Interference Screw Position and Hamstring Graft Location

for Anterior Cruciate Ligament Reconstruction

Peter T. Simonian, M.D., Patrick S. Sussmann, M.S., Todd H. Baldini, M.S.,

Heber C. Crockett, M.D., and Thomas L. Wickiewicz, M.D.

Summary: Anterior cruciate ligament reconstruction with hamstring tendon graft

and interference screw fixation has recently been considered. Concerns for the use
of interference screws with soft tissue grafts include damage to the graft during
screw insertion, decreased fixation strength, and a decrease in the bone-tendon
contact area for healing within the tunnel when the screw is placed in an eccentric
position. This last concern could be addressed by placing the interference screw
centrally between the four limbs of the hamstring graft. The purpose of this study
was to determine the mode of failure, the pullout force, and graft slippage before
graft fixation failure of hamstring tendons fixed with an interference screw
positioned eccentrically in relation to the hamstring tendons verses an interference
screw positioned centrally between the four graft limbs. The semitendinosus and
gracilis tendons were harvested from six, fresh cadaveric specimens. Each tendon
was divided into two segments of equal length. Both the semitendinosus and
gracilis tendon segments were looped to form four strands. The specimens were
then fixed with a bioabsorbable interference screw in the two different positions
and pulled from a standardized polyurethane foam. All tendons in both groups
failed by pulling out from between the interference screw and tunnel, regardless of
the screw position. No tendon was cut by the screw in either group. There was no
significant difference between the forces required to produce specific amounts of
graft slippage between the two fixation techniques tested. There was no significant
difference between the average total slippage at maximum pullout, 11.8 mm for the
screw placed in the eccentric position and 13.7 mm for the screw placed in the
central position. The maximum pullout force averaged 265.3 N for the screw
placed in the eccentric position, and 244.7 N for the screw placed in the central
position; these values were not significantly different. Placement of the interfer-
ence screw in the central position did not compromise strength and it improves
graft contact within the bone tunnel. Interference screw fixation, when applied
against a bone plug, has been shown to consistently have a pullout force of more
than 400 N. Key Words: ACL—Hamstring—Interference screw—Biodegradable—
Central—Eccentric.

A nterior cruciate ligament (ACL) reconstruction

has become an increasingly successful surgery.
There is much debate on the source of autologous
From The Sports Medicine Service, The Hospital for Special
Surgery, Affıliated with The New York Hospital-Cornell University
Medical College, New York, New York, U.S.A.
Address correspondence and reprint requests to Peter T. Simo-
nian, M.D., University of Washington Medical Center, Department
of Orthopaedic Surgery, Box 356500, Seattle, WA 98195, U.S.A.
r 1998 by the Arthroscopy Association of North America
0749-8063/98/1405-1783$3.00/0
tissue used for reconstruction. The two most common
sources for autologous graft include the central third of
the patellar tendon, or the semitendinosus and gracilis
tendons. Excellent functional results have been re-
ported using either source.1-9
Graft fixation is a potential concern with all ACL
reconstructions especially with the use of hamstring
tendons. Interference screw fixation, when applied
against a bone plug, has been shown to have superior
strength over other methods, including screw and

