The American Journal of Sports Medicine http://ajs.sagepub.com/

The Effect of Trochlear Dysplasia on Patellofemoral Biomechanics: A Cadaveric Study With Simulated Trochlear Deformities Annemieke Van Haver, Karel De Roo, Matthieu De Beule, Luc Labey, Patrick De Baets, David Dejour, Tom Claessens and Peter Verdonk Am J Sports Med published online March 4, 2015 DOI: 10.1177/0363546515572143 The online version of this article can be found at: http://ajs.sagepub.com/content/early/2015/03/04/0363546515572143

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:

P> OnlineFirst Version of Record - Mar 4, 2015 What is This?

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

AJSM PreView, published on March 4, 2015 as doi:10.1177/0363546515572143

The Effect of Trochlear Dysplasia on Patellofemoral Biomechanics A Cadaveric Study With Simulated Trochlear Deformities Annemieke Van Haver,*yz§ MSc, PhD, Karel De Roo,|| MD, Matthieu De Beule,{ MSc, PhD, Luc Labey,# MSc, PhD, Prof., Patrick De Baets,z MSc, PhD, Prof., David Dejour,** MD, Tom Claessens,y MSc, PhD, and Peter Verdonk,||yy MD, PhD, Prof. Investigation performed at Ghent University, Ghent, Belgium Background: Trochlear dysplasia appears in different geometrical variations. The Dejour classification is widely used to grade the severity of trochlear dysplasia and to decide on treatment. Purpose: To investigate the effect of trochlear dysplasia on patellofemoral biomechanics and to determine if different types of trochlear dysplasia have different effects on patellofemoral biomechanics. Study Design: Controlled laboratory study. Methods: Trochlear dysplasia was simulated in 4 cadaveric knees by replacing the native cadaveric trochlea with different types of custom-made trochlear implants, manufactured with 3-dimensional printing. For each knee, 5 trochlear implants were designed: 1 implant simulated the native trochlea (control condition), and 4 implants simulated 4 types of trochlear dysplasia. The knees were subjected to 3 biomechanical tests: a squat simulation, an open chain extension simulation, and a patellar stability test. The patellofemoral kinematics, contact area, contact pressure, and stability were compared between the control condition (replica implants) and the trochlear dysplastic condition and among the subgroups of trochlear dysplasia. Results: The patellofemoral joint in the trochlear dysplastic group showed increased internal rotation, lateral tilt, and lateral translation; increased contact pressures; decreased contact areas; and decreased stability when compared with the control group. Within the trochlear dysplastic group, the implants graded as Dejour type D showed the largest deviations for the kinematical parameters, and the implants graded as Dejour types B and D showed the largest deviations for the patellofemoral contact areas and pressures. Conclusion: Patellofemoral kinematics, contact area, contact pressure, and stability are significantly affected by trochlear dysplasia. Of all types of trochlear dysplasia, the models characterized with a pronounced trochlear bump showed the largest deviations in patellofemoral biomechanics. Clinical Relevance: Investigating the relationship between the shape of the trochlea and patellofemoral biomechanics can provide insight into the short-term effects (maltracking, increased pressures, and instability) and long-term effects (osteoarthritis) of different types of trochlear dysplasia. Furthermore, this investigation provides an empirical explanation for better treatment outcomes of trochleoplasty for Dejour types B and D dysplasia. Keywords: patellofemoral dislocation; experimental testing; rapid prototyping; patellofemoral kinematics; patellofemoral pressures

isolated patellofemoral arthritis.2,10,15,22 Classification of TD to assess the severity or to advise on treatment is an important topic in patellofemoral research. The most commonly used classification system is the 4-grade Dejour classification.8 The 4 Dejour classes are defined on true lateral radiographs and axial computed tomography or magnetic resonance scans: type A is characterized by the crossing sign and a fairly shallow trochlea; type B, by the crossing sign, a trochlear spur, and a flat or convex trochlea; type C, by the crossing sign, a double contour representing a hypoplastic medial condyle, asymmetry of trochlear facets, and a convex lateral facet; and type D,

Dysplasia of the femoral trochlea is currently universally accepted as one of the most important factors in patellar instability.3,11 Trochlear dysplasia (TD) is a geometrical abnormality of the shape and depth of the trochlear groove mainly at its proximal part, where the patella engages into the trochlea. It may lead to patellofemoral maltracking, increased contact pressures, patellar instability, and

The American Journal of Sports Medicine, Vol. XX, No. X DOI: 10.1177/0363546515572143 Ó 2015 The Author(s)

