Patella Alta: Lack of Correlation Between Patellotrochlear Cartilage Congruence and Commonly Used Patellar Height Ratios

Musculoskeletal Imaging Original Research Ali et al. Patella Alta Musculoskeletal Imaging Original Research Syed A. Ali 1 Robert Helmer Michael R. Terk Ali SA, Helmer R, Terk MR Keywords: Insall-Salvati index, MRI, patella alta, patellar height ratio, patellofemoral articular cartilage, patellotrochlear index DOI:10.2214/AJR.09.2729 Received March 12, 2009; accepted after revision May 5, 2009. 1 All authors: Department of Radiology, Emory University, 59 Executive Park S, 4th Fl., Ste. 4009, Atlanta, GA 30329. Address correspondence to S. A. Ali (saali4@emory.edu). AJR 2009; 193:1361 1366 0361 803X/09/1935 1361 American Roentgen Ray Society Patella Alta: Lack of Correlation Between Patellotrochlear Cartilage Congruence and Commonly Used Patellar Height Ratios OBJECTIVE. The purposes of this study were to assess the degree of patellotrochlear chondral overlap (patellotrochlear index), correlate it with the Insall-Salvati and modified Insall-Salvati indexes, and determine the association between these measurements and patellofemoral chondral defects. MATERIALS AND METHODS. Sagittal 1.5-T and 3-T MR images of 100 consecutively registered patients with symptoms were analyzed, and the Insall-Salvati index, modified Insall-Salvati index, patellotrochlear index, and patellophyseal index (ratio of the height of patella above the physeal line to the length of the patellar articular cartilage) were calculated. The upper and lower limits of 2 SDs were used to define patella alta and baja, and the correlation coefficient curves were plotted to compare techniques. The indexes in normal knees were compared with those in knees with severe chondral defects. RESULTS. The mean patellotrochlear index was 0.49 ± 0.15 (SD) (range, 0 0.88). On the basis of calculation of 2 SDs, patella alta was determined to have a patellotrochlear index less than 0.18 and patella baja, an index greater than 0.80. Weak correlation was found between the measured patellotrochlear index and Insall-Salvati index (r = 0.224) and between the patellotrochlear index and modified Insall-Salvati index (r = 0.073). A strong correlation was found between the patellotrochlear index and patellophyseal index (r = 0.813). A statistically significant (p < 0.05) difference in the modified Insall-Salvati index and patellophyseal index was found between knees with normal and those with severe cartilage defects. CONCLUSION. Our results indicate that the commonly used Insall-Salvati and modified Insall-Salvati indexes do not correlate with patellotrochlear articular cartilage congruence. We did find an association between the modified Insall-Salvati and patellophyseal indexes and the presence of severe chondral defects. T he position of the patella in the vertical plane has great clinical relevance. A high-riding patella, or patella alta, is considered a predisposing factor to the development of patellofemoral pain [1, 2]. It also is associated with recurrent dislocation of the patella, chondromalacia patellae, and joint effusion [3 5]. Therefore, assessment of patellar position is essential in evaluation of the anatomic alignment of the knee, especially in patients with anterior knee pain. Several methods are used to determine the presence of patella alta. Insall and Salvati [6] were the first to describe a method of establishing patellar height on the basis of the ratio of the length of the patellar tendon to the diagonal length of the patella on lateral radiographs. Various techniques, including those by Blackburne and Peel [7], Caton et al. [8], and de Carvalho et al. [9], have since been developed in the attempt to classify patellar position. In the modified Insall-Salvati ratio, patellar bony length is replaced with patellar articular cartilage length [10, 11]. The Insall- Salvati index, however, is the most widely accepted method because of its lack of dependence on the degree of knee flexion and its established MRI criteria [12, 13]. A comparison of the aforementioned radiographic methods has produced a high frequency of differing results in determination of patellar height. The differences have caused many observers to believe that classification of patella alta depends heavily on the ratio chosen [14]. Although patella alta is defined simply as a patella that rides high in the femoral trochlea, indexes have been developed to describe patellar position in relation to the proximal tibia. MRI has shown AJR:193, November 2009 1361

Ali et al. that abnormal contact between the patella and the femur due to superior displacement of the patella can produce malalignment of the patellofemoral joint [15]. This finding has led many clinicians to question the utility of commonly used ratios, such as the Insall-Salvati index and modified Insall-Salvati index, for accurate assessment of patellar position and evaluation of premorbid anatomic conditions. Calculation of the patellotrochlear index (Fig. 1C) described by Biedert and Albrecht [15] is the only technique with which patellar position can be determined by direct assessment of the chondral articulation. Because it is an indirect assessment of surface contact area an important factor in predicting the presence of pathologic changes in the patellofemoral region this index may replace the widely used Insall-Salvati and modified Insall-Salvati indexes as a measure of premorbid anatomic conditions. To our knowledge, in no previous investigation has the correlation between patellotrochlear index and commonly used patellar indexes been determined with MRI. For this purpose, our goals were to determine the MRI criteria for patella alta using the patellotrochlear, Insall- Salvati, and modified Insall-Salvati indexes; compare the Insall-Salvati and modified Insall-Salvati indexes with the patellotrochlear index; and assess the association between each index and the presence of patellofemoral chondral defects. A newer index known as the patellophyseal index was assessed as another way to classify patellar height. It has been observed [10] that the femoral physeal line on sagittal MR images approximates the proximal end of the trochlear cartilage. Unlike articular cartilage overlap, the femoral physeal line is visible on both MR images and radiographs. Therefore, it may be feasible to use the patellophyseal index on both types of images. We therefore studied the patellophyseal index and its correlation with the patellotrochlear index. Materials and Methods Patient Selection MR images of 103 consecutively registered patients with symptoms were analyzed. The patients arrived at our institution with different knee symptoms and underwent imaging for various conditions (Table 1). Patients with bipartite patella, patellar fractures, or previous surgery were excluded from the study. Three patients who had undergone anterior cruciate ligament surgery were excluded on the basis of these criteria. The mean patient age was 39.5 years (range, 12 88 years), and the sex distribution was 45 women and girls and 55 men and boys. Approval from our institutional review board was obtained for this retrospective chart and report review without the requirement for individual informed consent. MRI MRI was performed with either a 1.5-T system (Signa, GE Healthcare) (TR range/te range 3,500 4,000/12 35; matrix size, 256 192 224; field of view, 16 16 or 18 18 cm; slice thickness, 3 mm) or a 3-T system (Signa, GE Healthcare) (2,500 3,000/18 50; matrix size, 384 256; field of view, 16 16 cm; slice thickness, 3 mm) and an extremity coil. Sagittal proton density weighted A C Fig. 1 Calculation of patellar height ratios in 38-year-old woman with medial knee joint pain. A, MR image shows dimensions used for calculation of Insall-Salvati index: length of patellar tendon (PT) divided by patellar length (PL). B, MR image shows dimensions used for calculation of modified Insall-Salvati index: distance from distal end of patellar cartilage to tibial tuberosity (TT) divided by length of patellar articular cartilage (PC). C, MR image shows dimensions used for calculation of patellotrochlear index: length of trochlear articular cartilage overlapping patellar cartilage (TC) divided by length of PC. D, MR image shows dimensions used for calculation of patellophyseal index: height of patella above anterior physeal line (PHY) divided by length of PC. fast spin-echo and fat-suppressed sequences were used to determine the measurements on sagittal images. Measurements were made with electronic calipers (efilm Workstation 2.1 software, Merge Healthcare) on the sagittal slice with the greatest patellar length. This length was measured from the proximal patellar margin to the distal anterior tip (excluding osteophytes and enthesophytes). Imaging Analysis The following measurements were obtained: patellar length, length of the patellar tendon, length of patellar articular cartilage, distance from the distal end of the patellar cartilage to the tibial tuberosity, and length of trochlear articular cartilage overlapping the patellar cartilage. With these measurements, the Insall-Salvati index (Fig. 1A), B D 1362 AJR:193, November 2009

