1992, The British Journal of Radiology, 65, 662-667
Subluxation of the patella: evaluation of patellar articular cartilage with MR imaging By Katsuyuki Nakanishi, *Masahiro Inoue, Koushi Harada, Junpei Ikezoe, Takamichi Murakami, Hironobu Nakamura and Takahiro Kozuka Department of Radiology and * Department of Orthopaedic Surgery, Osaka University Medical School, 1-1-50, Fukushima, Fukushima-ku, Osaka, Japan {Received 19 August 1991, accepted 14 February 1992) Keywords: Subluxation of the patella, Patellar articular cartilage, MR imaging
Abstract. In patients with subluxation of the patella, injury of the patellar articular cartilage is frequently observed, and correct evaluation of this cartilage injury is extremely important for the management of these patients. Magnetic Resonance (MR) studies were performed on 102 patellofemoral (PF) joints of 51 patients with subluxation of the patella and 20 PF joints of 10 healthy volunteers. In 77 of the 102 PF joints with subluxation, arthroscopy and/or operation were performed. MR images were obtained with spin-echo and FLASH sequences, and para-axial images were obtained. We retrospectively analysed the MR findings of the 77 joints with special attention to the surface and thickness of the cartilage, and classified them into four grades. These MR grades were compared with the grades on arthroscopy, and the following results were obtained: MR grade 0, normal cartilage (n = 27, sensitivity 90.9%, specificity 74.2%); MR grade 1, thickening of the cartilage (n = 24, sensitivity 50%, specificity 89.1%); MR grade 2, surface irregularity of the cartilage (n = 20, sensitivity 85%, specificity 94.7%); MR grade 3, loss of the cartilage (n = 6, sensitivity 100%, specificity 100%). Although the early changes observed by arthroscopy were underestimated from the MR images, MR imaging proved to be extremely useful for evaluating moderately or advanced patellar cartilage injury.
Subluxation of the patella is a disorder of recurrent patellar pain, swelling and an unstable sensation of the patella. The onset of the disorder is usually at puberty. When patients are diagnosed as having subluxation of the patella, they usually have a history of giving way of the knees (Hughston, 1968; Merchant et al, 1974; Laurin et al, 1978; Aglietti et al, 1983; Install et al, 1983; Martinez et al, 1983; Inoue et al, 1988). It is very important for the treatment of subluxation of the patella to diagnose correctly the cartilage injury. However, before MR imaging, there was no method to visualize the patellar cartilage itself even when computed tomography (CT) was used. Therefore, invasive methods such as arthroscopy or CT with arthrography were necessary to evaluate the patellar articular cartilage (Boven et al, 1982a,b). Because it enables differentiation between cortical and cancellous bone, and hyaline cartilage and fibrocartilage, MR imaging has proved useful for examining the knee joints non-invasively. We undertook this study to determine whether MR imaging can accurately demonstrate the patellar articular cartilage, yield a correct grading, or replace arthroscopy. Materials and methods
MR studies were performed on 20 patellofemoral (PF) joints of 10 normal volunteers (age range 22-27 years; all female) as normal controls and 102 PF joints of 51 consecutive patients (age range 10-55 years; mean 19 years; 3 males, 48 females) who had a history of 662
giving way or dislocation of one or both PF joints and a radiographic diagnosis of subluxation of the patella. The roentgenographic criterion for diagnosing subluxation of the patella was a lateral PF angle of less than 0° on CT, with the knee in full extension (Inoue et al, 1988). From these 102 PF joints, we excluded four PF joints which had previously undergone arthroscopic debridement of cartilage, and 21 PF joints which had not undergone arthroscopy. As a result, 77 PF joints with subluxation were analysed. MR imaging was performed on a 1.5 T unit (Magnetom, Siemens) with a 10 cm planar surface coil. Patients were in the supine position with their knees in full extension. The planar surface coil was placed above the patella. Spin-echo (SE) images (Tx-weighted, spin density, and r2-weighted parameters) and FLASH images were obtained. T{ -weighted images were obtained with SE 400/26 (repetition time in milliseconds, echo time in milliseconds) pulse sequences and four excitations. The examination time was 6.88 min. Spin density and r2-weighted images were obtained with dual echo, SE 2000/26, 70 pulse sequences and one excitation. The examination time was 8.80 min. FLASH images were obtained with 320/15/flip angle 90° pulse sequences and four excitations. The examination time was 5.51 min. In all subjects, the image direction was perpendicular to the long axis of the patella. The section thickness was 4 mm with no gap, and the imaging matrix was 256 by 256 with FOV of 14-18 cm. MR images were prospectively interpreted by independent radiologists without knowledge of the results of The British Journal of Radiology, August 1992
Evaluation of patellar articular cartilage with MR imaging
(b) Figure 1. 25-year-old female with a normal appearance of the left patellar articular cartilage, (a) r r weighted image (SE400/26). The patellar articular cartilage (arrowhead) shows a homogeneous appearance, with a slightly higher intensity than skeletal muscle (arrow), (b) FLASH image (320/15, FA90). The patellar articular cartilage (white arrowhead) shows a relatively high intensity because of fat suppression in the bone marrow.