1 Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

2

Van Haver et al

The American Journal of Sports Medicine

by the crossing sign, a supratrochlear spur, a double-contour sign, asymmetry of the trochlear facets, and a cliff pattern.8 It has been investigated whether the 4 Dejour classes reflect differences in objective morphologic characteristics of the patellofemoral joint, the incidence of patellofemoral arthritis, and the outcome of trochleoplasty.1,14,15,20,21,31 Thus far, it has not been possible to determine if the 4 Dejour classes also reflect differences in patellofemoral biomechanics. Few in vitro experiments have been performed to evaluate the patellofemoral biomechanics in TD knees versus normal knees. Senavongse and Amis25 simulated TD by removing a wedge of bone to flatten the lateral trochlea and by lifting the articular cartilage to elevate the central groove.2 These surgical techniques were successful to demonstrate significant differences in patellar stability and tracking between normal and TD knees. The aims of this study were (1) to investigate the effect of TD on the patellofemoral biomechanics and (2) to determine if different types of TD have a different effect on the patellofemoral biomechanics. This study applied a novel validated method to replace the original cadaveric trochlea with different types of custom-made rapid prototyped (RPT) implants for experimental testing.27 This method facilitates the investigation of isolated TD in all its variations.

MATERIALS AND METHODS Cadaveric Specimens Four unmatched fresh-frozen cadaveric knees, 2 male and 2 female (aged 75-85 years), were thawed at room temperature. Based on the computed tomography arthrograms and 3-dimensional models of the cadaveric knees, it was observed that knee 1 had trochlear cartilage damage and that knee 4 displayed TD type A. No patella alta or other morphologic abnormalities of the patellofemoral joint were observed.

Trochlear Implants To evaluate the effect of the trochlear shape on patellofemoral biomechanics, a validated method to simulate bony deformities in cadaveric specimens was applied.27 For each cadaveric knee, 5 types of custom trochlear implants were created on a 3-dimensional printer (multimaterial 3D Connex350 printer; Objet Ltd). The trochlear cartilage layer was simulated by using a rubberlike material with

Figure 1. Three-dimensional models (A and B) and picture (C) of the cadaveric femur with trochlear implant: The cutting plane is defined parallel to the posterior condylar line (1), proximal to the anterior trochlea (2) and anterior to the notch (3), ensuring contact of the patella with the trochlear implant from 0° to 60° of knee flexion. a hardness comparable with cartilage (Shore A 80-90, tested according to ASTM standard D-2240); the rest of the implant was printed by a hard rubberlike material closer to bone (Shore A 90-100, tested according to ASTM standard D-2240). The first implant replicated the trochlear shape of the cadaver (replica implant), and the 4 other implants replicated different TD shapes (patient-based implants). The patient-based implants were based on computed tomography arthrograms of 4 TD patients classified as Dejour types A, B, C, and D with a history of recurrent patellar dislocation. These patient-based implants were designed to encompass the geometric characteristics of the pathologic abnormalities of the patients and, at the same time, fit the cadaveric specimens. To this end, the patient femur models were scaled and positioned to match the cadaveric femur model, and the cutting plane was defined (Figure 1). This plane was used to design the replica and patient-based implant as well as the guiding instruments to cut the cadaveric trochlea. After the trochleae were separated from the femur models, the cadaveric femur model could be combined with a patient trochlea model. Obviously, the connection between the cadaveric femur and patient trochlea will be stepwise, which is undesirable for this study. To ensure a perfectly smooth fit, a transition zone of approximately 1 cm was selected in the area where the trochlea made contact with the cadaver. In this transition zone, manual smoothing was applied without altering the cadaveric model and without altering the pathologic characteristics of the trochlea. In the design of the replica and patient-based implants, the loss of bone caused by the saw blade was compensated

*Address correspondence to Annemieke Van Haver, MSc, PhD, Monica Orthopaedic Research (MORE) Institute, Stevenslei 20, 2100 Antwerp, 0032 493181435, Belgium (email: [email protected]). y Department of Industrial Technology and Construction, Ghent University, Ghent, Belgium. z Department of Construction and Production, Ghent University, Ghent, Belgium. § Monica Orthopaedic Research (MORE) Institute, Antwerp, Belgium. || Department of Physical medicine and orthopaedic surgery, Ghent University, Ghent, Belgium. { Department of Civil Engineering, IBiTech–bioMMeda, Ghent University, Ghent, Belgium. # Department of Mechanical Engineering–Division of Biomechanics, Catholic University Leuven, Leuven, Belgium. **Department of Orthopaedics, Lyon-Ortho-Clinic, Clinique de la Sauvegarde, Lyon, France. yy Antwerp Orthopedic Center, Monica Hospitals, Antwerp, Belgium. One or more of the authors has declared the following potential conflict of interest or source of funding: Research was funded by the University College of Ghent. Tests were performed at the European Centre for Knee Research, funded by Smith & Nephew.