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 14, No 5 (July-August), 1998: pp 459–464 459
460 P. T. SIMONIAN ET AL.
washer post fixation,10 staple fixation, and button
fixation.11 A separate study did not demonstrate these
same differences in strength between these fixation
devices.12 Regardless of the pullout strength, a distinct
advantage of the interference screw is graft fixation
within the bone tunnel closer to the tunnel’s articular
aperture, thus shortening the length of the graft
between the two points of fixation.13 Interference
screw fixation of hamstring tendons would also pre-
clude the need for synthetic length augmentation of the
graft tissue with suture or tape.14
Because of these advantages, interference screw
fixation has recently been used for hamstring fixation
with successful short-term clinical results.15 Potential
concerns with interference fixation for hamstring ten-
dons include damage to the graft with screw insertion,
decreased pullout strength, and a decrease in the
bone-tendon contact area for healing in the tunnel
when the screw is placed in the eccentric position. This
last concern could be addressed by placing the interfer-
ence screw centrally between the four limbs of the
hamstring graft rather than eccentric screw placement.
Most studies have focused on peak pullout force
when evaluating different fixation devices. While this
is important, significant graft slippage before this peak
load would represent a clinical failure. Therefore, in
addition to reporting peak loads before failure, this
study focuses on the specific force required to incremen-
tally displace the graft from the fixation device and the
total slippage that occurred before reaching the peak
pullout force.
The purpose of this study is to determine the mode
of failure, the pullout force, and the graft slippage
before graft fixation failure of hamstring tendons fixed
with an interference screw placed eccentrically in
relation to the hamstring tendons verses an interfer-
ence screw placed centrally between the four graft
limbs.
MATERIALS AND METHODS
Specimens
Six fresh, cadaveric knee specimens, were used for
study. The mean age of the donors was 72 years (range,
54 to 83 years). The semitendinosus and gracilis
tendons were harvested from each specimen. These
tendons were removed distally from the pes anserinus
and stripped proximally from their respective muscle
bellies. All residual muscle was removed from the
tendons. Matched pairs of each tendon segment were
then made for comparison of the two fixation tech-
niques. Each tendon was divided into two equal length
segments. A No. 5, nonabsorbable, polyester suture
(Ethibond; Ethicon Inc, Somerville, NJ) was placed
into the ends of all the tendon segments in Bunnel
fashion to provide traction. Both the semitendinosus
and gracilis tendon segments were looped to form four
strands. The other matched segments of these two
tendons from the same specimen were also quadrupled
for direct comparison. This resulted in six matched
pairs of four-strand hamstring constructs. The proxi-
mal and distal tendon segments from each specimen
were randomized between the two testing groups.
Tendon Fixation
A 9 ⫻ 25 mm bioabsorbable, cannulated interfer-
ence screw (Bioscrew; Linvatec, Largo, FL) was
placed centrally between the four tendon limbs of one
group using a specially designed device that assures
central placement of the guide wire between the four
tendon limbs. This central fixation method was de-
signed to maximize the tendon-bone contact surface
area between each of the four hamstring limbs (Fig
1A). This group was compared with the other matched
group of tendons fixed with an identical interference
screw placed eccentrically between the four limbs of
the graft as a single group and the tunnel wall (Fig 1B).
A 10-mm tunnel hole was made in a new polyurethane
foam block for each pullout test.
Polyurethane Foam as a Model for Pullout Testing
The variations in apparent density, trabeculae orien-
tation, and mechanical properties of cancellous bone
within and between specimens are large.16 Human
bone is classified as cancellous if it has an apparent
density in the range of 0.09 to 1.26 g/cm3 or a porosity
between 30% and 90%.17,18 The compressive and
tensile strengths of cancellous bone have been shown
to be related to apparent density and thus vary by more
than two orders of magnitude over a typical range of
densities.19,20 To study the effects of specific param-
eters related to screw position variation independently
from variations in bone material properties and geom-
etry, rigid unicellular foam was selected as a test
material.
Each pullout test was performed in a new uniform
cube (2.5 cm3 ) made of unicellular polyurethane foam
(Last-O-Foam FR 6712; General Plastics, Tacoma,
WA) which has an apparent density of 0.19 gm/cm3;
this density is at the lower end of the range for
cancellous bone to produce a worst case scenario.
Polyurethane foam has been used previously in bone
screw design research.21-24 Cellular solids have been
extensively studied as engineering materials, and their
 INTERFERENCE SCREW AND HAMSTRING GRAFT LOCATION 461

FIGURE 1. A 9 ⫻ 25 mm bioabsorbable, cannulated interference screw was placed centrally between the four tendon limbs of one group using
a specially designed device that assures central placement of the guide wire between the four tendon limbs. This central fixation method was
designed to maximize the tendon-bone contact surface area between each of the four hamstring limbs. (B) This group was compared with the
other matched group of tendons fixed with an identical interference screw placed eccentrically between the four limbs of the graft as a single
group and the tunnel wall.