Patella Alta TABLE 1: Clinical Knee Symptoms Knee pain History or Symptom Anterior No. of Patients (n = 100) 21 a Medial 31 Lateral 14 Posterior 7 Two or more locations 17 Entire knee 4 Other (instability, swelling, loose body) a Five patients had a history of recurrent knee dislocation. modified Insall-Salvati index (Fig. 1B), patellotrochlear index (Fig. 1C), and patellophyseal index (Fig. 1D) were calculated. A musculoskeletal radiologist with 6 years of experience retrospectively evaluated the patellofemoral articular cartilage surface for the presence of chondral defects. The sagittal and axial proton density weighted fast spin-echo and proton density weighted fat-suppressed images were assessed for this purpose. The reader was blinded to patient age and patellar measurements. The severity of the chondral defect was based on the depth of the lesion according to the International Cartilage Repair Society classification [16] (Table 2). In cases in which more than one cartilage defect was observed, the defect with the greatest severity was chosen. Statistical Analysis All measurements were plotted on a histogram for evaluation of the distribution. The criteria for patella alta and baja were determined with the 2 SD empirical rule. Each index was plotted and compared with patellotrochlear index. Pearson s correlation coefficient was calculated and tested. Measurements of each index in knees with normal cartilage were compared with those in knees with mild and severe cartilage defects, and p was derived with the Student s t test. Statistical testing was performed with MedCalc software (version 9.6, MedCalc). Statistical significance was defined as p < 0.05. Results The mean patellotrochlear index was 0.49 ± 0.15 (SD) (range, 0 0.88). The measurements plotted on a histogram showed a symmetric gaussian curve (Fig. 2). The population lying within 2 SDs was described as having a normal patella, whereas the lower extreme and upper extreme populations were described as having patella alta and patella baja, respectively. With 6 TABLE 2: Grading System for Classification of Cartilage Defects Based on International Cartilage Repair Society Classification Severity Grade MRI Classification None 0 Normal Mild 1 Normal contour with abnormally increased signal intensity 2 Superficial erosion or ulceration of less than 50% of cartilage thickness Severe 3 Partial-thickness defect of more than 50% but less than 100% of cartilage thickness Note Adapted from [16]. Fig. 2 Histogram of distribution of patellotrochlear index in study population shows symmetric curve. Cutoff index for patella alta is 0.18 and for patella baja is 0.80. Solid line indicates mean (0.49); dotted line, proposed normal range, defined by calculation of 2 SDs at each end of curve. 4 Full-thickness cartilage loss No. of Subjects this definition, patella alta was determined to be a patellotrochlear index less than 0.18 and patella baja an index greater than 0.80. There was no statistically significant difference in patellotrochlear index measurements between the sexes (p > 0.05). The normal, patella alta, and patella baja values calculated for Insall-Salvati index, modified Insall-Salvati index, and patellophyseal index are shown in Table 3. Having described the criteria for each index, we sought to determine the correlation between indexes. The correlation coefficient curves were plotted for each method. A weak correlation of r = 0.224 was found between the Insall-Salvati and patellotrochlear indexes, and a weaker correlation of r = 0.073 was found between the modified Insall-Salvati and patellotrochlear indexes. A strong correlation (r = 0.813) was found between patellophyseal index and patellotrochlear index. Results of the comparisons of indexes are shown in Table 4. 30 25 20 15 10 5 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Patella Alta Patellotrochlear Index Patella Baja Patellofemoral chondral defects were detected in 54 knees: 25 knees with mild defects (grades 1 and 2) and 29 knees with severe defects (grades 3 and 4). Forty-six knees had no chondral defects in the patellofemoral region. The comparison of indexes revealed no statistical difference between knees with normal cartilage and those with mild cartilage defects. However, between knees with normal cartilage and those with severe cartilage defects, there was a statistically significant difference in modified Insall-Salvati index (p = 0.024) and patellophyseal index (p = 0.024). No such difference was found for the Insall- Salvati index (p = 0.12) or patellotrochlear index (p = 0.29) (Table 5). Discussion The findings of our study suggest that the patellotrochlear index can be easily measured to determine the presence of patella alta as an overlap ratio less than 0.18 and TABLE 3: Ratios for Patella Alta, Patella Baja, and Normal Patella Determined With Previously Described Indexes Index Normal Patella Patella Alta Patella Baja Patellotrochlear 0.49 < 0.18 > 0.80 Insall-Salvati 1.07 > 1.43 < 0.71 Modified Insall-Salvati 1.81 > 2.40 < 1.24 Patellophyseal 0.37 > 0.65 < 0.09 AJR:193, November 2009 1363

Ali et al. TABLE 4: Calculated Correlation Coefficients (r) for Patellar Height Ratios Index Patellotrochlear Insall-Salvati Modified Insall-Salvati Patellotrochlear 0.224 a 0.073 Patellophyseal 0.813 a 0.437 a 0.144 Note Dash [ ] indicates not applicable. a Significant correlation (p < 0.05). TABLE 5: Comparison of Patellar Measurements and Severity of Cartilage Defects Index Severity None (n = 46) Mild (n = 25) p Severe (n = 29) p Insall-Salvati 1.05 (1.01, 1.10) 1.04 (0.98, 1.11) 0.79 1.13 (1.06, 1.20) 0.12 Modified Insall-Salvati 1.87 (1.80, 1.95) 1.83 (1.72, 1.93) 0.46 1.71 (1.61, 1.82) 0.024 a Patellotrochlear 0.49 (0.45, 0.53) 0.53 (0.47, 0.59) 0.33 0.45 (0.39, 0.51) 0.29 Patellophyseal 0.36 (0.33, 0.40) 0.32 (0.27, 0.36) 0.11 0.44 (0.39, 0.49) 0.024 a Note Index values are mean with lower and upper limits of 2 SDs in parentheses; p, result of Student s t test of comparison with normal cartilage. a Significant difference (p < 0.05). patella baja as a ratio greater than 0.80. We found, however, that patellotrochlear index does not correlate with either the Insall-Salvati index or modified Insall Salvati index. Therefore, these measurement methods fail to reflect the degree of patellofemoral chondral overlap. The Insall-Salvati and modified Insall-Salvati indexes are the widely accepted ratios for classifying patella alta. In our study, we used MRI to reassess the classification of these patellar height ratios. We found our calculation of the Insall-Salvati index (1.07 ± 0.36) similar to the 1.1 ± 0.2 ratio previously determined by Miller et al. [12] and the 1.05 ± 0.45 determined by Shabshin et al. [13]. Similarly, our determined 1.43 Insall-Salvati index cutoff for patella alta was greater than the index (> 1.2) originally determined on radiographs. In the case of the modified Insall-Salvati index, however, it was difficult to compare our values because there are no published MRI criteria, to our knowledge. The modified Insall-Salvati index was originally determined on radiographs, and the authors relied on estimating the cartilage surface using osseous anatomy. There is evidence [17, 18] to suggest that a significant difference exists in articular cartilage joint geometry and the corresponding subchondral osseous anatomy of the patella. Therefore, we believe that the previous authors may have overestimated the articulating surface. This theory would explain why our calculated cutoff value for patella alta on A B Fig. 3 Sagittal 3-T MR images. A, 20-year-old woman with patella alta according to patellotrochlear index (0.12) and patellophyseal index (0.68). B, 47-year-old woman with patella baja according to patellotrochlear index (0.88) and patellophyseal index (0.09). MR images (> 2.4) was greater than the originally described ratio (> 2) on radiographs. Patellofemoral cartilage contact area can be calculated by measuring the curvilinear line of contact between the patella and the femur on each axial image and multiplying by slice thickness [19]. Clinically this process is tedious and cumbersome, so instead we measured the amount of cartilage overlap (patellotrochlear index) between the patella and trochlea on sagittal images as an indirect way of assessing contact area. Biedert and Albrecht [15] described the patellotrochlear index for determining the presence of patella alta and baja as ratios of less than 0.12 and greater than 0.50. Our calculated patellotrochlear index for patella alta (< 0.18) was similar to the originally described ratio (Fig. 3A). However, we found a high frequency of normal knees falsely classified as having patella baja according to the previously described criterion (patellotrochlear index, 0.50). Because the index was originally determined with a 95% CI and not the upper and lower extremes of 2 SDs, there was a difference in cutoff value for patella baja. In most classifications, patellar height traditionally is determined on the basis of 2 SDs above and below the mean [6, 10]. Therefore we believe our calculated ratio of greater than 0.80 is a correct estimate of this condition (Fig. 3B). We also found no statistical difference in the patellotrochlear indexes of male and female subjects. This finding suggests that unlike values in previous studies [11, 20, 21], our criteria can be used for all patients independent of sex. Fig. 4 42-year-old woman with high-riding patella. Sagittal proton density weighted fast spin-echo MR image shows Insall-Salvati (0.78) and modified Insall- Salvati (1.82) indexes are within normal limits. Only patellotrochlear index (0) shows patellar position is abnormal. 1364 AJR:193, November 2009

Patella Alta We found a weak correlation between the patellotrochlear index and the Insall-Salvati and modified Insall-Salvati indexes in the assessment of patellar height. It can be theorized that the lack of correlation between the Insall-Salvati index and patellotrochlear index occurred because in calculation of the Insall- Salvati ratio, the patellar articulating cartilage surface is not assessed in determination of patellar height. This theory, however, does not explain the weak correlation between the modified Insall-Salvati index and patellotrochlear index because the former takes into account patellar cartilage measurements. Therefore, the lack of correlation can be explained only by the fact that the Insall-Salvati and modified Insall- Salvati indexes, unlike the patellotrochlear index, depend on determining the position of the patella with respect to the tibia. In determining the various ratios, we found that some knees with a grossly highriding patella had normal Insall-Salvati and modified Insall-Salvati indexes. Grelsamer et al. [10] recommended concomitant use of the modified Insall-Salvati and Insall-Salvati indexes for assessment of knees with unusual patellar morphologic features. We found this recommendation incorrect because some patients with high-riding patella and unusual patellar configuration had both Insall-Salvati and modified Insall-Salvati indexes within normal limits. In those cases, the only index that determined the knee had patella alta was the patellotrochlear index (Fig. 4). This finding supports previous suggestions to define patella alta according to patellar position with respect to the femur rather than using tibial bony landmarks as reference points; justifies our use of femoral reference points such as the femoral trochlear cartilage (patellotrochlear index) and the femoral physeal line (patellophyseal index) to compare findings with previously established ratios; and shows the lack of correlation between the Insall-Salvati index, modified Insall-Salvati index, and patellotrochlear index. The articular cartilage is not visible on radiographs, making it difficult to assess the patellotrochlear index. We therefore studied use of the femoral physeal line as a reference point for developing the patellophyseal index. Miller et al. [12] described measuring cartilage using the femoral physeal line and found weak correlation with the Insall-Salvati index. In our MRI examinations of the knee, we found that the physeal line lay in approximately the same position as the proximal end of the trochlear cartilage and near the superior aspect of trochlear groove. Unlike Miller et al., we determined the ratio by assessing the height of the patella lying above the femoral physeal line. Our calculation of the patellophyseal index had weak correlation with the Insall-Salvati index, yet there was strong correlation with the patellotrochlear index, suggesting that the patellophyseal index is a good measure of cartilage congruence. Unlike the trochlear cartilage, the femoral physeal line is easy to visualize on radiographs and MR images, making the patellophyseal index an easy ratio to determine (Fig. 1D). There is evidence to suggest that persons with patella alta have altered knee extensor mechanics that may predispose them to increased patellofemoral joint reaction forces [22, 23]. In knees with a high-riding patella, an 8-mm change in patellar height can be associated with as much as a 25% change in the magnitude of the resultant contact force [24]. Persons with patella alta also are believed to have an average of 19% less contact area than persons with normal knees over the range of flexion [17]. The combined effect of increased joint forces and reduced contact area results in increased joint stress (force per unit area), which has been implicated [25] as the cause of patellofemoral pain and chondral injuries in these patients. Therefore, any classification for determining patellar height should take into consideration the associated pathologic changes in the joint. The association between patella alta and chondromalacia patellae has been studied extensively. Authors finding an association have reasoned that a high-riding patella results in malalignment and incongruity between the articulating surfaces of the patella and the trochlea. Endo et al. [26], however, did not find a significant correlation between patella alta and chondromalacia patellae. They determined patellar height using the Insall-Salvati index and patellotrochlear chondral overlap. Our results confirmed these findings because we also did not find a statistically significant difference between the presence of patella alta measured with the Insall-Salvati and patellotrochlear indexes and the presence of chondral defects. We did, however, find a statistically significant difference between the modified Insall-Salvati and patellophyseal indexes of knees with normal cartilage and those indexes in knees with severe cartilage defects. We believe this finding is important because it justifies the use of these two indexes in assessing premorbid anatomic conditions. There were limitations to this study. First, because the study was retrospective, only knees of patients with symptoms were analyzed. We included patients with various pathologic conditions, such as osteoarthritis, calcium pyrophosphate deposition syndrome, and chondral defects, that made assessment of the cartilage, tendon, and bone difficult. This decision might have led to inaccuracy in the measurements used for calculation of the indexes in these patients. Second, we did not directly compare indexes in patients with patella alta with indexes in patients with patella baja because the number of patients in each group was too small. Because there is no reference standard for assessing for patella alta, we did not directly determine the sensitivity and specificity of using the patellotrochlear index. We therefore had to correlate our criteria with commonly used methods described in the literature. Last, this study was performed with static MR images, so we ignored the dynamic components. For the purpose of the study, we considered the degree of knee flexion to be approximately the same in all subjects during imaging. Because differences in degree of flexion can alter the amount of cartilage contact area [18], results of future studies performed to assess a criterion for patellar height should account for this variable. Our study showed limited correlation between commonly used patellar height ratios and articular cartilage congruence. Because cartilage congruence is an important factor in determining patellar malalignment, we question the clinical utility of indexes defined by tibial reference points in determining patellar height. Our study showed, however, that unlike the Insall-Salvati and patellotrochlear indexes, the