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

Vol. XX, No. X, XXXX

Biomechanical Effects of Trochlear Dysplasia

3

Figure 2. Test rigs to evaluate the patellofemoral kinematics, contact area, and contact pressure (squat and open chain simulation). by adding a layer of 1.2 mm at the contact area of the trochlear implant.

Morphology and Classification of the Modified Cadaveric Knees Each of the 4 cadaveric knees was experimentally tested with 1 replica implant and 4 patient-based implants, resulting in 20 modified femur models (a cadaveric femur with a patient trochlea). These 20 modified femur models were presented to 2 blinded senior surgeons (P.V., D.D.) as 2-dimensional slices and 3-dimensional images. Consequently, each of the 20 femur models was assigned to 1 of the 4 Dejour classes upon consensus. When the surgeon had access to view the 4 patient models, he succeeded to assign the 20 modified femur models to the correct patient model. But when the surgeon was fully blinded, he did not always classify the modified femur model as the same type of TD as the original patient model on which the modified femur model was based. This discrepancy was mainly caused by a difference in shaft diameter between the cadaver and patient; when the distal femoral condyles were optimally positioned (see the previous section on trochlear implants) to replace the cadaveric trochlea by the patient trochlea, the femur with the smallest shaft diameter will show the largest anterior offset and vice versa. This strategy was applied to avoid too much anteriorization of the central trochlea, which has been reported as a possible cause of tightening of the medial retinaculum and medialization of the patella.2 Of the 20 modified cadaveric knees, 3 were classified as normal (15%), 7 as type A (35%), 3 as type B (15%), 5 as type C (25%), and 2 as type D (10%).

Measurement of Patellofemoral Kinematics and Contact Pressure and Area Patellofemoral kinematics have been shown to be different in open versus closed chain exercises and in weightbearing

versus nonweightbearing exercises. Moreover, patellar maltracking has been reported to be more pronounced in nonweightbearing open chain exercises.12,24 Therefore, the effects of the different trochlear shapes were evaluated during a squat simulation (from 35° to 75° of knee flexion) as well as an open chain extension simulation (from 65° to 5° of knee flexion). Although it is well accepted that lateral maltracking occurs during the last 20° to 30° of knee extension,12,23,28 it was not possible to obtain higher extension positions during the squat simulation, as the knee could be pulled in hyperextension, which would damage the specimen. A knee rig, based on the Oxford Knee Rig, was used to simulate a squat movement and an open chain extension movement (Figure 2).29 The motion of femur, tibia, and patella was recorded by 6 calibrated infrared cameras (MX401; Vicon Motion Systems), and the patellofemoral kinematics were defined according to Belvedere et al.4 The patellofemoral contact area and pressure were recorded by an I-scan 4000 pressure sensor (Tekscan). Before testing, the sensor was conditioned and calibrated according to manufacturer instructions using an axial/torsional loading frame (Bionix 370.02; MTS).18 Since the cadaveric specimens show substantial size differences—as demonstrated in the Appendix (available online at http://ajsm.sagepub.com/supplemental)—the mediolateral translation and the patellofemoral contact area were normalized by multiplying the translation with the rescaling factor and by multiplying the contact area by the squared rescaling factor described above.

Measurement of Patellofemoral Stability The stability of the patella was quantified by recording the mediolateral translation of the patella caused by applying a force of 100 N in the lateral and medial direction (Figure 3). Based on earlier studies, it was assumed that the patella would not reach its ultimate lateral or medial position before achieving the maximum force of 100 N.26

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

4

Van Haver et al

The American Journal of Sports Medicine

TABLE 1 Direction and Distribution of Weights Attached to the Tendons of the Quadricepsa Load Distribution Load Direction Angles VLL VLO VML VMO RF 1 VI ITB

14° 35° 15° 47°

lateral lateral medial medial

33° posterior 44° posterior 6° posterior

%

N

33 9 14 9 35 100

57.75 15.75 24.50 15.75 61.25 30.00

a ITB, iliotibial band; RF 1 VI, rectus femoris 1 vastus intermedius; VLL, vastus lateralis longus; VLO, vastus lateralis obliquus; VML, vastus medialis longus; VMO, vastus medialis obliquus.