properties have been shown to correspond to the
mechanical behavior of cancellous bone.17
Loading Method
Before mechanical testing, the specimens were
thawed at room temperature and rehydrated in warm
saline solution. The specimens were also continually
hydrated with a saline atomizer before testing. All the
test comparisons were between graft segments from
the same specimen.
The polyurethane blocks were held to the actuator of
the materials testing machine with a clamp and the
closed loop of semitendinosus and gracilis tendons
were attached to the load cell of the materials testing
machine with a metal loop.
The load magnitude and actuator displacement were
monitored throughout the test by a computer data
acquisition system at a rate of 10 Hz. The specimens
were first preloaded to 10 N and then loaded at a rate of
45 N/s25 to failure on a uniaxial materials test system
(MTS Systems Corporation, Minneapolis, MN).
Data Analysis
Paired, two-tailed, t tests were used to evaluate the
differences in matched hamstring pair pullout force
with centrally positioned and eccentrically positioned
interference screws at 1-mm increments of graft slip-
page and at the maximum pullout force. The mean
pullout force and standard deviations were calculated
at each millimeter increment of displacement and at
the maximal pullout force for both the centrally posi-
tioned and eccentrically positioned interference screw
pairs.
Additionally, paired, two-tailed, t tests were also
used to evaluate the differences in the amount of graft
slippage at the maximum pullout force for the centrally
placed and the eccentrically placed interference screw
groups. The mean graft slippage and standard devia-
tions at maximum pullout force were calculated.
RESULTS
Mechanism of Failure
All tendons in both groups failed by pulling out
from between the interference screw and tunnel,
regardless of the screw position. There were no
failures through the substance of the tendons. No
tendon was cut by the screw in either group.
Graft Slippage Verses Pullout Forces
Graft slippage from around the interference screw
was the mode of failure in every trial. Although most
studies have focused on peak pullout force, slippage
occurring before this peak value is just as relevant. A
minimum of 6 mm of slippage occurred in every trial
before reaching the peak pullout force. There was no
significant difference between the forces required to
produce specific amounts of slippage between the two
fixation techniques tested (Table 1 and Fig 2). The
approximate increase in pullout force per millimeter
graft slippage was 30 N in both groups.
There was no significant difference between the
total slippage at maximum pullout between the two
fixation techniques tested; for the screw placed in the
eccentric position, mean slippage was 11.8 mm, and
462 P. T. SIMONIAN ET AL.

TABLE 1. Forces Required to Produce Graft Slippage sons were made between hamstring grafts from the
same specimen.
Mean Eccentric Mean Central
Graft Slippage Screw (SD) Screw (SD) Because of the success and general strength advan-
tages of interference screw fixation of a bone plug,
1 mm 38.9 N (9.9 N) 36.6 N (5.4 N) some have started using this same fixation technique
2 mm 69.5 N (11.6 N) 59.7 N (9.8 N)
3 mm 102.8 N (12.3 N) 89.6 N (19.7 N) on soft tissue grafts with encouraging early clinical
4 mm 134.8 N (18.4 N) 116.6 N (24.3 N) results.15 However, the interference screws used in the
5 mm 164.0 N (20.8 N) 141.5 N (31.8 N) present study were made of a different material with
6 mm 190 N (28.2 N) 158.9 N (38.2 N) different geometry.
Maximum pullout force 265.3 N (47.7 N) 244.7 N (60.5 N)
Slippage at maximum Potential concerns with interference fixation for
pullout force 11.8 mm (3.1 mm) 13.7 mm (3.6 mm) hamstring tendons include damage to the graft on
screw insertion, decreased pullout strength, and a
decrease in the bone-tendon contact area for healing in
for the screw placed in the central position mean the tunnel when the screw is positioned eccentrically.
slippage was 13.7 mm (Table 1). The first concern of screw damage to the hamstring
Maximal Pullout Force tendons was not evident in the present study.
There was no significant difference between the Interference screw fixation with a bone plug has
maximum pullout force of the two fixation techniques been shown to have strength superior to alternative
tested. For the screw placed in the eccentric position, fixation devices in most studies10,11; however, this
the maximum pullout force averaged 265.3 N, and for finding has not been conclusively substantiated.12
the screw placed in the central position the maximum Matthews et al.12 evaluated four fixation methods of
pullout force averaged 244.7 N (Table 1 and Fig 2). patellar tendon-bone grafts in cadaveric knees. Group I
and II grafts were fixed with interference screws;
group III and IV grafts were anchored with nonabsorb-
DISCUSSION
able sutures tied over a screw and washer, post
Although many other factors affect the success of fixation. The mean force to failure in groups I, II, III,
ACL reconstruction, the method of surgical fixation is and IV were 435.0 N, 458.2 N, 454.2 N, and 415.8 N,
the major factor influencing the graft’s mechanical respectively. There was no significant difference in the
properties in the immediate postoperative period.11 An force to failure among the four methods of fixation.
attempt was made in the present study to limit the Neither of the soft tissue interference screw techniques
number of fixation variables. The size and type of utilized in the present study had peak pullout forces
interference screw, the tunnel diameter, and the foam comparable to these values.
density were kept consistent; additionally, compari- Paschal et al.10 compared biomechanical characteris-

FIGURE 2. Graphic representa-

tion of the mean force required
to cause 1, 2, 3, 4, 5, and 6 mm
of four-stranded hamstring graft
slippage from around an inter-
ference screw positioned in two
different configurations. The let-
ter E represents the eccentri-
cally positioned interference
screw, and the letter C repre-
sents the centrally positioned
interference screw for each
matched pair. The approximate
increase in pullout force per
millimeter of graft slippage was
30 N in both groups. The last
data pair represent the mean
maximal pullout force before
graft failure. There were no sta-
tistical differences between any
of these data pairs.
 INTERFERENCE SCREW AND HAMSTRING GRAFT LOCATION
tics of interference screw fixation to sutures tied over a
cancellous screw and washer using bone–patellar
tendon–bone grafts in porcine anatomic specimen
knees. Maximum pullout strength and displacement of
the bone graft at 110 N of force were compared in a
progressive load-to-failure test. Interference screw
fixation demonstrated statistically significantly higher
mean ultimate failure loads at 535 N compared with
post fixation at 309 N. Interference fixation also
showed statistically significantly less displacement of
the bone graft at 110 N of force, 0.32 mm versus 2.21
mm. All failures occurred at the fixation site. The data
from the present study again show that neither of the
soft tissue interference screw techniques used were
comparable to the maximum pullout strength of inter-
ference screw with porcine bone–patellar tendon–bone
grafts reported by Paschal et al.; however, our values
approximated the values of their cancellous screw and
washer, post fixation. There was also considerably
more graft slippage at approximately 110 N of force in
the present study measuring 3 to 4 mm when compared
with the porcine bone–patellar tendon–bone graft fixed
with an interference screw at 0.32 mm reported by
Paschal et al.; however, our slippage results approxi-
mated the values of their cancellous screw and washer,
post fixation at 2.21 mm.
All hamstring grafts failed by slipping from around
the interference screw, regardless of screw position.
The amount of tendon slippage before reaching the
maximal pullout force value would represent a clinical
failure at more than 10 mm of graft slippage in both
groups. As mentioned, the maximum pullout strength
was substantially less than reported for interference
screws in conjunction with bone plugs.10,12 However,
the maximal pullout strength and tendon slippage at
approximately 110 N were nearly comparable to the
clinically proven screw and washer, post fixation
technique.10 The strength deficiency could be ad-
dressed by augmenting soft tissue interference screw
fixation with another device like a suture button. This
augmentation would likely increase the strength and
decrease the slippage. This combination would allow
the potential advantage of interference screw tunnel
fixation without a compromise in strength. Regardless
of the pullout strength, a distinct advantage of the
interference screw is graft fixation within the bone
tunnel closer to its articular aperture, thus shortening
the length of the graft between the two points of
fixation.
The last concern of decreased bone-tendon contact
resulting from an eccentrically positioned interference
screw was addressed by positioning the screw cen-
463
trally between the four limbs of the hamstring graft.
Placing the screw in this central position did not affect
the amount of graft slippage or maximum pullout
strength compared with the eccentric screw position. A
potential problem of this technique is the propensity
for the graft limbs to rotate as the screw is turned; this
can be minimized by applying significant traction on
all the graft limbs during screw insertion.
A potential limitation of this study is the use of
polyurethane foam rather than actual bone specimens.
This synthetic substance provided a more uniform
density than human bone specimens. However, in this
discussion comparisons were made with other studies
that used actual bone specimens.
Acknowledgment: The authors thank Alexander Fernan-
dez for his assistance with the anatomical dissections.
REFERENCES
1. Cho KO. Reconstruction of the anterior cruciate ligament by
