Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carrillon, Y.
Right arrow Articles by Tran-Minh, V. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carrillon, Y.
Right arrow Articles by Tran-Minh, V. A.
(Radiology. 2000;216:582-585.)
© RSNA, 2000

Technical Developments

Patellar Instability: Assessment on MR Images by Measuring the Lateral Trochlear Inclination-Initial Experience1

Yannick Carrillon, MD, Hassane Abidi, PhD, David Dejour, MD, Olivier Fantino, MD, Bernard Moyen, MD and Van A. Tran-Minh, MD

1 From the Departments of Radiology (Y.C., O.F., V.A.T.M.), Medical Statistics (H.A.), and Orthopedics (D.D., B.M.), Centre Hospitalier Lyon-Sud, Chemin du grand Revoyet, 69495 Pierre Bénite Cedex, France. Received June 22, 1999; revision requested August 10; final revision received November 9; accepted November 22. Address correspondence to Y.C. (e-mail: carrillon@grisn.univ-lyon1.fr).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
Results
 Discussion
 REFERENCES
 
Lateral trochlear inclination (LTI) of the knee was compared on magnetic resonance (MR) images obtained in 30 patients with patellar instability (PI) and 30 patients with nonspecific internal knee derangement. Differences in LTI values between the two populations were significant (P < .001). Reproducibility of the measurement was judged excellent with an intraclass correlation superior to 0.98. Below a threshold value fixed at 11°, LTI appears to be an excellent diagnostic test of PI with a sensitivity of 0.93 (28/30), a specificity of 0.87 (26/30), and an accuracy of 0.90 (54/60).

Index terms: Joints, abnormalities, 452.42 • Knee, abnormalities, 452.42, 452.43, 453.42, 453.43 • Knee, MR, 452.121415, 453.121415 • Patella, 453.42, 453.43


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
Results
 Discussion
 REFERENCES
 
Among patellofemoral disorders, patellar instability (PI) is one of the most frequent clinical syndromes. It is attributed to the abnormal course of the patella during knee flexion (1–3). Numerous clinical presentations may reveal this entity. The diagnosis of PI is easy when there are clinical and physical evidences of prior dislocations of the patella. The diagnosis can be more confusing when the patients have a normal knee examination and no history of dislocation of the patella. These patients may have only a nonspecific knee pain that could be confused with other forms of internal knee derangement (2–4). In these cases, imaging techniques play a major role, since they may demonstrate findings that could be related to abnormal tracking of the patella (5). Dysplasia of the femoral trochlea is one of the findings that could be detected on knee radiographs obtained in patients with PI. Its frequency has been studied and reported by numerous authors on the basis of either transverse or lateral views of the patellofemoral joint (6–8). Both views have limits for assessing femoral trochlear dysplasia, however, since small dysplasia may be present on only the proximal portion of the trochlea, a location that is difficult to analyze with conventional radiography (9).

Magnetic resonance (MR) imaging is an excellent imaging technique for exploring meniscal and other forms of internal knee derangements, but it is not yet considered as useful for the diagnosis of PI. Unlike conventional radiographs, however, transverse MR images allow complete and clear visualization of the entire patellofemoral joint by clearly delineated the proximal portion of the trochlea. Measurement of the lateral trochlear inclination (LTI) on transverse images of the patellofemoral joint is one method for assessing femoral trochlear dysplasia (10). To our knowledge, this method has not been previously described on MR images. The purpose of this study was first to determine if measuring the LTI on MR images depicting the proximal portion of the femoral trochlea was a feasible and reproducible method. Then, LTI values obtained in a group of patients with PI were compared with those obtained in a group of control subjects to evaluate the reliability of this technique for corroborating the diagnosis of PI.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion
 REFERENCES
 
Subjects
From November 1996 through March 1998, 30 consecutive patients with PI (22 female and eight male patients; age range, 15–42 years; mean age, 24 years) were referred to undergo MR imaging to plan a surgical procedure (12 right and 18 left knees). No evidence of knee osteoarthritis was depicted on radiographs for any of the patients. The condition for inclusion in this group of PI patients was a history of at least two episodes of patellar dislocation. Patellar dislocation was diagnosed and noted by the referring physician at the time of the physical examination.