Preparation of the Cadaveric Specimens Figure 3. Schematic representation of the test rig to evaluate the patellofemoral stability. ITB, iliotibial band; RF 1 VI, rectus femoris 1 vastus intermedius; VLL, vastus lateralis longus; VLO, vastus lateralis obliquus; VML, vastus medialis longus; VMO, vastus medialis obliquus.

A stability rig, partially based on the stability rig described by Senavongse et al,26 was used to evaluate the effect of the different trochlear shapes on the patellar stability in different knee flexion angles. The stability rig consisted of a frame in which the knee specimen was positioned horizontally with the lateral side of the knee on top. The femur was rigidly fixed, and the distal tibia was positioned in an open cylindrical container, which was attached to the rig in a circular track around the knee flexion-extension axis. With this container, the knee was sequentially fixed in 10°, 20°, 30°, 45°, and 60° of knee flexion, while all other degrees of freedom of the tibial and patellar motion were left unconstrained. In higher extension angles, the patella was not engaged in the trochlea; therefore, no significant differences were expected in full extension.2 A total load of 175 N was applied to the quadriceps tendons, and 30 N was applied on the iliotibial band by hanging a series of calibrated weights according to the physiologic cross-sectional areas and directions of the muscles relative to the femoral axis as indicated in Table 1.5,13,19 To displace the patella laterally and medially from its stable neutral position, a force of 100 N was applied manually to a screw in the most medial and most lateral border of the patella. This force was monitored by a digital dynamometer. The patellofemoral kinematics were recorded analogous to the kinematics during the squat and open chain simulation. The mediolateral position of the patella in its stable neutral position and in its ultimate medial and lateral position (after applying 100 N medially and laterally) was normalized for size differences.

The femur and tibia were rigidly fixed with polymethylmethacrylate in 2 containers, taking the cadaver-specific physiologic valgus angle into account. To place the trochlear implants, a custom-made guiding instrument was fitted on the articular surface of the distal femur. Care was taken to place the instrument in the best-fitting and most stable position. Monitoring of the position was facilitated by the transparent material and perforations in the guiding instrument.27 The instrument was then fixed with pins at the anterior and lateral side. After fixation, an oscillating saw blade was guided through the lateral slot in the instrument to remove the femoral trochlea. When the guiding instrument and the trochlea were removed from the distal femur, a second guiding instrument was used to ream a cylindrical socket in the surface of the trabecular bone of the femur. In this socket, a fixation component was cemented to enable easy replacement of one trochlear implant by another during the experimental tests. After the implants were placed, the knee joint was closed with a Vicryl 2 suture (Ethicon NV; Johnson & Johnson) to minimize the effect of the manipulations on the patellofemoral biomechanics.

Statistical Analysis A 2-way analysis of variance for unequal sample sizes was conducted to examine the effect of TD on the patellofemoral kinematics, contact area and pressure, and patellar stability. Analyses were first done for the TD models (TD group) versus the normal models (control group). Subsequently, analyses were done for the different subgroups. If significant differences between the subgroups were found, Tukey post hoc tests were performed for pairwise comparisons.

RESULTS Effect on Patellofemoral Kinematics Squat. During the squat simulation, the mean patellar tracking showed progressive external rotation, lateral

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

Vol. XX, No. X, XXXX

Biomechanical Effects of Trochlear Dysplasia

Effect on Patellofemoral Stability The neutral position of the patella tended to be more lateral in the TD group when compared with the control group (on average, 4 mm; ns) (Figure 5). After applying 100 N in the lateral direction, the patella shifted, on average, 3 mm more lateral in the TD group versus the control group (P = .007). The largest difference (4 mm) in displacement between the normal and TD groups was observed at 20° of knee flexion. The final mediolateral position was, on average, 7 mm more lateral in the TD group versus the control group (P = .031).

External Rotaon (°) Internal Medial Tilt (°) Lateral Lateral ML transl. (mm) Medial

Squat. During the squat simulation, the mean patellofemoral contact pressure and area increased with knee flexion from 35° to 75°. On average, the contact pressure was higher (P = .028) and the contact area lower (P \ .001) in the TD group versus the control group (Figure 4). Analysis of the 4 classes revealed that types B and D showed higher patellofemoral contact pressures (P = .002) and lower patellofemoral contact areas (P = .006) when compared with types A and C. Open Chain. During the open chain extension simulation, the mean patellofemoral contact pressure increased with knee extension from 65° to 15° and decreased beyond 15° of knee flexion. The contact area increased from 65° to 45° of knee flexion and decreased beyond 45° of knee flexion. On average, the contact pressure was higher (P = .001) and the contact area lower (P \ .001) in the TD group versus the control group (Figure 4). Analysis of the 4 classes revealed that types B and D showed higher patellofemoral contact pressures (P \ .001 ) and higher patellofemoral contact areas (P \ .001) when compared with types A and C.

