1. 315www.JournalofKneeSurgery.com
ABSTRACT: The rotational constraint characteristics
of 4 commercially available posterior-stabilized im-
plants (Triathlon, NexGen, PFC, and Genesis II) were
determined. The torques required to achieve up to
620° of internal and external rotation were quantified
at hyperextension, 0°, 15°, 60°, and 90° of flexion. The
effectoflubricationwastestedbycomparingthetorque
values generated in dry and serum environments. Re-
sults indicated that substantially higher torques were
generated in the dry environment, compared with the
serum environment. In extension, the principal design
features influencing the torques were interaction of the
box and post, as well as the insert anterior lip. In high-
er flexion angles, the amount of rollback, coupled with
the posterior lip geometry, most heavily influenced
the generation of torque. Comparison of the 4 designs
showed that the Triathlon generated the least torques
to 610° of rotation, whereas the PFC and Genesis II
were the most constrained.
[J Knee Surg. 2008;21:315-319.]
Introduction
Several studies have documented evidence of rotary
motion between the insert and baseplate of a total knee
arthroplasty (TKA).6,8,13-15,18
This motion, which can
cause backside wear, osteolysis, and loosening,15
results
from rotational stresses generated at the articulating sur-
face. The rotational constraint of a knee design thus plays
an important role. The more constrained the design, the
greater the potential for torques generated at the articulat-
ing surface to be transferred to the interface between the
polyethylene tibial insert and the metallic baseplate and
to the bone-implant interface. For example, Bradley et al4
found evidence of rotational backside wear in 71% of the
more constrained PFC posterior-stabilized trays; however,
only 31% of the PFC cruciate-retaining components dis-
played the same.
Articular surface constraint can be quantified experi-
mentally using a multi-axis test frame. Testing can be
performed dry or with a lubricant. The American Society
for Testing and Materials standard on constraint testing
specifies the use of a lubricant.2
However, results can vary
depending on which method is used.
The purpose of this study was to quantify the effects
of measuring constraint in a dry versus a lubricated envi-
ronment and to quantify rotational constraint of 4 TKA
designs in a lubricated environment. The Triathlon PS
(Stryker Orthopaedics, Mahwah, NJ), PFC Sigma Stabi-
lized (DePuy-Mitek, Warsaw, Ind), NexGen LPS (Zimmer
Inc, Warsaw, Ind), and Genesis II PS (Smith & Nephew,
Andover, Mass) designs were studied.
Materials and Methods
Test Samples
This study evaluated 3 samples of each knee design.
All samples represented medium-sized components with
the thinnest modular polyethylene insert offered by the
manufacturer.
Test Apparatus
All testing was performed on a multi-axis servo-
hydraulic test frame (MTS Bionix 858; MTS Corp, Eden
Rotational Constraint of Posterior-Stabilized
Total Knee Prostheses
Safia Bhimji, MS
Mark Kester, PhD
Thomas Schmalzried, MD
Ms Bhimji and Dr Kester are from Stryker Orthopaedics, Mahwah,
NJ; and Dr Schmalzried is from Orthopaedic Hospital, Los Angeles,
Calif.
Ms Bhimji and Dr Kester are employees of and Dr Schmalzried is a
consultant for Stryker Orthopaedics. However, no financial support was
received for this article.
Correspondence: Mark Kester, PhD, Stryker Orthopaedics, 325
Corporate Drive, Mahwah, NJ 07430.
Downloadedby:StrykerOrthopaedics.Copyrightedmaterial.

2. 316
THE JOURNAL OF KNEE SURGERY
October 2008 / Vol 21 No 4
Prairie, Minn). The test frame consisted of 2 actuators: a
coupled axial-torsional actuator oriented vertically to ap-
ply compressive loads and internal-external torques to the
components, and a side actuator, which is an axial actua-
tor oriented parallel to the floor to apply shear loads.
Test Structure
The femoral components and tibial baseplates were
rigidly mounted to stainless steel holding fixtures. The in-
serts were assembled to the baseplates according to surgi-
cal protocol. The insert-baseplate assembly was mounted
at a 0° slope to a bath, and the bath was then mounted to
a bearing table connected to the side actuator of the test
frame. The assembly was mounted so that the side actua-
tor applied a load in the anteroposterior (AP) direction of
the component assembly, and the bearing table allowed
unrestricted mediolateral motion. The femoral component
was mounted to the axial-torsional actuator via fixturing
that allowed flexion angle adjustment and unrestricted
varus-valgus rotation. Figure 1 presents a photograph of
the test setup. To initially align the femoral component
to the insert at each flexion angle, a 200-N compressive
force was applied while the femoral component was ro-
tated through 610°. The insert was allowed to float in the
AP and mediolateral directions to find its natural resting
position.