arthroscopy or operation, paying special attention to the surface and thickness of the cartilage. The greatest thickness was measured. If there was some discrepancy among the three observers, agreement was obtained by consensus. We then classified these MR findings into four grades (MR grade 0, normal; MR grade 1, thickening; MR grade 2, surface irregularity; MR grade 3, loss of the cartilage). We compared these MR grades with the arthroscopic grades. Arthroscopy was performed by an experienced orthopaedic surgeon, and the results were recorded by arthroscopic grading, which usually consisted of classification into four grades (arthroscopy (AS) grade 0, normal; AS grade 1, softening and fissure; AS grade 2, fragmentation; AS grade 3, exposure of the subchondral bone) (Table I).
Table I. MR findings and arthroscopic gradings (n = 77) MR grades
Athroscopic (AS) grades 0
1 2
Sensitivity Specificity
3
0: Normal appearance 10 17 1: Thickening 1 20 3 2: Surface irregularity 3 17 3: Loss of the cartilage 6
90.9% 50% 85% 100%
74.2% 89.1% 94.7% 100%
AS grades: grade 0, normal appearance; grade 1, softening and fissure; grade 2, fragmentation; grade 3, exposure of the subchondral bone.
Vol. 65, No. 776
Results
Healthy volunteers On r,-weighted images (Fig. la), the normal patellar articular cartilage was seen as a homogeneous region with a smooth contour. It had a slightly higher intensity than skeletal muscle. On FLASH images (Fig. lb), the normal patellar articular cartilage was demonstrated as a relatively high intensity area, because the intensity of the fat in the bone marrow was suppressed on this sequence. The thickness measured from the FLASH image of normal volunteers ranged from 2.8 to 4.4 mm, mean 3.56 mm. There were no significant differences between the right patellar cartilages and the left in each normal volunteer. Patients with subluxation of the patella Taking account of the thickness of cartilage in normal volunteers, we defined cartilage with a smooth surface and more than 5 mm thick as "thickening" (MR grade 1), and that with a smooth surface less than 5 mm thick as "normal" (MR grade 0). When irregularity of the surface of the cartilage was observed, we defined it as "surface irregularity" (MR grade 2) regardless of its thickness. If no articular cartilage could be seen on the MR image, we defined it as "loss of the cartilage" (MR grade 3). Using these criteria of MR imaging (Table I), 27 PF joints were diagnosed as MR grade 0, 24 as MR grade 1, 20 as MR grade 2, and six as MR grade 3. 10 (thickness range 2.8-4.7 mm; mean 3.54 mm) of the 27 PF joints visualized as MR grade 0 were also diagnosed as normal AS grade 0 by arthroscopy, but the other 17 (thickness range 2.9-4.8 mm; mean 4.13 mm) were 663
K. Nakanishi et al
Figure 2. 19-year-old female with thickening (MR grade 1) of the right patellar articular cartilage, (a) Tl-weighted image. The patellar articular cartilage is thick (white arrowhead). In this image, joint fluid is adjacent to the cartilage (arrow), and its intensity is similar to that of cartilage, (b) FLASH image. The patellar cartilage (white arrowhead) shows relatively high intensity, while the joint fluid shows low intensity (arrow), allowing them to be easily differentiated. In this case, the thickness of the cartilage is 6.9 mm. (c) Arthroscopy. The surface of the cartilage is depressed by the probe (AS grade 1).