Squat

Open chain

15

15

10

10

5

5

0

0

**

-5

**

-5

-10

-10 30 40 50 60 70 80

20 15 10 5 0 -5 -10

0

15 30 45 60 75

20 15 10 5 0 -5 -10

*

**

0

30 40 50 60 70 80

Contact P (MPa)

Effect on Patellofemoral Contact Pressure and Area

TD group Control group

10

10

5

5

0

0

-5

-5

15 30 45 60 75

**

-10

-10 30 40 50 60 70 80

0

5

5

4

4

15 30 45 60 75

3

3 *

2

2

1

1

0

0 30 40 50 60 70 80

Contact area (mm2)

tilt, and lateral translation with increasing knee flexion from 35° to 75° (Figure 4). The patellae showed more internal rotation (P \ .001) and lateral tilt (P = .016) in the TD group compared with the control group. For the mediolateral translation, there was no significant difference. Among the 4 Dejour classes, analyses of the kinematics showed no specific differences. Open Chain. During the open chain extension simulation, the mean patellar tracking in the control group showed a progressive internal rotation, medial tilt, and medial translation with increasing knee extension from 65° to 25° and no rotation, a lateral tilt, and lateral translation from 25° to 5° of knee flexion (Figure 4). The patella was consistently more internally rotated (P \ .001), laterally tilted (P \ .001), and more laterally positioned (P \ .001) in the TD group versus the control group. Tukey post hoc testing revealed that type D showed more internal rotation (P = .001) and more lateral tilt (P = .049) compared with types A, B, and C. For the mediolateral translation, no difference was observed between type D and types A, B, and C.

5

** 0

15 30 45 60 75

0

15 30 45 60 75 Knee flexion angle (°)

500

500 400

**

400

300

300

200

200

100

100 **

0

0 30 40 50 60 70 80 Knee flexion angle (°)

Figure 4. Patellofemoral rotation, tilt, mediolateral (ML) translation, contact pressure (P), and contact area during the squat simulation and open chain extension simulation as a function of knee flexion for the control group and trochlear dysplastic group (TD group). Significant differences between the TD group and the control group: *P \ .05. **P \ .01.

Applying 100 N on the patella in the medial direction did not demonstrate a significant difference between the control and TD group. For the 4 Dejour classes, the analysis of the patellar stability in the lateral and medial direction showed no significant differences.

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

6

Van Haver et al

The American Journal of Sports Medicine

Control group

TD group

Lateral PF posion (mm) Medial

35

25 Medial posion 15

Medial transl.15

Neutral posion Lateral transl.** 5 Lateral posion*

5

-5

-5

-15 10 20 30 45 60 Knee flexion angle

10 20 30 45 60 Knee flexion angle

Figure 5. Results of the patellar stability test for control group and the trochlear dysplastic group (TD group). Patellofemoral (PF) mediolateral position in rest (neutral position) and after applying 100 N on the patella in the medial direction (medial position) and lateral direction (lateral position) as a function of knee flexion. Significant differences between the TD group and the control group: * P \ .05. **P \ .01.

DISCUSSION This new research method, using RPT implants to simulate TD, showed that TD has a significant effect on patellofemoral biomechanics, especially in TD knees with a pronounced trochlear bump. Dejour type D showed the largest deviation in terms of patellofemoral kinematics. Dejour types B and D showed the largest deviation in terms of patellofemoral contact mechanics. Further differentiation in the biomechanical behavior according to the 4 Dejour classes was not possible. The highly significant differences in patellofemoral kinematics, contact area, contact pressure, and stability between the TD and control groups are generally in line with earlier studies by Amis and colleagues.2 However, Amis and colleagues observed that the patellae in the TD knees tended to move medially as compared with the normal knees, which was probably due to tightening of the medial retinacular restraints caused by the elevation of the central groove in the earlier study. In the current study, care was taken to avoid excessive anterior elevation of the trochlea. Consequently, the medial retinaculum was not tightened, but the trochlear bump was sometimes less pronounced than initially intended (as described below). Classification of TD is an important topic in patellofemoral research, and to date, the 4-grade Dejour classification is the most widely used system to grade the severity of TD. Nevertheless, differences in patellofemoral biomechanics among the 4 Dejour classes have, to our knowledge, not been investigated before. Several important findings are based on this 4-grade classification; however, the observer reliability of the classification has recently been questioned, and a more reliable 2-grade classification

was proposed by grading Dejour type A as low grade and grouping types B, C, and D as high grade TD.17 This alternative grouping is supported by 2 studies, showing that only type A could be morphologically distinguished from types B, C, and D by applying the Dejour classification as well as by applying morphologic cutoff values.17,20 Notwithstanding the superior observer reliability of this 2grade classification, other studies, including the current study, do not seem to support that the 2-grade classification (separating type A from types B, C, and D) is the most appropriate grouping in terms of prognosis and treatment. The only significant predictor of a better outcome after sulcus-deepening trochleoplasty was the presence of a supratrochlear spur or bump, which resulted in a more restricted indication of this technically demanding procedure for type D and severe type B, while TD types A and C and mild type B are advised to be treated with other techniques.6,9,14,21 Furthermore, it has been reported that isolated patellofemoral arthritis is particularly associated with TD in the presence of a supratrochlear bump (Dejour types B and D).1,15,31 The current experimental study showed that TD knees with a distinct trochlear bump trigger more severe biomechanical deviations. Deviations in the patellofemoral kinematics were most severe in type D during the open chain extension simulation. This observation was not found during the squat simulation. This difference between open chain and squat indicates that open chain exercises are indeed more provocative for patellar maltracking than closed chain exercises.12,24 In the current study, the patella was more externally rotated and laterally tilted in type D when compared with the other types. Previous studies on patient populations have also shown that lateral patellar tilt,

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

Vol. XX, No. X, XXXX

Biomechanical Effects of Trochlear Dysplasia

measured statically on axial images, is highly correlated with the severity of TD.7,8 Deviations in the patellofemoral contact pressure and area were most severe in Dejour types B and D; contact areas were lower and contact pressures were higher compared with types A and C. These findings confirm and explain results of in vivo studies. Isolated patellofemoral arthritis is highly associated with TD, particularly in the presence of a supratrochlear spur (Dejour types B and D).1,15,31 It is assumed that the prominence of the proximal trochlea causes an increase of the patellofemoral compression in knee flexion, like a reverse Maquet effect.15,31 This explanation is also believed to be the reason for the better outcome of trochleoplasty in types B and D.14 The current study shows experimental evidence that the patellofemoral contact pressures are indeed substantially increased in types B and D. However, in living patients, the pathogenesis of TD is a lot more complex when compared with this analytic in vitro study. While the current study analyzed the ‘‘instant’’ biomechanical effect of TD as an isolated anatomic abnormality of the trochlea, in vivo TD is part of a multifactorial condition in which adaptive changes occur over the years. Moreover, experimental animal studies have suggested that the origin of TD can be secondary to an abnormal patellar position.16,30 The reduced patellofemoral stability in the TD group was most pronounced at 20° of knee flexion, which agrees with previous research.25 The patellofemoral stability showed no significant differences among the 4 classes. This in vitro study has some inherent limitations. The Dejour classification was applied to grade the TD models, while the observer agreement of this classification is still subject to discussion. Contradictory suggestions have been made on the classification; further differentiation of the 4 classes (type A, mild type B, severe type B, type C, and type D) is sometimes used to improve guidelines for treatment decision, while grouping of the 4 classes (highand low-grade TD) has been proposed to improve observer agreement. In addition, TD is sometimes undiagnosed or underestimated in patients.6 These statements in the literature on TD classification indicate that further research on trochlear shape variations is essential to further unravel the complex relation between trochlear shape and patellofemoral biomechanics. An important consequence for TD research is the possibility that study results based on this classification, including the results of the current study, might be biased to some grouping incorrectness. A second limitation is that TD was simulated without altering the patellar articular shape nor the patellar height, leading to potential mismatch of the patellofemoral joint. Amis et al2 already discussed that this mismatch is justified because the lateral patellar facet stays in contact with the trochlea, while the medial facet lifts off the trochlear surface, which is actually similar to clinical situations. Compared with the TD simulations of Amis et al, the current method may be better at replicating different shape variations. However, the RPT material differs from the native trochlear material. Apart from the hardness (which was cartilage-like), the friction coefficient, Young modulus,

7

and Poisson ratio are also potentially important characteristics that can affect the patellofemoral joint biomechanics. These material properties are not provided by the supplier and should be further investigated to fully understand their effect on biomechanical experiments. To assess the effect of using RPT material in cadaveric experiments, the differences in knee biomechanics between native knees and knees with a RPT replica were described by Van Haver et al.27 Because the RPT material differs from cartilage, the results of this study can be used to compare different types of TD with one another, but differences in material properties should be taken into account when extrapolating these results to living patients. Furthermore, as in other in vitro studies, the test setups used have wellknown restrictions that can affect the kinematics and stability of the patella. An important restriction in the stability rig is that the patellar rotation and tilt are influenced by the applied lateral and medial load. Ideally, the patella should be loaded in only the mediolateral direction, which could be achieved by cementing a ball bearing in the center of the patella.26 Nevertheless, the current setup provided an important indication of the patellar stability. Finally, another limitation is the limited number of tested knees and the unequal number of tested cases in each subgroup. Therefore, statistical analysis for unequal sample size was performed. Notwithstanding these limitations, this study is the first controlled cadaveric study demonstrating that a pronounced trochlear bump (Dejour types B and D) induces higher contact pressures. This experimental observation supports the hypothesis that a pronounced trochlear bump increases patellofemoral compression like a reverse Maquet effect, which has been reported as a mechanism in the development of isolated patellofemoral arthritis in Dejour types B and D and the better outcome of trochleoplasty in Dejour types B and D.14,15,31 To conclude, this experimental study on modified cadaveric knees demonstrated that TD adversely affects patellofemoral kinematics, contact mechanics, and stability. The presence of a trochlear bump showed to be an important provocative factor for severely deviating patellofemoral kinematics and increased contact pressures.

ACKNOWLEDGMENT The authors thank Dr Wouter Huysse (Department of Radiology, University Hospital Ghent) and Tom Van Hoof (Department of Anatomy, Ghent University) for the imaging data of the cadaveric specimens and Ronny De Corte (European Centre for Knee Research–Leuven, Smith & Nephew) for his assistance during the experimental testing.

REFERENCES 1. Allain J, Dejour D, Argenson JN, et al. Isolated patello-femoral osteoarthritis. Rev Chir Orthop Reparatrice Appareil Moteur. 2004;90:69129.

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015

8

Van Haver et al

The American Journal of Sports Medicine

2. Amis AA, Oguz C, Bull AMJ, Senavongse W, Dejour D. The effect of trochleoplasty on patellar stability and kinematics: a biomechanical study in vitro. J Bone Joint Surg Br. 2008;90:864-869. 3. Arendt EA, Dejour D. Patella instability: building bridges across the ocean a historic review. Knee Surg Sports Traumatol Arthrosc. 2013;21:279-293. 4. Belvedere C, Catani F, Ensini A, de la Barrera JLM, Leardini A. Patellar tracking during total knee arthroplasty: an in vitro feasibility study. Knee Surg Sports Traumatol Arthrosc. 2007;15:985-993. 5. Bull AMJ, Andersen HN, Basso O, Targett J, Amis AA. Incidence and mechanism of the pivot shift: an in vitro study. Clin Orthop Relat Res. 1999;363:219-231. 6. Dejour D, Byn P, Ntagiopoulos PG. The Lyon’s sulcus-deepening trochleoplasty in previous unsuccessful patellofemoral surgery. Int Orthop. 2013;37:433-439. 7. Dejour D, Le Coultre B. Osteotomies in patello-femoral instabilities. Sports Med Arthrosc Rev. 2007;15:39-46. 8. Dejour D, Reynaud P, Lecoultre B. Douleurs et instabilite´ rotulienne: essai de classification. Me´decin et Hygie`ne. 1998;56:1466-1471. 9. Dejour D, Saggin P. The sulcus deepening trochleoplasty: the Lyon’s procedure. Int Orthop. 2010;34:311-316. 10. Dejour H, Walch G, Neyret P, Adeleine P. Dysplasia of the femoral trochlea. Rev Chir Orthop Reparatrice Appareil Moteur. 1990;76:45-54. 11. Dejour H, Walch G, Nove-Josserand L, Guier C. Factors of patellar instability: an anatomic radiographic study. Knee Surg Sports Traumatol Arthrosc. 1994;2:19-26. 12. Doucette SA, Child DD. The effect of open and closed chain exercise and knee joint position on patellar tracking in lateral patellar compression syndrome. J Orthop Sports Phys Ther. 1996;23:104-110. 13. Farahmand F, Senavongse W, Amis AA. Quantitative study of the quadriceps muscles and trochlear groove geometry related to instability of the patellofemoral joint. J Orthop Res. 1998;16:136-143. 14. Fucentese SF, Zingg PO, Schmitt J, Pfirrmann CW, Meyer DC, Koch PP. Classification of trochlear dysplasia as predictor of clinical outcome after trochleoplasty. Knee Surg Sports Traumatol Arthrosc. 2011;19:1655-1661. 15. Grelsamer RP, Dejour D, Gould J. The pathophysiology of patellofemoral arthritis. Orthop Clin North Am. 2008;39:269-274. 16. Huri G, Atay OA, Ergen B, Atesok K, Johnson DL, Doral MN. Development of femoral trochlear groove in growing rabbir after patellar instability. Knee Surg Sports Traumatol Arthrosc. 2012;20:232-238. 17. Lippacher S, Dejour D, Elsharkawi M, et al. Observer agreement on the Dejour trochlear dysplasia classification: a comparison of true lateral radiographs and axial magnetic resonance images. Am J Sports Med. 2012;40:837-843. 18. Maurer J, Ronsky J, Loitz-Ramage B, Andersen M, Zernicke R, Harder J. Prosthetic socket interface pressures: customized calibration technique for the TEKSCAN F-socket system. Summer

