitendinosus and gracilis tendon grafts. A prospective, randomized clinical trial. J Bone Joint Surg Am. 2004;86(10):2143-2155.
Alentorn-Geli E, Lajara F, Samitier G, et al. The transtibial versus the
anteromedial portal technique in the arthroscopic bone-patellar
tendon-bone anterior cruciate ligament reconstruction. Knee Surg
Sports Traumatol Arthrosc. 2010;18(8):1013-1037.
Ayeni OR, Chahal M, Tran MN, et al. Pivot shift as an outcome measure for ACL reconstruction: a systematic review. Knee Surg Sports
Traumatol Arthrosc. 2012;20(4):767-777.
Barenius B, Nordlander M, Ponzer S, et al. Quality of life and clinical outcome after anterior cruciate ligament reconstruction using patellar tendon graft or quadrupled semitendinosus graft: an 8-year follow-up of
a randomized controlled trial. Am J Sports Med. 2010;38(8):1533-1541.
Barrett AM, Craft JA, Replogle WH, et al. Anterior cruciate ligament
graft failure: a comparison of graft type based on age and Tegner
activity level. Am J Sports Med. 2011;39(10):2194-2198.
Biau DJ, Katsahian S, Kartus J, et al. Patellar tendon versus hamstring tendon autografts for reconstructing the anterior cruciate ligament: a meta-analysis based on individual patient data. Am J Sports
Med. 2009;37(12):2470-2478.
Blickenstaff KR, Grana WA, Egle D. Analysis of a semitendinosus
autograft in a rabbit model. Am J Sports Med. 1997;25(4):554-559.
Danish Health and Medicines Authority. Declaration on Approval of
Nationwide and Regional Clinical Quality Databases [in Danish]. Statutory Order 459. June 2006.
Ejerhed L, Kartus J, Sernert N, et al. Patellar tendon or semitendinosus tendon autografts for anterior cruciate ligament reconstruction?
A prospective randomized study with a two-year follow-up. Am J
Sports Med. 2003;31(1):19-25.
Feller JA, Webster KE. A randomized comparison of patellar tendon
and hamstring tendon anterior cruciate ligament reconstruction. Am
J Sports Med. 2003;31(4):564-573.
Freedman KB, DAmato MJ, Nedeff DD, et al. Arthroscopic anterior
cruciate ligament reconstruction: a metaanalysis comparing patellar
tendon and hamstring tendon autografts. Am J Sports Med.
2003;31(1):2-11.
Frobell RB, Roos EM, Roos HP, et al. A randomized trial of treatment
for acute anterior cruciate ligament tears. N Engl J Med.
2010;363(4):331-342.
Fu FH, Bennett CH, Lattermann C, et al. Current trends in anterior
cruciate ligament reconstruction, part 1: biology and biomechanics
of reconstruction. Am J Sports Med. 1999;27(6):821-830.
Gifstad T, Sole A, Strand T, et al. Long-term follow-up of patellar tendon grafts or hamstring tendon grafts in endoscopic ACL reconstructions. Knee Surg Sports Traumatol Arthrosc. 2013;21(3):576-583.
Grana WA, Egle DM, Mahnken R, et al. An analysis of autograft fixation after anterior cruciate ligament reconstruction in a rabbit model.
Am J Sports Med. 1994;22(3):344-351.
Granan LP, Forssblad M, Lind M, et al. The Scandinavian ACL registries
2004-2007: baseline epidemiology. Acta Orthop. 2009;80(5):563-567.
Jonsson H, Riklund-Ahlstrom K, Lind J. Positive pivot shift after ACL
reconstruction predicts later osteoarthrosis: 63 patients followed 5-9
years after surgery. Acta Orthop Scand. 2004;75(5):594-599.
Kopf S, Kauert R, Halfpaap J, et al. A new quantitative method for pivot
shift grading. Knee Surg Sports Traumatol Arthrosc. 2012;20(4):718-723.
Laxdal G, Kartus J, Hansson L, et al. A prospective randomized comparison of bone-patellar tendon-bone and hamstring grafts for anterior cruciate ligament reconstruction. Arthroscopy. 2005;21(1):34-42.
Li S, Chen Y, Lin Z, et al. A systematic review of randomized controlled clinical trials comparing hamstring autografts versus bonepatellar tendon-bone autografts for the reconstruction of the anterior
cruciate ligament. Arch Orthop Trauma Surg. 2012;132(9):1287-1297.
Liden M, Ejerhed L, Sernert N, et al. Patellar tendon or semitendinosus tendon autografts for anterior cruciate ligament reconstruction:

23.

24.

25.

26.

27.

28.

29.
30.

31.

32.

33.
34.

35.

36.

37.

38.

39.

40.