recognized as having softening and fissure (AS grade 1) during arthroscopy. 20 of the 24 PF joints diagnosed as "thickening" (MR grade 1) by MR images were recognized as AS grade 1 (Fig. 2), three as AS grade 2 and the remaining one as AS grade 0 by arthroscopy. The mean thickness of these 24 patellar cartilages was 5.93 mm on the MR images. 17 (range of thickness on MR images 2.9-5.4 mm; mean 4.59 mm) of the 20 PF joints visualized as MR grade 2 were recognized as having fragmentation (AS grade 2) on arthroscopy (Fig. 3). The remaining three (thickness range 5.2-6.5 mm; mean 5.97 mm) were diagnosed as AS grade 1. All six PF 664
joints diagnosed as MR grade 3 were observed as AS grade 3 in arthroscopy (Fig. 4). Discussion
Cartilage may be classified as hyaline or fibrocartilage. The patellar articular cartilage is hyaline. Cartilage is composed of chondrocytes and matrix. The matrix is composed of collagen fibres and ground substance, which contains various mucopolysaccharides. Hyaline cartilage contains a lot of mucopolysaccharide, which reflects the signal intensity of articular cartilage on MR imaging (Bentley & Dowd, 1984; Shahriaree, 1985). The British Journal of Radiology, August 1992
Evaluation of patellar articular cartilage with MR imaging
(b)
Figure 3. 19-year-old female with surface irregularity (MR grade 2) of the right knee, (a) FLASH image. From the central ridge to the medial portion, the surface of the articular cartilage is irregular (white arrow), and the cartilage layer is inhomogeneous (white arrowhead), (b) Operative findings. The lateral portion of articular cartilage is almost intact (arrow), but there are many fragmentations of the medial portion (arrowhead). This change corresponds to AS grade 2
Hyaline cartilage shows slightly higher intensity than skeletal muscle on Tx-weighted images (Li et al, 1984; Reicher et al, 1985a,b; Beltran et al, 1986; Konig et al, 1987; Yulish et al, 1987). All 20 PF joints of 10 normal volunteers showed normal alignment. The thickness was almost equal in all
portions of the cartilage. Subluxation of the patella is usually preceded by malalignment of the PF joint. In many cases, the lateral portion of the articular cartilage appears soft on arthroscopy. In this condition, the ground substance in the cartilage has degenerated. Actually, mucopolysaccharide degenerates, and carti-
(a)
Figure 4. 50-year-old female with loss of the cartilage (MR grade 3) of the left knee, (a) rrweighted image. The lateral portion of the patellar articular cartilage is lost and the subchondral bone is thickened (arrow). The medial portion of the articular cartilage remains, but its surface is irregular (arrowhead), (b) Arthroscopy. In the lateral portion of the petalla, the subchondral bone is exposed (arrow) (AS grade 3). Vol.y-65, No. 776
665
K. Nakanishi et al
(b) Figure 5. 19-year-old female with thickening (MR grade 1) of the left patellar articular cartilage, (a) ^-weighted image. The thickness of the cartilage is 6.2 mm and the surface appears smooth. A low-intensity band is seen in the deep layer of the cartilage (arrow), (b) Operative findings. Many large fragmentations are seen in the cartilage (arrowhead) (AS grade 2).
lage containing a large amount of water as well as vertical collagen fibres is disrupted (Bentley & Dowd, 1984; Shahriaree, 1985). If patients with malalignment of a PF joint suffer from giving way or dislocation, subtle fissures form on the surface of the cartilage. As these episodes repeat, larger fragmentations occur and the surface becomes irregular. With repeated dislocation over a long period, the subchondral bone is finally exposed (condition is observed as exposure of the subchondral bone by arthroscopy). From the analysis of our results, AS grade 1 lesions were underestimated on MR imaging. 17 PF joints whose thickness did not exceed 5 mm on the MR images (MR grade 0) were evaluated as having softening on arthroscopy. In these 17 PF joints, the mean thickness was 4.13 mm, which was slightly thicker than 10 PF joints seen to be normal (AS grade 0) on arthroscopy. Furthermore, for these 17 PF joints, the mean age was 16.8 years, which was younger than the mean of 23.3 years for the normal volunteers. Since the cartilage thickness of patients at puberty (between 10 and 19 years) might be physiologically thinner than that of normal volunteers, the criteria for "thickening" might be less than 5 mm. Three of the 24 PF joints observed to be thickened on MR images (MR grade 1) showed fragmentation on arthroscopy (AS grade 2). In these three cases, an inhomogeneous appearance was clearly observed around the central ridge of the patella (Fig. 5). The other 21 PF joints diagnosed as MR grade 1 had a relatively homogeneous appearance. Once an obviously inhomogeneous appearance is seen, more advanced changes must be considered. Fragmentation on arthroscopy (AS grade 2) was accurately diagnosed by MR imaging. Although 20 PF joints were observed to have an irregular surface by MR imaging, three of them proved to have many deep fissures around the central ridge of the patella and were diagnosed as AS grade 1. When an irregular surface is observed in this portion by 666
MR imaging, fissure formation or fragmentation must be considered (MR grade 3). Loss of the articular cartilage (MR grade 3) corresponds to exposure of the subchondral bone (AS grade 3) by arthroscopy. All patients in this group were over 30 years in age and had a long history of giving way or dislocation of the patella. In some cases, it is difficult to identify the cartilage itself on Tx-weighted images. In such cases, a FLASH image with a flip angle of 90° is useful. This is because when the fluid in the joint capsule is adjacent to the cartilage, it is difficult to differentiate the cartilage from the joint fluid on ^-weighted images. With FLASH images with a flip angle 90°, however, cartilage shows a relatively high intensity (Konig et al, 1987) while the joint fluid shows a low intensity. For this reason, FLASH images with a flip angle of 90° provide better anatomic detail than rrweighted images. However, Tx and spin-density images are useful for visualizing subtle changes in the cartilage layer. On r2-weighted images, the cartilage appears as a low intensity region, while the joint fluid shows high intensity (Beltran et al, 1986); however, the imaging time is long. We conclude that the combination of Tx -weighted images and FLASH images with a flip angle of 90° is extremely effective for detecting abnormalities of the patellar articular cartilage. In summary, MRI can accurately determine pathological changes in the patellar articular cartilage in patients with subluxation of the patella, especially advanced changes. Cartilage with a thickness exceeding 5 mm (MR grade 1) corresponds to softening and fissure seen in arthroscopy (AS grade 1). An irregular surface observed on MR imaging (MR grade 2) corresponds to fragmentation (AS grade 2) on arthroscopy. Loss of the cartilage observed in MR imaging (MR grade 3) corresponds to exposure of the subchondral bone (AS grade 3). The combination of rrweighted The British Journal of Radiology, August 1992
Evaluation of patellar articular cartilage with MR imaging
images and FLASH images with a flip angle of 90° is extremely effective for detecting abnormalities of patellar articular cartilage.