semitendinosus tenodesis. J Bone Joint Surg Am 1975;57:
608-612.
2. Clancy WG, Nelson DA, Reider B, Narechania R. ACL
reconstruction using one third of the patellar ligament, aug-
mented by extra-articular tendon transfers. J Bone Joint Surg
Am 1982;64:352-359.
3. Lipscomb AB, Johnston RK, Snyder RB, Brothers JC. Second-
ary reconstruction of the anterior cruciate ligament in athletes
using the semitendinosus tendon. Am J Sports Med 1979;7:
81-84.
4. O’Brien SJ, Warren RF, Pavlov H, Panariello R, Wickiewicz
TL. Reconstruction of the chronically insufficient ACL with the
central third of the patellar ligament. J Bone Joint Surg Am
1991;73:278-286.
5. Sgaglione NA, Warren RF, Wickewicz TL, Gold DA, Panari-
ello RA. Primary repair with semitendinosus tendon augmenta-
tion of acute cruciate ligament injuries. Am J Sports Med
1990;18:64-73.
6. Shelbourne DK, Al E. Anterior cruciate ligament injury:
Evaluation of intraarticular reconstruction of acute tears with-
out repair. Am J Sports Med 1990;18:484-489.
7. Bach BR Jr, Jones GT, Sweet FA, Hager CA. Arthroscopy-
assisted anterior cruciate ligament reconstruction using patellar
tendon substitution. Two- to four-year follow-up results. Am J
Sports Med 1994;22:758-767.
8. Maeda A, Shino K, Horibe S, Nakata K, Buccafusca G.
Anterior cruciate ligament reconstruction with multistranded
autogenous semitendinosus tendon. Am J Sports Med 1996;24:
504-509.
9. Marder RA, Raskind JR, Carroll M. Prospective evaluation of
arthroscopically assisted anterior cruciate ligament reconstruc-
tion. Patellar tendon versus semitendinosus and gracilis ten-
dons. Am J Sports Med 1991;19:478-484.
10. Paschal SO, Seemann MD, Ashman RB, Allard RN, Montgom-
ery JB. Interference fixation versus postfixation of bone–
patellar tendon–bone grafts for anterior cruciate ligament
reconstruction. A biomechanical comparative study in porcine
knees. Clin Orthop 1994;281-287.
11. Kurosaka M, Yoshiya S, Andrish JT. A biomechanical compari-
son of different surgical techniques of graft fixation in anterior
cruciate ligament reconstruction. Am J Sports Med 1987;15:
225-229.
464 P. T. SIMONIAN ET AL.
12. Matthews LS, Lawrence SJ, Yahiro MA, Sinclair MR. Fixation
strengths of patellar tendon-bone grafts. Arthroscopy 1993;9:
76-81.
13. Ishibashi Y, Rudy T, Kim H, Fu F, Woo S. The effect of the
anterior cruciate ligament graft fixation level on knee stability.
Arthroscopy 1995;11:373.
14. Barrett GR, Papendick L, Miller C. Endobutton button endo-
scopic fixation technique in anterior cruciate ligament recon-
struction. Arthroscopy 1995;11:340-343.
15. Pinczewski L, Corry I, Clingeleffer A, Webb J. Endoscopic
ACL reconstruction comparing 4 strand hamstring tendon with
patellar tendon autograft—Two year results. Presented at the
23rd Annual Meeting of the American Orthopaedic Society of
Sports Medicine, Sun Valley, ID, 1997.
16. Sell P, Collins M, Dove J. Pedicle screws: Axial pull-out
strength in the lumbar spine. Spine 1988;13:1075-1076.
17. Gibson L, Ashby M. Cancellous bone. In: Cellular solids:
Structure and properties. New York: Pergamon, 1988;316-
331.
18. Hayes WC. Biomechanics of cortical and trabecular bone:
Implications for assessment of fracture risk. In: Mow VC,
Hayes WC, eds. Basic orthopaedic biomechanics. New York:
Raven, 1991;93-142.
19. Carter DR, Spengler DM. Mechanical properties and composi-
tion of cortical bone. Clin Orthop 1978;135:192-217.
20. Carter DR, Hayes WC. Bone compressive strength: The in-
fluence of density and strain rate. Science 1976;194:1174-1176.
21. De Coster TA, Heetderks DB, Downey DJ, Ferries JS, Jones W.
Optimizing bone screw pullout force. J Orthop Trauma 1990;4:
169-174.
22. Finlay JB, Harada I, Bourne RB, Rorabeck CH, Hardie R, Scott
MA. Analysis of the pullout strength of screws and pegs used to
secure tibial components following total knee arthroplasty.
Clin Orthop 1989;247:220-231.
23. Hearn TC, Schatzker J, Wolfson N. Extraction strength of
cannulated cancellous bone screws. J Orthop Trauma 1993;7:
138-141.
24. Shaw JA. A biomechanical comparison of scaphoid screws. J
Hand Surg Am 1987;12:347-353.
25. Liu SH, Kabo JM, Osti L. Biomechanics of two types of
bone-tendon-bone graft for ACL reconstruction. J Bone Joint
Surg Br 1995;77:232-235.