The control population comprised 30 consecutive patients referred during the same period as the PI patients to undergo MR evaluation of an internal derangement of the knee. The control population was cross matched by sex and age with the patient population. For this reason, four men were excluded on the basis of their age (range, 46–64 years). We reviewed 20 right and 10 left knees in 24 women and six men (age range, 19–37 years; mean age, 26 years). Various knee disorders were found on MR images: meniscal tears in 22 cases and anterior cruciate ligament tears in 12. Clinical examination of these patients was performed by the same physicians who examined the study patients. They noticed no physical sign that could be related to PI.

Imaging Protocol
All the MR examinations were performed on the same 1.5-T unit (Gyroscan ACS; Philips Medical Systems, Eindhoven, the Netherlands) with use of a birdcage knee coil (diameter, 14 cm). The knee was totally extended within the coil and not rotated. The MR examination was performed with the same imaging protocol for all patients and control subjects with a transverse fat-suppressed T2-weighted turbo spin-echo sequence: repetition time msec/echo time msec of 2,000/60, eight turbo factors, six signals acquired, 256 x 192 matrix, 16-cm field of view, 4-mm-thick sections, and 0.4-mm intersection gap. The sequence was used to explore the entire patellofemoral joint.

Measurement of the LTI
To analyze the proximal portion of the trochlea, an area prone to dysplasia, we decided to measure LTI on the first craniocaudal image that demonstrated cartilaginous trochlea. This image was considered the reference image. On fat-suppressed T2-weighted images, cartilage is clearly delineated as a thick band of intermediate signal intensity adjacent to subchondral bone that contrasts with the high signal intensity of the articular fluid. At this level, the lateral facet of the trochlea is formed (Fig 1). Reference images from all the subjects were then transferred onto a personal computer. LTI measurements were obtained by using an adapted software (IMAGE; Scion, Frederick, Md). LTI was calculated by means of a line tangential to the subchondral bone of the posterior aspect of the two femoral condyles crossed with a line tangential to the subchondral bone of the lateral trochlear facet (Fig 2). Both choice of reference image and LTI measurement were performed independently by two observers (Y.C., D.D.).



View larger version (98K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Successive caudocranial transverse T2-weighted images with fat suppression explore the entire patellofemoral joint. The reference image allows analysis of the proximal portion of the cartilaginous trochlea (arrow).

 


View larger version (157K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Reference images obtained in (a) a control subject and (b) a patient with PI. LTI was measured by crossing a line adjacent to the posterior edges of the two condyles (open arrow) with another line tangential to the lateral trochlear facet (solid arrow). LTI was 16° in a and 3° in b.

 


View larger version (159K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Reference images obtained in (a) a control subject and (b) a patient with PI. LTI was measured by crossing a line adjacent to the posterior edges of the two condyles (open arrow) with another line tangential to the lateral trochlear facet (solid arrow). LTI was 16° in a and 3° in b.

 
Statistical Analysis
Interobserver reproducibility was analyzed by means of a regression procedure and intraclass correlation (SAS; SPSS, Cary, NC). For the regression procedure, the test of intercept equals 0 and slope equals 1 was performed.

The Student t test was used to compare the means of cases and controls. The Shapiro-Wilks test was used to test the normality (P = .599), and the Levene test was used to test the equality of variances (P = .127). Software (MEDCALC; MedCalc, Mariekerke, Belgium) was used to build the observed receiver operating characteristic curve. This curve was fitted by means of a maximum likelihood method (ROCKIT; Metz CE, University of Chicago, Ill). The area under the curve and its CI were then calculated. The LTI threshold value was selected by determining the value that provided the highest accuracy (minimal false-negative and false-positive findings). The effect of this choice on type I errors (1 - sensitivity) and type II errors (1 - specificity) was then illustrated by adjusting the observed LTI distribution by the normal probability density functions.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Distinctive regression lines obtained in study cases and controls suggested a strong interobserver reliability for measurement of LTI (Fig 3). Furthermore, intraclass correlations and 95% CIs were, respectively, 0.986 and 0.972–0.993 for controls and 0.988 and 0.976–0.994 for cases. The distinctive regression lines also confirmed the high level of concordance between observers. The measurement of the first physician was retained for the data analysis.