KONIG, H., SAUTER, R., DEIMLING, M. & VOGT, M., 1987.
Cartilage disorders: comparison of spin-echo, CHESS, and FLASH sequence MR images. Radiology, 164, 753-758. LAURIN, C. A., LEVESQUE, H. P., DUSSAULT, R., LABELLE, H. &
PEIDES, J. P., 1978. The abnormal lateral patellofemoral angle. Journal of Bone and Joint Surgery, 60, 55-60.
References AGLIETTI, P., INSTALL, J. N. & CERULLI, G., 1983. Patellar pain
Li, K. C , HENKELMAN, R. M., POON, P. Y. & RUBINSTEIN, J.,
and incongruence. Measurement of incongruence. Clinical Orthopaedics and Related Research, 176, 217-224.
1984. MR imaging of the normal knee. Journal of Computer Assisted Tomography, 8(6), 1147-1154.
BELTRAN, J., NOTO, A. M., HERMAN, L. J., MOSURE, J. C ,
MARTINEZ, S., KOROBKIN, M., FONDREN, F. B., HEDLUND,
BURK, J. M. & CHRISTOFORIDIS, A. J., 1986. Joint effusions:
L. W. & GOLDNER, J. L., 1983. Diagnosis of patellofemoral malalignment by computed tomography. Journal of Computer Assisted Tomography, 7, 1050-1053.
MR imaging. Radiology, 158, 133-137. BENTLEY, G. & DOWD, G., 1984. Current concept of etiology and treatment of chondromalacia patellae. Clinical Orthopaedics and Related Research, 189, 209-228. BOVEN, F., BELLEMANS, M.-A., GEURTS, J. & POTVLIEGE, R.,
1982a. A Comparative study of the patello-femoral joint on axial roentgenogram, axial arthrogram, and computed tomography following arthrography. Skeletal Radiology, 8, 179-181. BOVEN, F., BELLEMANS, M.-A., GEURTS, J., BOECK, H. D. &
POTVLIEGE, R., 1982b. The value of computed tomography scanning in Chrondromalacia patellae. Skeletal Radiology, 8, 183-185. HUGHSTON, J. C , 1968. Subluxation of the patella. Journal of Bone and Joint Surgery, 50, 1003-1026. INOUE, M., SHINO, K., HIROSE, H., HIROBE, S. & ONO, K., 1988.
Subluxation of the patella. Journal of Bone and Joint Surgery, 70, 1331-1337. INSTALL, J. N., AGLIETTI, P. & TRIA, A. J., JR., 1983. Patellar
pain and incongruence. Clinical application. Orthopaedics and Related Research, 176, 225-232.
Vol. 65, No. 776
Clinical
MERCHANT, A. C , MERCER, R. L., JACOBSON, R. H. & COOL,
C. R., 1974. Roentgenographic analysis of patellofemoral congruence. Journal of Bone and Joint Surgery, 56, 1391-1396. REICHER, M. A., RAUSCHNING, W., GOLD, R. H., BASSETT, L. W., LUFKIN, R. B. & GLEN, W., JR., 1985a.
High-resolution magnetic resonance imaging of the knee joint: normal anatomy. American Journal of Roentgenology, 145, 895-902. REICHER, M. A., BASSETT, L. W. & GOLD, R. H., 1985b.
High-resolution magnetic resonance imaging of the knee joint: pathologic correlations. American Journal of Roentgenology, 145, 903-909. SHAHRIAREE, H., 1985. Chondromalacia. Contemporary Orthopaedics, 11, 27-39. YULISH, B. S., MONTANEZ, J., GOODFELLOW, D . B . , BRYAN, P. J., MULOPULOS, G. P. & MODIC, M. T., 1987.
Chondromalacia patellae: assessment with MR imaging. Radiology, 164, 763-766.