41.
42.

a prospective, randomized study with a 7-year follow-up. Am J

Sports Med. 2007;35(5):740-748.
Lind M, Menhert F, Pedersen AB. The first results from the Danish
ACL reconstruction registry: epidemiologic and 2 year follow-up
results from 5,818 knee ligament reconstructions. Knee Surg Sports
Traumatol Arthrosc. 2009;17(2):117-124.
Lind M, Menhert F, Pedersen AB. Incidence and outcome after revision anterior cruciate ligament reconstruction: results from the Danish registry for knee ligament reconstructions. Am J Sports Med.
2012;40(7):1551-1557.
Magnussen RA, Carey JL, Spindler KP. Does autograft choice determine intermediate-term outcome of ACL reconstruction? Knee Surg
Sports Traumatol Arthrosc. 2011;19(3):462-472.
Magnussen RA, Granan LP, Dunn WR, et al. Cross-cultural comparison
of patients undergoing ACL reconstruction in the United States and
Norway. Knee Surg Sports Traumatol Arthrosc. 2010;18(1):98-105.
Magnussen RA, Lawrence JT, West RL, et al. Graft size and patient age
are predictors of early revision after anterior cruciate ligament reconstruction with hamstring autograft. Arthroscopy. 2012;28(4):526-531.
Maletis GB, Cameron SL, Tengan JJ, et al. A prospective randomized
study of anterior cruciate ligament reconstruction: a comparison of
patellar tendon and quadruple-strand semitendinosus/gracilis tendons fixed with bioabsorbable interference screws. Am J Sports
Med. 2007;35(3):384-394.
Maletis GB, Incasio MCS, Desmod JL, et al. Reconstruction of the
anterior cruciate ligament. Bone Joint J. 2013;95-B(5):623-628.
Mohtadi NG, Chan DS, Dainty KN, et al. Patellar tendon versus hamstring tendon autograft for anterior cruciate ligament rupture in
adults. Cochrane Database Syst Rev. 2011;9:CD005960.
Papageorgiou CD, Ma CB, Abramowitch SD, et al. A multidisciplinary
study of the healing of an intraarticular anterior cruciate ligament
graft in a goat model. Am J Sports Med. 2001;29(5):620-626.
Pedersen AB, Graversen HP, Klareskov B, et al. Dansk Korsbands
Rekonstruktions Register: Arsrapport 2011. Aarhus, Denmark: SUNTRYK; 2012.
Pedersen CB. The Danish Civil Registration System. Scand J Public
Health. 2011;39(7 Suppl):22-25.
Rahr-Wagner L, Thillemann TM, Pedersen AB, Lind MC. Increased
risk of revision after anteromedial compared with transtibial drilling
of the femoral tunnel during primary anterior cruciate ligament reconstruction: results from the Danish Knee Ligament Reconstruction
Register. Arthroscopy. 2013;29(1):98-105.
Rahr-Wagner L, Thillemann TM, Pedersen AB, Lind MC. Validation of
14,500 operated knees registered in the Danish Knee Ligament
Reconstruction Register: registration completeness and validity of
key variables. Clin Epidemiol. 2013;5(1):219-228.
Roe J, Pinczewski LA, Russell VJ, et al. A 7-year follow-up of patellar
tendon and hamstring tendon grafts for arthroscopic anterior cruciate
ligament reconstruction: differences and similarities. Am J Sports
Med. 2005;33(9):1337-1345.
Roos EM, Toksvig-Larsen S. Knee injury and Osteoarthritis Outcome
Score (KOOS): validation and comparison to the WOMAC in total
knee replacement. Health Qual Life Outcomes. 2003;1:17.
Rothman KJ, Greenland S, Lash T. Causal diagrams. In: Modern Epidemiology. 3rd ed. Philadelphia: Lippincott Williams & Wilkins;
2008:183-200.
Sorensen HT, Sabroe S, Olsen J. A framework for evaluation of secondary data sources for epidemiological research. Int J Epidemiol.
1996;25(2):435-442.
Taylor DC, DeBerardino TM, Nelson BJ, et al. Patellar tendon versus
hamstring tendon autografts for anterior cruciate ligament reconstruction: a randomized controlled trial using similar femoral and tibial
fixation methods. Am J Sports Med. 2009;37(10):1946-1957.
Tegner Y, Lysholm J. Rating systems in the evaluation of knee ligament injuries. Clin Orthop Relat Res. 1985;198:43-49.
Zebis MK, Bencke J, Andersen LL, et al. The effects of neuromuscular
training on knee joint motor control during sidecutting in female elite
soccer and handball players. Clin J Sport Med. 2008;18(4):329-337.

For reprints and permission queries, please visit SAGEs Web site at http://www.sagepub.com/journalsPermissions.nav
Downloaded from ajs.sagepub.com at Philadelphia College of Osteopathic Medicine on February 2, 2014