View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Distinctive regression lines for cases and controls for the LTI values measured by the second observer (LTI2) versus the first observer (LTI1). Regression equations were LTI2 = 0.184 + 1.0124 x LTI1 for cases and LTI2 = 0.927 + 0.953 x LTI1 for controls. The hypothesis that the intercept is null and the slope is equal to unity could not be rejected on the basis of the observed lines; therefore, the regression lines appear identical to the diagonal line.

 
The mean LTI value was 6.17° (SD, 4.97) in patients and 16.93° (SD, 4.76) in controls. The difference in means between these groups was significant (P < .001).

The area under the receiver operating characteristic curve (Fig 4) and its 95% CI were, respectively, 0.95 and 0.88–0.99. The choice of 11° as the threshold value for LTI results in excellent discrimination between the two groups, with a sensitivity of 0.93 (28/30), a specificity of 0.87 (26/30), and an accuracy of 0.90 (54/60) (Table). The effect of this choice on the type I errors ({alpha} or false-positive findings) and type II errors (ß or false-positive findings) is illustrated in Figure 5. Values of {alpha} and ß were, respectively, 0.166 and 0.106. Consequently, the true-positive (1 - {alpha}) and true-negative (1 - ß) values were, respectively, 0.834 and 0.894. These values were slightly different from sensitivity and specificity obtained directly from data as a consequence of the adjustment process.



View larger version (46K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Observed and fitted receiver operating characteristic curves of LTI values. The dotted line represents the distinct actual LTI values. The area under the curve is 0.95 (95% chance that case has a lower value than control), which indicates an excellent discrimination between the two groups.

 

View this table:
[in this window]
[in a new window]

  		Case Distribution for LTI Threshold Value of 11°
 


View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5. Adjusted probability density function of LTI in cases and controls on the basis of observed means and variances with 11° as the threshold value. Type I error (probability of LTI greater than 11° in cases) is 0.116, and type II error (probability of LTI of 11° or less in controls) is 0.106. Sensitivity and specificity obtained with this method are, respectively, 0.384 (95% CI: 1, 0.116) and 0.894 (95% CI: 1, 0.106).

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion
REFERENCES
 
Femoral trochlear dysplasia is one of the anomalies that could be observed on radiographs of patients with PI. On the basis of surgical observation, PI was described as a filling process involving the intercondylar groove (11). The origin of this process remains unknown. It could be due to bone production caused by excessive pressure over the trochlear groove during flexion of the knee. Brattström (6) described femoral trochlear dysplasia on conventional radiographs depicting the transverse view of the patellofemoral joint as a flattening of the trochlear angle. Later, Malghem and Maldague (7) and Dejour et al (8) studied the depth insufficiency of the trochlea as seen on lateral radiographs in patients with PI. By means of comparison with the transverse view of the patellofemoral joint, the lateral view of the knee makes it possible in theory to analyze the whole aspect of the patellofemoral joint and more particularly the proximal portion of the trochlea. This zone is very important to consider, since it is at this point that dysplasia occurs more frequently. Indeed, when the dislocating anomalies are moderate, patellar subluxation occurs on only the proximal portion of the patellofemoral joint, and thus only the first millimeters of the trochlea are subject to excessive pressure. Unfortunately, the lateral view is also not sufficient for analyzing the proximal portion of the trochlea since, at this point, the various lines that form the edges and bottom of the condyles become superimposed. For these reasons, other imaging techniques such as computed tomography were proposed to explore more accurately the patellofemoral joint in PI (9,12,13).