667
Subluxation of the patella: evaluation of patellar articular cartilage with MR imaging By Katsuyuki Nakanishi, *Masahiro Inoue, Koushi Harada, Junpei Ikezoe, Takamichi Murakami, Hironobu Nakamura and Takahiro Kozuka Department of Radiology and * Department of Orthopaedic Surgery, Osaka University Medical School, 1-1-50, Fukushima, Fukushima-ku, Osaka, Japan {Received 19 August 1991, accepted 14 February 1992) Keywords: Subluxation of the patella, Patellar articular cartilage, MR imaging
Abstract. In patients with subluxation of the patella, injury of the patellar articular cartilage is frequently observed, and correct evaluation of this cartilage injury is extremely important for the management of these patients. Magnetic Resonance (MR) studies were performed on 102 patellofemoral (PF) joints of 51 patients with subluxation of the patella and 20 PF joints of 10 healthy volunteers. In 77 of the 102 PF joints with subluxation, arthroscopy and/or operation were performed. MR images were obtained with spin-echo and FLASH sequences, and para-axial images were obtained. We retrospectively analysed the MR findings of the 77 joints with special attention to the surface and thickness of the cartilage, and classified them into four grades. These MR grades were compared with the grades on arthroscopy, and the following results were obtained: MR grade 0, normal cartilage (n = 27, sensitivity 90.9%, specificity 74.2%); MR grade 1, thickening of the cartilage (n = 24, sensitivity 50%, specificity 89.1%); MR grade 2, surface irregularity of the cartilage (n = 20, sensitivity 85%, specificity 94.7%); MR grade 3, loss of the cartilage (n = 6, sensitivity 100%, specificity 100%). Although the early changes observed by arthroscopy were underestimated from the MR images, MR imaging proved to be extremely useful for evaluating moderately or advanced patellar cartilage injury.
Subluxation of the patella is a disorder of recurrent patellar pain, swelling and an unstable sensation of the patella. The onset of the disorder is usually at puberty. When patients are diagnosed as having subluxation of the patella, they usually have a history of giving way of the knees (Hughston, 1968; Merchant et al, 1974; Laurin et al, 1978; Aglietti et al, 1983; Install et al, 1983; Martinez et al, 1983; Inoue et al, 1988). It is very important for the treatment of subluxation of the patella to diagnose correctly the cartilage injury. However, before MR imaging, there was no method to visualize the patellar cartilage itself even when computed tomography (CT) was used. Therefore, invasive methods such as arthroscopy or CT with arthrography were necessary to evaluate the patellar articular cartilage (Boven et al, 1982a,b). Because it enables differentiation between cortical and cancellous bone, and hyaline cartilage and fibrocartilage, MR imaging has proved useful for examining the knee joints non-invasively. We undertook this study to determine whether MR imaging can accurately demonstrate the patellar articular cartilage, yield a correct grading, or replace arthroscopy. Materials and methods
MR studies were performed on 20 patellofemoral (PF) joints of 10 normal volunteers (age range 22-27 years; all female) as normal controls and 102 PF joints of 51 consecutive patients (age range 10-55 years; mean 19 years; 3 males, 48 females) who had a history of 662
giving way or dislocation of one or both PF joints and a radiographic diagnosis of subluxation of the patella. The roentgenographic criterion for diagnosing subluxation of the patella was a lateral PF angle of less than 0° on CT, with the knee in full extension (Inoue et al, 1988). From these 102 PF joints, we excluded four PF joints which had previously undergone arthroscopic debridement of cartilage, and 21 PF joints which had not undergone arthroscopy. As a result, 77 PF joints with subluxation were analysed. MR imaging was performed on a 1.5 T unit (Magnetom, Siemens) with a 10 cm planar surface coil. Patients were in the supine position with their knees in full extension. The planar surface coil was placed above the patella. Spin-echo (SE) images (Tx-weighted, spin density, and r2-weighted parameters) and FLASH images were obtained. T{ -weighted images were obtained with SE 400/26 (repetition time in milliseconds, echo time in milliseconds) pulse sequences and four excitations. The examination time was 6.88 min. Spin density and r2-weighted images were obtained with dual echo, SE 2000/26, 70 pulse sequences and one excitation. The examination time was 8.80 min. FLASH images were obtained with 320/15/flip angle 90° pulse sequences and four excitations. The examination time was 5.51 min. In all subjects, the image direction was perpendicular to the long axis of the patella. The section thickness was 4 mm with no gap, and the imaging matrix was 256 by 256 with FOV of 14-18 cm. MR images were prospectively interpreted by independent radiologists without knowledge of the results of The British Journal of Radiology, August 1992
Evaluation of patellar articular cartilage with MR imaging
(b) Figure 1. 25-year-old female with a normal appearance of the left patellar articular cartilage, (a) r r weighted image (SE400/26). The patellar articular cartilage (arrowhead) shows a homogeneous appearance, with a slightly higher intensity than skeletal muscle (arrow), (b) FLASH image (320/15, FA90). The patellar articular cartilage (white arrowhead) shows a relatively high intensity because of fat suppression in the bone marrow.
arthroscopy or operation, paying special attention to the surface and thickness of the cartilage. The greatest thickness was measured. If there was some discrepancy among the three observers, agreement was obtained by consensus. We then classified these MR findings into four grades (MR grade 0, normal; MR grade 1, thickening; MR grade 2, surface irregularity; MR grade 3, loss of the cartilage). We compared these MR grades with the arthroscopic grades. Arthroscopy was performed by an experienced orthopaedic surgeon, and the results were recorded by arthroscopic grading, which usually consisted of classification into four grades (arthroscopy (AS) grade 0, normal; AS grade 1, softening and fissure; AS grade 2, fragmentation; AS grade 3, exposure of the subchondral bone) (Table I).
Table I. MR findings and arthroscopic gradings (n = 77) MR grades
Athroscopic (AS) grades 0
1 2
Sensitivity Specificity
3
0: Normal appearance 10 17 1: Thickening 1 20 3 2: Surface irregularity 3 17 3: Loss of the cartilage 6
90.9% 50% 85% 100%
74.2% 89.1% 94.7% 100%
AS grades: grade 0, normal appearance; grade 1, softening and fissure; grade 2, fragmentation; grade 3, exposure of the subchondral bone.
Vol. 65, No. 776
Results
Healthy volunteers On r,-weighted images (Fig. la), the normal patellar articular cartilage was seen as a homogeneous region with a smooth contour. It had a slightly higher intensity than skeletal muscle. On FLASH images (Fig. lb), the normal patellar articular cartilage was demonstrated as a relatively high intensity area, because the intensity of the fat in the bone marrow was suppressed on this sequence. The thickness measured from the FLASH image of normal volunteers ranged from 2.8 to 4.4 mm, mean 3.56 mm. There were no significant differences between the right patellar cartilages and the left in each normal volunteer. Patients with subluxation of the patella Taking account of the thickness of cartilage in normal volunteers, we defined cartilage with a smooth surface and more than 5 mm thick as "thickening" (MR grade 1), and that with a smooth surface less than 5 mm thick as "normal" (MR grade 0). When irregularity of the surface of the cartilage was observed, we defined it as "surface irregularity" (MR grade 2) regardless of its thickness. If no articular cartilage could be seen on the MR image, we defined it as "loss of the cartilage" (MR grade 3). Using these criteria of MR imaging (Table I), 27 PF joints were diagnosed as MR grade 0, 24 as MR grade 1, 20 as MR grade 2, and six as MR grade 3. 10 (thickness range 2.8-4.7 mm; mean 3.54 mm) of the 27 PF joints visualized as MR grade 0 were also diagnosed as normal AS grade 0 by arthroscopy, but the other 17 (thickness range 2.9-4.8 mm; mean 4.13 mm) were 663
K. Nakanishi et al
Figure 2. 19-year-old female with thickening (MR grade 1) of the right patellar articular cartilage, (a) Tl-weighted image. The patellar articular cartilage is thick (white arrowhead). In this image, joint fluid is adjacent to the cartilage (arrow), and its intensity is similar to that of cartilage, (b) FLASH image. The patellar cartilage (white arrowhead) shows relatively high intensity, while the joint fluid shows low intensity (arrow), allowing them to be easily differentiated. In this case, the thickness of the cartilage is 6.9 mm. (c) Arthroscopy. The surface of the cartilage is depressed by the probe (AS grade 1).
recognized as having softening and fissure (AS grade 1) during arthroscopy. 20 of the 24 PF joints diagnosed as "thickening" (MR grade 1) by MR images were recognized as AS grade 1 (Fig. 2), three as AS grade 2 and the remaining one as AS grade 0 by arthroscopy. The mean thickness of these 24 patellar cartilages was 5.93 mm on the MR images. 17 (range of thickness on MR images 2.9-5.4 mm; mean 4.59 mm) of the 20 PF joints visualized as MR grade 2 were recognized as having fragmentation (AS grade 2) on arthroscopy (Fig. 3). The remaining three (thickness range 5.2-6.5 mm; mean 5.97 mm) were diagnosed as AS grade 1. All six PF 664
joints diagnosed as MR grade 3 were observed as AS grade 3 in arthroscopy (Fig. 4). Discussion
Cartilage may be classified as hyaline or fibrocartilage. The patellar articular cartilage is hyaline. Cartilage is composed of chondrocytes and matrix. The matrix is composed of collagen fibres and ground substance, which contains various mucopolysaccharides. Hyaline cartilage contains a lot of mucopolysaccharide, which reflects the signal intensity of articular cartilage on MR imaging (Bentley & Dowd, 1984; Shahriaree, 1985). The British Journal of Radiology, August 1992
Evaluation of patellar articular cartilage with MR imaging
(b)
Figure 3. 19-year-old female with surface irregularity (MR grade 2) of the right knee, (a) FLASH image. From the central ridge to the medial portion, the surface of the articular cartilage is irregular (white arrow), and the cartilage layer is inhomogeneous (white arrowhead), (b) Operative findings. The lateral portion of articular cartilage is almost intact (arrow), but there are many fragmentations of the medial portion (arrowhead). This change corresponds to AS grade 2
Hyaline cartilage shows slightly higher intensity than skeletal muscle on Tx-weighted images (Li et al, 1984; Reicher et al, 1985a,b; Beltran et al, 1986; Konig et al, 1987; Yulish et al, 1987). All 20 PF joints of 10 normal volunteers showed normal alignment. The thickness was almost equal in all
portions of the cartilage. Subluxation of the patella is usually preceded by malalignment of the PF joint. In many cases, the lateral portion of the articular cartilage appears soft on arthroscopy. In this condition, the ground substance in the cartilage has degenerated. Actually, mucopolysaccharide degenerates, and carti-
(a)
Figure 4. 50-year-old female with loss of the cartilage (MR grade 3) of the left knee, (a) rrweighted image. The lateral portion of the patellar articular cartilage is lost and the subchondral bone is thickened (arrow). The medial portion of the articular cartilage remains, but its surface is irregular (arrowhead), (b) Arthroscopy. In the lateral portion of the petalla, the subchondral bone is exposed (arrow) (AS grade 3). Vol.y-65, No. 776
665
K. Nakanishi et al
(b) Figure 5. 19-year-old female with thickening (MR grade 1) of the left patellar articular cartilage, (a) ^-weighted image. The thickness of the cartilage is 6.2 mm and the surface appears smooth. A low-intensity band is seen in the deep layer of the cartilage (arrow), (b) Operative findings. Many large fragmentations are seen in the cartilage (arrowhead) (AS grade 2).
lage containing a large amount of water as well as vertical collagen fibres is disrupted (Bentley & Dowd, 1984; Shahriaree, 1985). If patients with malalignment of a PF joint suffer from giving way or dislocation, subtle fissures form on the surface of the cartilage. As these episodes repeat, larger fragmentations occur and the surface becomes irregular. With repeated dislocation over a long period, the subchondral bone is finally exposed (condition is observed as exposure of the subchondral bone by arthroscopy). From the analysis of our results, AS grade 1 lesions were underestimated on MR imaging. 17 PF joints whose thickness did not exceed 5 mm on the MR images (MR grade 0) were evaluated as having softening on arthroscopy. In these 17 PF joints, the mean thickness was 4.13 mm, which was slightly thicker than 10 PF joints seen to be normal (AS grade 0) on arthroscopy. Furthermore, for these 17 PF joints, the mean age was 16.8 years, which was younger than the mean of 23.3 years for the normal volunteers. Since the cartilage thickness of patients at puberty (between 10 and 19 years) might be physiologically thinner than that of normal volunteers, the criteria for "thickening" might be less than 5 mm. Three of the 24 PF joints observed to be thickened on MR images (MR grade 1) showed fragmentation on arthroscopy (AS grade 2). In these three cases, an inhomogeneous appearance was clearly observed around the central ridge of the patella (Fig. 5). The other 21 PF joints diagnosed as MR grade 1 had a relatively homogeneous appearance. Once an obviously inhomogeneous appearance is seen, more advanced changes must be considered. Fragmentation on arthroscopy (AS grade 2) was accurately diagnosed by MR imaging. Although 20 PF joints were observed to have an irregular surface by MR imaging, three of them proved to have many deep fissures around the central ridge of the patella and were diagnosed as AS grade 1. When an irregular surface is observed in this portion by 666
MR imaging, fissure formation or fragmentation must be considered (MR grade 3). Loss of the articular cartilage (MR grade 3) corresponds to exposure of the subchondral bone (AS grade 3) by arthroscopy. All patients in this group were over 30 years in age and had a long history of giving way or dislocation of the patella. In some cases, it is difficult to identify the cartilage itself on Tx-weighted images. In such cases, a FLASH image with a flip angle of 90° is useful. This is because when the fluid in the joint capsule is adjacent to the cartilage, it is difficult to differentiate the cartilage from the joint fluid on ^-weighted images. With FLASH images with a flip angle 90°, however, cartilage shows a relatively high intensity (Konig et al, 1987) while the joint fluid shows a low intensity. For this reason, FLASH images with a flip angle of 90° provide better anatomic detail than rrweighted images. However, Tx and spin-density images are useful for visualizing subtle changes in the cartilage layer. On r2-weighted images, the cartilage appears as a low intensity region, while the joint fluid shows high intensity (Beltran et al, 1986); however, the imaging time is long. We conclude that the combination of Tx -weighted images and FLASH images with a flip angle of 90° is extremely effective for detecting abnormalities of the patellar articular cartilage. In summary, MRI can accurately determine pathological changes in the patellar articular cartilage in patients with subluxation of the patella, especially advanced changes. Cartilage with a thickness exceeding 5 mm (MR grade 1) corresponds to softening and fissure seen in arthroscopy (AS grade 1). An irregular surface observed on MR imaging (MR grade 2) corresponds to fragmentation (AS grade 2) on arthroscopy. Loss of the cartilage observed in MR imaging (MR grade 3) corresponds to exposure of the subchondral bone (AS grade 3). The combination of rrweighted The British Journal of Radiology, August 1992
Evaluation of patellar articular cartilage with MR imaging
images and FLASH images with a flip angle of 90° is extremely effective for detecting abnormalities of patellar articular cartilage.
KONIG, H., SAUTER, R., DEIMLING, M. & VOGT, M., 1987.
Cartilage disorders: comparison of spin-echo, CHESS, and FLASH sequence MR images. Radiology, 164, 753-758. LAURIN, C. A., LEVESQUE, H. P., DUSSAULT, R., LABELLE, H. &
PEIDES, J. P., 1978. The abnormal lateral patellofemoral angle. Journal of Bone and Joint Surgery, 60, 55-60.
References AGLIETTI, P., INSTALL, J. N. & CERULLI, G., 1983. Patellar pain
Li, K. C , HENKELMAN, R. M., POON, P. Y. & RUBINSTEIN, J.,
and incongruence. Measurement of incongruence. Clinical Orthopaedics and Related Research, 176, 217-224.
1984. MR imaging of the normal knee. Journal of Computer Assisted Tomography, 8(6), 1147-1154.
BELTRAN, J., NOTO, A. M., HERMAN, L. J., MOSURE, J. C ,
MARTINEZ, S., KOROBKIN, M., FONDREN, F. B., HEDLUND,
BURK, J. M. & CHRISTOFORIDIS, A. J., 1986. Joint effusions:
L. W. & GOLDNER, J. L., 1983. Diagnosis of patellofemoral malalignment by computed tomography. Journal of Computer Assisted Tomography, 7, 1050-1053.
MR imaging. Radiology, 158, 133-137. BENTLEY, G. & DOWD, G., 1984. Current concept of etiology and treatment of chondromalacia patellae. Clinical Orthopaedics and Related Research, 189, 209-228. BOVEN, F., BELLEMANS, M.-A., GEURTS, J. & POTVLIEGE, R.,
1982a. A Comparative study of the patello-femoral joint on axial roentgenogram, axial arthrogram, and computed tomography following arthrography. Skeletal Radiology, 8, 179-181. BOVEN, F., BELLEMANS, M.-A., GEURTS, J., BOECK, H. D. &
POTVLIEGE, R., 1982b. The value of computed tomography scanning in Chrondromalacia patellae. Skeletal Radiology, 8, 183-185. HUGHSTON, J. C , 1968. Subluxation of the patella. Journal of Bone and Joint Surgery, 50, 1003-1026. INOUE, M., SHINO, K., HIROSE, H., HIROBE, S. & ONO, K., 1988.
Subluxation of the patella. Journal of Bone and Joint Surgery, 70, 1331-1337. INSTALL, J. N., AGLIETTI, P. & TRIA, A. J., JR., 1983. Patellar
pain and incongruence. Clinical application. Orthopaedics and Related Research, 176, 225-232.
Vol. 65, No. 776
Clinical
MERCHANT, A. C , MERCER, R. L., JACOBSON, R. H. & COOL,
C. R., 1974. Roentgenographic analysis of patellofemoral congruence. Journal of Bone and Joint Surgery, 56, 1391-1396. REICHER, M. A., RAUSCHNING, W., GOLD, R. H., BASSETT, L. W., LUFKIN, R. B. & GLEN, W., JR., 1985a.
High-resolution magnetic resonance imaging of the knee joint: normal anatomy. American Journal of Roentgenology, 145, 895-902. REICHER, M. A., BASSETT, L. W. & GOLD, R. H., 1985b.
High-resolution magnetic resonance imaging of the knee joint: pathologic correlations. American Journal of Roentgenology, 145, 903-909. SHAHRIAREE, H., 1985. Chondromalacia. Contemporary Orthopaedics, 11, 27-39. YULISH, B. S., MONTANEZ, J., GOODFELLOW, D . B . , BRYAN, P. J., MULOPULOS, G. P. & MODIC, M. T., 1987.
Chondromalacia patellae: assessment with MR imaging. Radiology, 164, 763-766.
667