modified Insall-Salvati index and patellophyseal indexes are associated with severe chondral defects. This association with patellofemoral abnormality may be evidence for clinicians to apply these two indexes in predicting the presence of premorbid anatomic conditions. Furthermore, the patellophyseal index correlates well with articular cartilage congruence and can be easily measured on both MR images and radiographs, making it an ideal technique for assessing patellar height. References 1. Kannus PA. Long patellar tendon: radiographic sign of patellofemoral pain syndrome. Radiology 1992; 185:859 863 2. Kujala UM, Osterman K, Kvist M, Aalto T, Fri berg O. AJR:193, November 2009 1365

Ali et al. Factors predisposing to patellar chondropathy and Orthop Relat Res 1992; 282:170 176 joint contact area using magnetic resonance imag- patellar apicitis in athletes. Int Orthop 1986; 12. Miller TT, Staron RB, Feldman F. Patellar height ing. Magn Reson Imaging 2003; 21:955 959 10:195 200 on sagittal MR imaging of the knee. AJR 1996; 20. Aglietti P, InsaIl IN, Cerulli G. Patellar pain and 3. Insall J, Goldberg V, Salvati E. Recurrent disloca- 167:339 341 incongruence. Part 1. Measurements of incongru- tion and the high-riding patella. Clin Orthop 13. Shabshin N, Schweitzer ME, Morrison WB, Park- ence. Clin Orthop Relat Res 1983; 176:217 224 1972; 88:67 69 4. Insall J, Falvo KA, Wise DW. Chondromalacia patellae: a prospective study. J Bone Joint Surg Am 1976; 58:1 8 5. Moller BN, Krebs B, Jurik AG. Patellar height and patellofemoral congruence. Arch Orthop Trauma Surg 1986; 104:380 381 6. Insall J, Salvati E. Patella position in the normal knee joint. Radiology 1971; 101:101 104 7. Blackburne JS, Peel TE. A new method of measuring patellar height. J Bone Joint Surg Br 1977; 59:241 242 8. Caton J, Deschamps G, Chambat P, et al. Patella infera: apropos of 128 cases [in French]. Rev Chir Orthop Reparatrice Appar Mot 1982; 68:317 325 9. de Carvalho A, Holst Andersen A, Topp S, Jurik AG. A method for assessing the height of the patella. Int Orthop 1985; 9:195 197 10. Grelsamer RP, Proctor CS, Bazos AN. Evaluation of patellar shape in the sagittal plane: a clinical analysis. Am J Sports Med 1994; 22:61 66 11. Grelsamer RP, Meadows S. The modified Insall- Salvati ratio for assessment of patellar height. Clin er L. MRI criteria for patella alta and baja. Skeletal Radiol 2004; 33:445 450 14. Seil R, Muller B, Georg T, Kohn D, Rupp S. Reliability and interobserver variability in radiological patellar height ratios. Knee Surg Sports Traumatol Arthrosc 2000; 8:231 236 15. Biedert RM, Albrecht A. The patellotrochlear index: a new index for assessing patellar height. Knee Surg Sports Traumatol Arthrosc 2006; 14: 707 712 16. Brittberg M, Winalski CS. Evaluation of cartilage injuries and repair. J Bone Joint Surg Am 2003; 85:58 69 17. Ward SR, Terk MR, Powers CM. Patella alta: association with patellofemoral alignment and changes in contact area during weight-bearing. J Bone Joint Surg Am 2007; 89:1749 1755 18. Staeubli HU, Bosshard C, Porcellini P, Rauschning W. Magnetic resonance imaging for articular cartilage: cartilage bone mismatch. Clin Sports Med 2002; 21:417 433 19. Heino-Brechter J, Powers CM, Terk MR, Ward SR, Lee TQ. Quantification of patellofemoral 21. Dowd GS, Bentley G. Radiographic assessment in patellar instability and chondromalacia patellae. J Bone Joint Surg Br 1986; 68:297 300 22. Heino-Brechter J, Powers CM. Patellofemoral stress during walking in persons with and without patellofemoral pain. Med Sci Sports Exerc 2002; 34:1582 1593 23. Yamaguchi GT, Zajac FE. A planar model of the knee joint to characterize the knee extensor mechanism. J Biomech 1989; 22:1 10 24. Singerman R, Davy DT, Goldberg VM. Effects of patella alta and infera on patellofemoral contact forces. J Biomech 1994; 27:1059 1065 25. Fulkerson JP, Shea KP. Mechanical basis for patellofemoral pain and cartilage breakdown. In: Ewing JW, ed. Articular cartilage and knee joint function: basic science and arthroscopy. New York: Raven Press, 1990:93 101 26. Endo Y, Schweitzer ME, Bordalo-Rodrigues M, Rokito AS, Babb JS. MRI quantitative morphologic analysis of patellofemoral region: lack of correlation with chondromalacia patellae at surgery. AJR 2007; 189:1165 1168 1366 AJR:193, November 2009