MR imaging is a very efficient tool for exploring the various types of internal knee derangement. This technique is particularly powerful for the study of the meniscus and ligaments. In PI, however, the value of MR imaging is not clear. Patellar tracking, which is abnormal in PI, was studied with kinematic imaging by numerous authors (14–17). In these studies, fast acquisition allows direct visualization of lateral subluxation of the patella that occurs during the first 30° of knee flexion. This kind of analysis is difficult to perform routinely, however, since it requires specific hardware and additional time. Koskinen and Kujala (18) and Koskinen et al (19) studied on MR images the distal insertion of the vastus medialis muscle, the relationships of the patella with the femoral trochlea, and the depth of the femoral trochlea. They found that none of these data help differentiate healthy volunteers from patients with patellar dislocation. Moreover, they found that measurement of the sulcus angle was not reproducible. In our opinion, these results confirm that study of the patellar position in relation to the femoral trochlea is not relevant when the knee is totally extended and, thus, not yet engaged. Their findings also confirm that trochlear angle, which is another way to assess femoral trochlear dysplasia, cannot be reliably measured since it varies as a function of the level at which it is measured.

We were interested in estimating trochlear dysplasia by measuring the LTI to provide a tool that can be used at the proximal portion of the trochlea. Our results showed that determination of this level is reproducible and is an accurate method with which to assess the diagnosis of femoral trochlear dysplasia. Measurements of the LTI in patients with PI demonstrated a lateral trochlear facet more horizontally oriented than normal in comparison with the bicondylar lines. In severe cases, the inclination of the lateral facet of the trochlea can be inversely oriented (LTI, <0°), a finding so evident that LTI measurement is not required to assess the dysplasia. When the anomaly is moderate, our results suggest that differentiation of patients with PI from those with internal knee derangement is possible with a threshold value of 11°. In our opinion, this reproducible method may be beneficial and contribute to the diagnosis of subclinical forms of PI.

In this preliminary study, measurement of LTI appeared to be a reproducible and accurate technique for differentiation of dislocating forms of PI from other forms of internal knee derangement. This technique is reproducible and represents a good diagnostic test. The threshold value of 11° for LTI should be tested in a larger population, including patients with subclinical forms of PI, to consider this value as the minimum to tolerate for a normal lateral trochlear facet inclination.


   FOOTNOTES
 
Abbreviations: LTI = lateral trochlear inclination, PI = patellar instability

Author contributions: Guarantors of integrity of entire study, Y.C., V.A.T.M.; study concepts, Y.C., D.D.; study design, Y.C.; definition of intellectual content, Y.C.; literature research, Y.C.; clinical studies, Y.C., D.D., B.M.; data acquisition and analysis, Y.C., O.F.; statistical analysis, H.A.; manuscript preparation and editing, Y.C.; manuscript review, Y.C., V.A.T.M.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion
 REFERENCES
 