19.

20.

21.

22. 23.

24.

25.

26.

27.

28.

29.

30.

31.

Bioengineering Conference. 2003;29. Available at: http://www.tulane .edu/~sbc2003/pdfdocs/1073.PDF. Merican AM, Kondo E, Amis AA. The effect on patellofemoral joint stability of selective cutting of lateral retinacular and capsular structures. J Biomech. 2009;42:291-296. Nelitz M, Lippacher S, Reichel H, Dornacher D. Evaluation of trochlear dysplasia using MRI: correlation between the classification system of Dejour and objective parameters of trochlear dysplasia [published online November 30, 2012]. Knee Surg Sports Traumatol Arthrosc. doi:10.1007/s00167-012-2321-y Ntagiopoulos PG, Byn P, Dejour D. Midterm results of comprehensive surgical reconstruction including sulcus-deepening trochleoplasty in recurrent patellar dislocations with high-grade trochlear dysplasia. Am J Sports Med. 2013;41:998-1004. Pfirrmann CWA, Zanetti M, Romero J, Hodler J. Femoral trochlear dysplasia: MR findings. Radiology. 2000;216:858-864. Powers CM. Patellar kinematics, part II: the influence of the depth of the trochlear groove in subjects with and without patellofemoral pain. Phys Ther. 2000;80:965-973. Powers CM, Ward SR, Fredericson M, Guillet M, Shellock FG. Patellofemoral kinematics during weight-bearing and non-weight-bearing knee extension in persons with lateral subluxation of the patella: a preliminary study. J Orthop Sports Phys Ther. 2003;33:677-685. Senavongse W, Amis AA. The effects of articular, retinacular, or muscular deficiencies on patellofemoral joint stability: a biomechanical study in vitro. J Bone Joint Surg Br. 2005;87:577-582. Senavongse W, Farahmand F, Jones J, Andersen H, Bull AMJ, Amis AA. Quantitative measurement of patellofemoral joint stability: forcedisplacement behavior of the human patella in vitro. J Orthop Res. 2003;21:780-786. Van Haver A, De Roo K, Claessens T, et al. The use of rapid prototyping to simulate knee joint abnormalities for experimental cadaver research: a validation study with replica implants of the native trochlea. Proc Inst Mech Eng H. 2014;228:833-842. Varadarajan KM, Freiberg AA, Gill TJ, Rubash HE, Li GA. Relationship between three-dimensional geometry of the trochlear groove and in vivo patellar tracking during weight-bearing knee flexion. J Biomech Eng. 2010;132:061008. Victor J, Van Glabbeek F, Vander Sloten J, Parizel PM, Somville J, Bellemans J. An experimental model for kinematic analysis of the knee. J Bone Joint Surg Am. 2009;91(suppl 6):150-163. Yamada Y, Toritsuka Y, Yoshikawa H, Sugamoto K, Horibe S, Shino K. Morphological analysis of the femoral trochlea in patients with recurrent dislocation of the patella using three-dimensional computer models. J Bone Joint Surg Br. 2007;89:746-751. Zaffagnini S, Dejour D, Arendt EA. Patellofemoral Pain, Instability, and Arthritis: Clinical Presentation, Imaging, and Treatment. Berlin, Germany: Springer-Verlag; 2010.

For reprints and permission queries, please visit SAGE’s Web site at http://www.sagepub.com/journalsPermissions.nav

Downloaded from ajs.sagepub.com at FLORIDA ATLANTIC UNIV on May 23, 2015