Loading Parameters
A 3800-N compressive load was applied between the
femoral component and insert while the femoral compo-
nent was rotated between 65° of internal-external rota-
tion at 5° per second for 5 cycles. The rotation was then
incrementally increased to 610°, 615°, and 620° for
5 cycles each. The insert was allowed to float in the AP
and mediolateral directions throughout testing. Testing
was performed in hyperextension and at 0°, 15°, 60°, and
90° of flexion. The hyperextension angle was design de-
pendent and represented the angle at which the femoral
component first contacted the anterior side of the insert
post.
Test Procedure
Each sample of the Triathlon PS design was first test-
ed in air at all flexion angles.Axial load and displacement,
side load and displacement, and torsional torque and angle
were recorded at 12.5 Hz throughout the loading process.
After testing had been completed in air, Alpha Calf
Fraction Serum (Hyclone Labs, Logan, Utah) diluted with
water (1 part serum to 6 parts water) was added to the
bath, with the fluid level completely covering the level of
tibiofemoral articulation. The temperature was heated to
37°C using a heater (approximately 30 min to heat), and
circulation was maintained using a pump. The data collec-
tion process was then repeated at each flexion angle for all
samples of all 4 knee designs.
Analysis
Articular surface constraint was measured at each
flexion angle as the amount of torque required to rotate
the femoral component on the insert from 220° to 120°.
Comparisons between designs and between test environ-
Figure 1. Test setup.
1
Figure 2. Rotational constraint at 15° flexion. Both air and
serum are depicted.
2
Figure 3. Rotational constraint at 90° flexion. Both air and
serum are depicted.
3
Downloadedby:StrykerOrthopaedics.Copyrightedmaterial.

3. 317
Rotational Constraint of Posterior-Stabilized Total Knee Prostheses
www.JournalofKneeSurgery.com
ments were made by overlaying plots of rotational torque
data as a function of femoral rotation. The peak-to-peak
torque values between 610° of rotation were also com-
puted and compared across designs.
Results
Effects of Test Environment on Constraint
Plots of rotational constraint for one sample of the Tri-
athlon knee design at 15° and 90° are displayed in Figures
2 and 3, respectively. The plots show results of testing in
both a dry and a lubricated environment. A positive angle
denotes internal rotation of the femoral component on the
insert, and a negative angle denotes external rotation. The
results demonstrate a substantial increase in the torque
required for rotation in the dry environment, compared
with the lubricated environment. This trend remained
consistent across the 620° of femoral rotation and across
flexion angles. On average, torques generated in the dry
environment were 45% larger than those generated in the
lubricated environment.
Effects of Knee Design on Constraint
Plots comparing rotational constraint across all 4 knee
designs in the lubricated environment can be found in Fig-
ure 4. Between 610° of rotation, Triathlon was the least
constrained followed by either NexGen or Genesis II, de-
pending on flexion angle. The PFC system demonstrated
the most constraint across all angles. Constraint charac-
teristics within 610° of rotation have clinical effects, as
this range represents rotations likely to occur in vivo at the
flexion angles tested.1,3,7,11,12
Beyond 610° of rotation, all designs demonstrated
higher torques. In internal rotation, NexGen was the least
constrained at hyperextension, 60°, and 90°; and Triathlon
was the least constrained at 0° and 15°. PFC and Gen-
esis II were the most constrained across all flexion angles.
In external rotation, NexGen was the least constrained at
hyperextension and at 15°; Triathlon was the least con-
strained at 0°; and Genesis II was the least constrained
at 60° and 90°. The PFC components did not reach 620°
of rotation in hyperextension due to impingement of
the femoral component on the anterior side of the tibial
post, which causes the rotational torque levels to exceed
a 35-Nm limit set to prevent damage to the inserts. The
NexGen components were not tested to 620° of rotation
at 90° of flexion. At this angle, the femoral component
articulated close to the posterior edge of the insert due
4B
4D
Figure 4. Rotational constraint across designs in a lubricat-
ed environment.
4A
4E
4C
Downloadedby:StrykerOrthopaedics.Copyrightedmaterial.

4. 318
THE JOURNAL OF KNEE SURGERY
October 2008 / Vol 21 No 4
to the very posterior location of the femoral condyles on
the insert. At 20° of rotation, the femoral condyles tended
to sublux off the insert and, in some cases, deformed the
posterior lip of the insert.
The peak-to-peak torque values generated between
610° of femoral rotation were computed for each flex-
ion angle. Results at 0° and 15° of flexion are shown in
Figures 5 and 6, respectively. The graphs display the av-
erage and standard deviation over the 3 samples tested.
Comparisons across designs were made using a 1-factor
analysis of variance (alpha = 0.05). Between 610° of ro-
tation, Triathlon demonstrated the least amount of torque
(differences were statistically significant with at least 1
design at each flexion angle). Consistent with their high-
er conformity, PFC and Genesis II generated the largest
amount of torque.
Discussion
The effects of friction between components contrib-
ute substantially to constraint characteristics. Our results
demonstrated that testing in a dry environment increased
the torques generated at the articulating surface by 45%
over those obtained in the lubricated environment. A con-
sistent test environment is necessary to compare results
across various loading conditions and implant designs.
It is recommended that a lubricant of dilute calf serum
be used to emulate the synovial environment of the knee
joint.
When comparing across the 4 designs tested in the lu-
bricated environment, Triathlon was the least constrained
between 610° of rotation, whereas PFC and Genesis II
were the most constrained. Resistance to rotation in the
PFC design was caused mainly by box-post impingement,
as previously reported.10
Impingement was greatest in hy-
perextension and full extension, where the box and the
post usually play a larger role in constraint. Anterior tibial
lips can also add to the resistance, as was demonstrated
by the Genesis II design. Rotational constraint can be re-
duced by increasing clearance around the tibial post, by
rounding the post with respect to the coronal dimension of
the intercondylar box, and by lowering anterior lip height.
Mobile tibial bearings are another design-dependent way
to reduce rotational constraint.
Whereas the tibial post and intercondylar box influ-
enced rotational constraint in extension and hyperexten-
sion, the tibiofemoral condylar geometry influenced ro-
tational torques in deeper flexion. The inability to test the
NexGen at 90° to 620° without damaging the insert bears
discussion. As the anatomic knee flexes, the femur rolls
posteriorly on the tibia while externally rotating on the
tibia. If this motion is to be permitted after arthroplasty,
the articular surfaces must accommodate rotation as the
cam and post drive rollback. The NexGen was not able to
accommodate as much rotation as the other designs tested
due to the posterior position of the femoral condyles at
90° of flexion and the posterior insert geometry. When the
femur was rotated to 620°, the NexGen tibial articulation
would not permit the rotation without subluxation and
damage to the posterior lip of the insert. To allow more
rotation, the NexGen would have to decrease the rollback,
extend the articular track, or remove material from the
posterior intercondylar aspect of the femoral component.
Design features such as rotary arcs in the articular surface
of the tibial insert can enhance the ability to accommodate
rotation.9
The arcs should extend far enough anteriorly
and posteriorly to allow rotation at the extremes of flexion
and extension.
There are shortcomings with this investigation. The
test rig used a fixed center of rotation for internal and
external rotations, which may not fully represent the in
situ variability. The maximum flexion angle tested was
90°, which is less than most clinical series indicate as
the maximum flexion angle. However, 90° covers joint
loading under the majority of ambulatory conditions, in-
Figure 6. Peak-to-peak torque, at 5° flexion for 610° rota-
tion. Differences were significant between Triathlon (Stryker
Orthopaedics) and all other designs and between PFC
(Depuy-Mitek) and NexGen (Zimmer Inc) and NexGen and
Genesis II (Smith & Nephew).
6
Figure 5. Peak-to-peak torque, at 0° flexion for 610° rota-
tion. Differences were significant between Triathlon (Stryker
Orthopaedics) and Genesis II (Smith & Nephew), and Tri-
athlon and PFC (Depuy-Mitek).
5
Downloadedby:StrykerOrthopaedics.Copyrightedmaterial.

5. 319
Rotational Constraint of Posterior-Stabilized Total Knee Prostheses
www.JournalofKneeSurgery.com
cluding stair climbing. The contact areas of the different
designs were not determined, which could affect the ser-
vice life of the implant.5
However, reduction of oxidation
appears to be more important than contact area for low
wear and implant survival.16
Essentially all manufacturers
now sterilize and package tibial polyethylene inserts in
an oxygen-free environment. Retrieval studies indicate a
favorable performance, with a dramatic reduction in pit-
ting and delamination.17
Some manufacturers offer tibial
inserts with no or negligible free radicals, although clini-
cal results are not yet available.
Conclusion
The rotational torques generated by 4 posterior-
stabilized knee implant systems were compared at dif-
ferent flexion angles. Substantially higher torques were
generated in air, compared with those generated in serum.
The posterior-stabilized box-post design and anterior tib-
ial lip affected rotational constraint in extension, whereas
the tibiofemoral articular geometry had a greater effect
on the ability to accommodate rotation in flexion. In ex-
tension, implants should design in clearance between the
cam and the post and avoid high anterior lips that can re-
sist rotation. In flexion, the tibial and femoral geometries
should accommodate rotation without damaging the pos-
terior poly or allowing femoral subluxation off the tibial
posterior lip.
References
1. Andriacchi TP, Dyrby CO, Johnson TS. The use of func-
tional analysis in evaluating knee kinematics. Clin Or-
thop. 2003;(410):44-53.
2. ASTM International. Standard Test Method for Determi-
nation of Total Knee Replacement. West Conshohocken,
PA: ASTM International; 2005.
3. Benson LC, DesJardins JD, Harman MK, LaBerge M.
Effect of stair descent loading on ultra-high molecular
weight polyethylene wear in a force-controlled knee sim-
ulator. Proc Inst Mech Eng [H]. 2002;216:409-418.
4. Bradley MP, Mayor MB, Collier JP. Differences in articu-
lar track area of posterior-stabilized and cruciate-retaining
retrieved total knee implants. Orthopedics. 2004;27:1273-
1278.
5. Buechel FF Sr, Buechel FF Jr, Pappas MJ, Dalessio J.
Twenty-year evaluation of the New Jersey LCS Rotating
Platform Knee Replacement. J Knee Surg. 2002;15:84-89.
6. Conditt MA, Ismaily SK, Alexander JW, Noble PC. Back-
side wear of modular ultra-high molecular weight polyeth-
ylene tibial inserts. J Bone Joint Surg Am. 2004;86:1031-
1037.
7. Costigan PA, Deluzio KJ, Wyss UP. Knee and hip kinetics
during normal stair climbing. Gait Posture. 2002;16:31-
37.
8. Engh GA, Lounici S, Rao AR, Collier MB. In vivo dete-
rioration of tibial baseplate locking mechanisms in con-
temporary modular total knee components. J Bone Joint
Surg Am. 2001;83:1660-1665.
9. Harwin SF, Greene KA, Hitt K. Early experience with a
new total knee implant: Maximizing range of motion and
function with gender-specific sizing. Surg Technol Int.
2007;16:199-205.
10. Klein R, Serpe L, Kester MA, et al. Rotational constraint
in posterior-stabilized total knee prostheses. Clin Orthop.
2003;(410):82-89.
11. Lafortune MA, Cavanagh PR, Sommer HJ III, Kalenak A.
Three-dimensional kinematics of the human knee during
walking. J Biomech. 1992;25:347-357.
12. Li G, Zayontz S, DeFrate LE, Most E, Suggs JF, Rubash
HE. Kinematics of the knee at high flexion angles: An in
vitro investigation. J Orthop Res. 2004;22:90-95.
13. Mikulak SA, Mahoney OM, dela Rosa MA, Schmalzried
TP. Loosening and osteolysis with the press-fit condylar
posterior-cruciate-substituting total knee replacement. J
Bone Joint Surg Am. 2001;83:398-403.
14. Parks NL, Engh GA, Topoleski LD, Emperado J. The
Coventry Award. Modular tibial insert micromotion. A
concern with contemporary knee implants. Clin Orthop.
1998;(356):10-15.
15. Rao AR, Engh GA, Collier MB, Lounici S. Tibial inter-
face wear in retrieved total knee components and correla-
tions with modular insert motion. J Bone Joint Surg Am.
2002;84:1849-1855.
16. Ritter MA. Direct compression molded polyethyl-
ene for total hip and knee replacements. Clin Orthop.
2001;(393):94-100.
17. Silva M, Kabbash CA, Tiberi JV III, et al. Surface dam-
age on open box posterior-stabilized polyethylene tibial
inserts. Clin Orthop. 2003;(416):135-144.
18. Wasielewski RC. The causes of insert backside wear in to-
tal knee arthroplasty. Clin Orthop. 2002;(404):232-246.
Downloadedby:StrykerOrthopaedics.Copyrightedmaterial.