  1. Grelsamer RP. Classifications of patellofemoral disorders. Am J Knee Surg 1997; 10:96-100.[Medline]
  2. MacNab L. recurrent dislocation of the patella. J Bone Joint Surg Am 1952; 34:957-976.[Abstract/Free Full Text]
  3. Fulkerson JP, Shea KP. Disorders of patellofemoral alignment. J Bone Joint Surg Am 1990; 72:1424-1429.[Free Full Text]
  4. Kummel BM. The diagnosis of patellofemoral derangements. Prim Care 1980; 7:199-216.[Medline]
  5. Wiberg G. Roentgenographic and anatomic studies on the femoropatellar joint, with special reference to chondromalacia patellae. Acta Orthop Scand 1941; 12:319-340.
  6. Brattström H. Shape of the intercondylar groove normally and in recurrent dislocation of the patella. Acta Scand 1964; 68(suppl):1-148.
  7. Malghem J, Maldague B. Depth insufficiency of the proximal trochlear groove on lateral radiographs of the knee: relation to patellar dislocation. Radiology 1989; 170:507-510.[Abstract/Free Full Text]
  8. Dejour H, Walch G, Nove-Josserand L, Guier C. Factors of patellar instability: an anatomic radiographic study. Knee Surg Sports Traumatol Arthrosc 1994; 2:219-226.[Medline]
  9. Walker C, Cassar-Pullicino VN, Vaisha R, McCall IW. The patello-femoral joint: a critical appraisal of its geometric assessment utilizing conventional axial radiography and computed arthro-tomography. Br J Radiol 1993; 66:755-761.[Abstract/Free Full Text]
  10. Bernageau J, Goutallier D, Larde D, Guérin L. L’obliquité de la joue externe de la trochlée fémorale. Encyclop Med Chir 1981; 30:39-42.
  11. Albee FH. The bone graft wedge in the treatment of habitual dislocation of the patella. Med Rec 1915; 88:257-259.
  12. Guzzanti V, Gigante A, Di Lazzaro A, Fabbriciani C. Patellofemoral malalignment in adolescents: computerized tomographic assessment with or without quadriceps contraction. Am J Sports Med 1994; 22:55-60.[Abstract/Free Full Text]
  13. Martinez S, Korobkin M, Fondren FB, Hedlund LW, Goldner JL. Computed tomography of the normal patellofemoral joint. Invest Radiol 1983; 18:249-253.[Medline]
  14. Shellock FG, Mink JH, Deutsch AL, Foo TK, Sullenberger P. Patellofemoral joint: identification of abnormalities with active-movement, "unloaded" versus "loaded" kinematic MR imaging techniques. Radiology 1993; 188:575-578.[Abstract/Free Full Text]
  15. Powers CM, Shellock FG, Pfaff M. Quantification of patellar tracking using kinematic MRI. J Magn Reson Imaging 1998; 8:724-732.[Medline]
  16. Brown SM, Bradley WG, Jr. Kinematic magnetic resonance imaging of the knee. Magn Reson Imaging Clin N Am 1994; 2:441-449.[Medline]
  17. Brossmann J, Muhle C, Schroder C, et al. Patellar tracking patterns during active and passive knee extension: evaluation with motion-triggered cine MR imaging. Radiology 1993; 187:205-212.[Abstract/Free Full Text]
  18. Koskinen SK, Kujala UM. Patellofemoral relationships and distal insertion of the vastus medialis muscle: a magnetic resonance imaging study in nonsymptomatic subjects and in patients with patellar dislocation. Arthroscopy 1992; 8:465-468.[Medline]
  19. Koskinen SK, Taimela S, Nelimarkka O, Komu M, Kujala UM. Magnetic resonance imaging of patellofemoral relationships. Skeletal Radiol 1993; 22:403-410.[Medline]



This article has been cited by other articles:


Home page
 Am J Sports MedHome page
C. S. Ahmad, M. McCarthy, J. A. Gomez, and B. E. Shubin Stein
The Moving Patellar Apprehension Test for Lateral Patellar Instability
Am. J. Sports Med., April 1, 2009; 37(4): 791 - 796.
[Abstract] [Full Text] [PDF]

Home page
 Rheumatology (Oxford)Home page
L. Kalichman, Y. Zhang, J. Niu, J. Goggins, D. Gale, D. T. Felson, and D. Hunter
The association between patellar alignment and patellofemoral joint osteoarthritis features an MRI study
Rheumatology, August 1, 2007; 46(8): 1303 - 1308.
[Abstract] [Full Text] [PDF]

Home page
 J Bone Joint Surg BrHome page
J. S. Mulford, C. J. Wakeley, and J. D. J. Eldridge
Assessment and management of chronic patellofemoral instability
J Bone Joint Surg Br, June 1, 2007; 89-B(6): 709 - 716.
[Abstract] [Full Text] [PDF]

Home page
 J Bone Joint Surg BrHome page
A. L. van Huyssteen, M. R. G. Hendrix, A. J. Barnett, C. J. Wakeley, and J. D. J. Eldridge
Cartilage-bone mismatch in the dysplastic trochlea: AN MRI STUDY
J Bone Joint Surg Br, May 1, 2006; 88-B(5): 688 - 691.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
D. A. Elias, L. M. White, and D. C. Fithian
Acute Lateral Patellar Dislocation at MR Imaging: Injury Patterns of Medial Patellar Soft-Tissue Restraints and Osteochondral Injuries of the Inferomedial Patella
Radiology, December 1, 2002; 225(3): 736 - 743.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carrillon, Y.
Right arrow Articles by Tran-Minh, V. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carrillon, Y.
Right arrow Articles by Tran-Minh, V. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE