Abstract

Objective. The aim of our study was to evaluate the association between patellar alignment (using standard MRI images of extended knees) and MRI indices of patellofemoral (PF) osteoarthritis (OA) features.

Methods. In this cross-sectional observational study, subjects were recruited to participate in the Boston Osteoarthritis of the Knee Study (BOKS). The association of patellar alignment [patellar length ratio (PLR), sulcus angle (SA), lateral patellar tilt angle (LPTA) and bisect offset (BO)] with measures of PF OA [cartilage morphology and bone marrow lesion (BML) in the medial and lateral PF compartment] were examined using a logistic regression model while adjusting for age, sex and BMI.

Results. Study sample comprised 126 males (mean age 68.0, BMI 31.2) and 87 females (mean age 64.7, BMI 31.6). All measurements of patellar alignment were statistically significantly associated with cartilage morphology and BML in the lateral compartment of PF joint. PLR and SA were significantly associated with medial cartilage loss. With increasing PLR there was an increased prevalence of lateral and medial cartilage loss as well as of lateral BML. Increasing SA was positively associated with increased lateral and medial cartilage loss and lateral BML. LPTA range was negatively associated with lateral cartilage loss and BML. More laterally displaced patella (higher BO) was associated with increased lateral cartilage loss and BML.

Conclusions. The results of our study clearly indicated that patellar alignment is associated with manifestations of PF OA such as cartilage thickness loss and BML.

Introduction

Osteoarthritis (OA) is a major public health problem due to its high prevalence, costs and levels of pain and disability. It is the most widespread chronic joint disorder and the knee is frequently affected [1]. Increases in life expectancy and ageing of the population are expected to make OA the fourth leading cause of disability in the general population by the year 2020 [2].

The knee joint comprises both the tibiofemoral and patellofemoral (PF) joints. The structural changes that can occur in knee OA include alterations in a number of tissue structures such as cartilage loss, bone marrow lesions (BMLs), synovitis, osteophyte formation, subchondral sclerosis and subchondral cysts [3]. Whilst the structural changes in the tibiofemoral joint with OA and their associated risk factors are well described, the literature on alterations in the PF joint is sparser.

One of the potent risk factors for structural alteration in the PF joint is patellar malalignment. Patellae that are located centrally in the trochlear groove, and not malaligned are thought to be less likely to develop OA [4–6]. Patellar malalignment could be a reason for excessive stress on the articular surfaces of the PF joint, and therefore could potentially be a reason for aberrant mechanical forces and osteoarthritic changes in the knee.

Most of the previous studies of patellar malalignment have used plain X-ray evaluations of the knee from lateral and skyline radiographs [7–11]. Various methods to evaluate patellar malalignment were proposed: (i) in the lateral plane—evaluation of relationship between patellar height and patellar ligament length [12, 13]; (ii) on the axial (skyline) view—evaluation of trochlear sulcus angle (SA) and depth [14], evaluation of lateral patello-femoral angle [11, 15], lateral patellar tilt angle [16], bisect offset of patella [17] and evaluation of congruence angle [16].

The last decade has seen the growth of the use of MRI for evaluation of knee OA. This non-invasive method helps to directly visualize the patellar and femoral cartilages as well as BMLs. In recent research, Felson et al. [18] demonstrated that BMLs are strongly associated with pain and structural progression of knee OA. In the tibiofemoral joint, BMLs were strongly related to mechanical alignment.

Very few studies have evaluated PF alignment using MRI [19–21]. Muellner et al. [19] performed measurements analogous to those that were accepted in X-ray evaluation. In this study, the MRI images were obtained with knees flexed to between 20° and 45°. Knee flexion allows evaluation of PF relations when the patella is located in opposition to the femoral trochanteric sulcus. None of these studies evaluated the relation of patellar alignment to other structural changes that could be evaluated on knee MRI.

In common clinical practice, MRI of the knees is usually obtained with the patient in the supine position with fully extended knees. Thus, in this study we assessed the association of patellar alignment (in fully extended knees) with structural manifestations of PF OA. Our hypothesis was that increasing patellar malalignment would be associated with increasing structural changes (cartilage thickness loss and BML) in persons with symptomatic knee OA.

Materials and methods

Study design

This was a cross-sectional observational study.

Sample

Subjects were recruited to participate in a natural history study of symptomatic knee OA, called the Boston Osteoarthritis of the Knee Study (BOKS). Subjects in this study were a subset of subjects whose recruitment has been described in detail elsewhere [22]. Briefly, subjects were recruited from two prospective studies of the quality of life of veterans (one of men and one of women), from clinics at the Veterans Administration Boston Health Care System and from advertisements in local newspapers. Potential participants were asked two questions: ‘Do you have pain, aching or stiffness in one or both knees on most days?’ and ‘Has a doctor ever told you that you have knee arthritis?’ For subjects who answered positively to both questions, we conducted a follow-up interview in which we asked about other types of arthritis that could cause knee symptoms. If no other forms of arthritis were identified, then the individual was eligible for recruitment. To determine whether subjects had radiographic OA, they underwent a series of standard knee radiographs (antero-posterior, lateral and skyline views). For the skyline view each knee was imaged separately with the participant in a standing position and the limb flexed at 30°–40° during weight bearing. If the subject had a definite osteophyte on any view in the symptomatic knee, they were eligible for the study. By having frequent knee symptoms and radiographic OA, all subjects met ACR criteria for symptomatic knee OA [23]. For the natural history study, we enrolled subjects who were interested in participating and who could walk with or without a cane. The examinations were approved by the Boston University Medical Center and the Veterans Administration Boston Healthcare System IRBs. Each subject's written consent was obtained according to the Declaration of Helsinki.

MRI evaluations

All studies were performed with a Signa 1.5T MRI system (General Electric Corp., Milwaukee, WI, USA) using a phased-array knee coil. A standard anchoring device for the ankle and knee was used to ensure uniformity of positioning between patients and for follow-up. The imaging protocol included sagittal spin-echo proton density and T2-weighted images [repetition time (TR), 2200 ms; time to echo (TE) 20/80 ms] with a slice thickness of 3 mm, a 1 mm interslice gap, 1 excitation, a field of view (FOV) of 11–12 cm, and a matrix of 256 × 192 pixels; and coronal and axial spin-echo fat-suppressed proton density- and T2-weighted images (TR 2200 ms; TE 20/80) with a slice thickness of 3 mm, a 1 mm interslice gap, 1 excitation and with the same FOV and matrix. The 213 MRIs from BOKS used in this analysis were those that were available in a digital archive thus allowing us to measure PF alignment as subsequently described. Each of the 213 subjects had one knee, on which the MRI was performed and each knee had cartilage morphology and BML scored on four plates in the PF joint, i.e. the medial patella, the medial anterior femur, the lateral patella and the lateral anterior femur.

Cartilage morphology

PF cartilage morphology on MRI was scored paired and unblinded to sequence on two plates (anterior femur and patella), for the medial and lateral PF compartment, respectively, using the WORMS semi-quantitative method [24]. Cartilage morphology was scored using a 0–6 scale: 0 = normal thickness and signal; 1 = normal thickness but increased signal on T2-weighted images; 2 = solitary focal defect of <1 cm in greatest width; 3 = areas of partial-thickness defects (<75% of the plate) with areas of preserved thickness; 4 = diffuse partial-thickness loss of cartilage (≥75% of the plate); 5 = areas of full-thickness loss (<75% of the plate) with areas of partial thickness loss; 6 = diffuse full-thickness loss (≥75% of the plate). Intra-class correlation coefficient (ICC) on agreement for cartilage readings ranged from 0.75 to 0.97.

Scores of grade 1 and 2 were extremely infrequent among the MRIs read in our study population [25]. Further, grade 1 represents a change in signal in cartilage of otherwise normal morphology, while grades 2 and 3 represent similar types of abnormality of the cartilage, focal defects without overall thinning. Therefore, to create a consistent and logical scale for evaluating cartilage morphology and to ensure that the assumptions of proportional odds modelling were fulfilled we collapsed the original WORMS cartilage scores that were read on a 0–6 scale, into three groups for the purposes of our analyses, where the original WORMS score of 0 and 1 were considered as the first group, the original scores of 2 and 3 were considered as the second group and the original scores of 4, 5 and 6 were considered as the third group, respectively, in the new scale [25]. Using this new scale, cartilage morphology at each region was then defined on a scale from 0 to 2.

Bone marrow lesions

BMLs in the subarticular marrow are defined as poorly marginated areas of increased signal intensity in the normally hypointense fatty marrow on the fat-suppressed spin-echo images and graded in each region from 0 to 3 based on the extent of regional involvement; 0 = none; 1 < 25% of the region; 2 = 25–50% of the region; 3 ≥50% of the region. Inter-observer agreement (ICC) for reading BMLs was 0.82. In this scoring system, BMLs in the PF joint were graded in the anterior femur and posterior (articular) surface of patella in the medial and lateral PF compartment respectively, blinded to sequence [24].

Patellar alignment evaluation

The patellar alignment evaluation in this study was performed using the digital film viewing software eFilm Workstation (Version 2.0.0) software. We measured patellar alignment in two planes: sagittal and transverse (axial). In the sagittal plane, we measured patellar length ratio (PLR) according to the Insall and Salvati method [12]. For these measurements we found the slice with clearly recognizable patellar margins and where the patellar bone volume appeared to be maximal. To measure patellar length and patellar ligament length according to the Insall and Salvati method two lines were drawn (Fig. 1A): PL—from upper to lower point of inner (articulating) surface of the patella excluding osteophytes; TL (patellar ligament length)—from lower inner point of patella to the highest point of tibial tuberosity. PLR was calculated according to the equation: PLR = PL/TL.

Fig. 1.

Schema of measured patellar alignment indices. In a sagittal plane: (A) PL, inner patellar length and TL, patellar tendon length (PLR, patellar length ratio was computed as PL / TL); in a transverse (axial) plane; (B) SA, sulcus angle; (C) LPTA, lateral patellar tilt angle; and (D) schema of bisect offset (BO) measurement (BO = a*100/a + b).

Fig. 1.

Schema of measured patellar alignment indices. In a sagittal plane: (A) PL, inner patellar length and TL, patellar tendon length (PLR, patellar length ratio was computed as PL / TL); in a transverse (axial) plane; (B) SA, sulcus angle; (C) LPTA, lateral patellar tilt angle; and (D) schema of bisect offset (BO) measurement (BO = a*100/a + b).

In the transverse plane, we measured two groups of indices: (i) one measure that describes the trochlear depth: SA and (ii) two measures that described patellar position: lateral patellar tilt angle (LPTA) and bisect offset (BO). For the measurements of SA we found the axial slice that referred to the proximal 1/3 of the femoral trochlear curve by using the 3D cursor on the sagittal image. SA is an angle between two lines: from the lowest point of the trochlear sulcus, one on lateral bony margin and the second on the medial bony margin (Fig. 1B). For the measurements of patellar alignment we found the axial slice that referred to the middle of the patella by using the 3D cursor on the sagittal image. LPTA is an angle between the posterior condylar line and the line drawn through the lateral inferior bony margin of the patella (Fig. 1C). For BO measurements we drew the posterior condylar line and the perpendicular line up though the lowest point of the femoral sulcus and through the patella. We measured the distance between the lateral border of the patella and this vertical line (a) and between the medial border of the patella and this vertical line (b) (Fig. 1C). BO was calculated according to the formula: BO = a*100/(a + b).

Reliability of MRI readings

First, we (L.K. and D.H.) read one batch of MRIs and developed a reading protocol for evaluation of patellar alignment. Using this protocol each of us read and re-read 10 MRIs separately to assess the intra- and inter-rater reliability of the readings for each of the patellar alignment features. One investigator (L.K.), blinded to patient identifiers, read all of the MRIs. To evaluate for reader-drift, we re-assessed intra-rater reliability by inserting one original reliability scan for every 10 new scans. Before reading each batch of MRIs investigator (L.K.) re-read five MRIs, which had been previously read, to ‘calibrate’ his readings to a standard. The intra-observer reliability for reading for different patellar alignment indices varied between 0.86 and 0.96. The inter-observer reliability ranged from 0.68 to 0.93.

Statistical analysis

We first categorized each of the four patellar alignment measures into quartiles. PF cartilage morphology took on whole number values from 0 (no loss) to 2 (maximum loss) and was analysed as ordered categories. We examined the relation between quartiles of each patellar alignment measurement and medial PF (medial patella and medial anterior femur) cartilage morphology using the proportional odds logistic regression model. Generalized estimating equation models were used to account for the correlation in the cartilage morphology outcome between the anterior femur and patellar plates within a knee. We tested for linear trend for each measurement of patellar alignment and medial PF cartilage using the patellar alignment measure as a continuous variable in the model. If there was a potential U shape or J-shaped relation between the patellar alignment measure and medial PF cartilage, we tested the U-shaped trend by including both the patellar alignment measure and its square in the regression. We used the same approach to examine the relation between each patellar alignment evaluation and lateral PF cartilage. All the models were adjusted for age, sex and BMI.

Since the number of knees with BMLs >1 in PF joint was small, we collapsed BML grades 1–3 into one category thus generating a dichotomous variable. We examined the relation between quartiles of each patellar alignment measure and medial BML using a logistic regression model while adjusting for age, sex, and BMI. Generalized estimating equation was used to adjust for the correlation between plates within a knee. Linear trend and U-shaped trend were tested using the methods described earlier. The same approach was used to examine the relation between each patellar alignment measure and lateral BML. Statistical analyses were performed using SAS software, (SAS Institute Inc, Cary, NC, USA release 9.1).

Results

Of the 324 patients entering BOKS, 311 obtained an MRI of their more symptomatic knee at baseline. Table 1 demonstrates the characteristics of the 213 study participants with available digitally archived films used for this analysis. We compared the group of individuals who were included in the present study (n = 213) with the group of individuals who were not (n = 111). There were no statistically significant differences between these groups in terms of age {mean (S.D.) [66.7 (9.3) years vs 67.8 (9.1) years], respectively, P = 0.28{ and BMI [31.4 (5.5) kg/m2vs 31.5 (6.1) kg/m2, respectively, P = 0.87]. Our study sample was composed of 126 males (mean age 68.0) and 87 females (mean age 64.7). On average, the subjects were obese with a mean BMI of 31.2 for males and 31.6 for females and had radiographic knee OA (K&L grade ≥2 in 65.9% of males and 87.4% of females). Those who did not fulfill the criteria for radiographic OA based on their K&L grade had PF OA.

Table 1.

Characteristics of the studied sample

Characteristics n Mean (s.d.) Range 
Age 213 66.7 (9.3) 47–93 
Sex (women) 213 40.8%  
BMI 213 31.4 (5.5) 21.5–55.9 
K & L ≥ 2 212 75.0% 0–4 
Characteristics n Mean (s.d.) Range 
Age 213 66.7 (9.3) 47–93 
Sex (women) 213 40.8%  
BMI 213 31.4 (5.5) 21.5–55.9 
K & L ≥ 2 212 75.0% 0–4 

(Tables 2–5) show the relation between the patellar alignment measures (predictors) and indices of PF OA (outcomes). Each table presents the number of knees in each quartile, the range of predictor in each quartile, odds ratios and the P for trend if the model. As mentioned before, each of the 213 subjects had one knee on which the MRI was performed and each knee had cartilage morphology and BML scored on four plates in the PF joint, that is, the medial patella, the medial anterior femur, the lateral patella and the lateral anterior femur. The number of plates with missing value was seven for medial cartilage loss, seven for lateral cartilage loss, four for medial BML and three in lateral BML. Two hundred and five knees had PLR and SA measured and 204 knees had LPTA measured and 202 knees had BO measured. So the number of plates used for the analyses were different for each combination of MRI feature and patella alignment.

Table 2.

Association between patella alignment (four groups) and adjusted means of cartilage loss in a lateral compartment of PF joint

 Lateral cartilage loss 
 Quartiles Lowest 2nd 3rd Highest P for trend 
PLR No. of plates 99 102 100 102 Linear: 0.0068 
 Range of predictor 0.66–0.87 0.88–0.98 0.98–1.12 1.13–1.71  
 Cartilage morphology 2–3 (%) 19.2 18.6 18.0 13.7  
 Cartilage morphology 4–6 (%) 29.3 33.3 48.0 55.9  
 Odds ratio (95% CI) 1.0 1.1 (0.7–1.9) 1.7 (1.0–2.9) 2.0 (1.2–3.6)  
SA No. of plates 101 103 99 101 Linear: 0.0009 
 Range of predictor 98–113 114–119 120–124 125–155  
 Cartilage morphology 2–3 (%) 15.8 20.4 19.2 16.8  
 Cartilage morphology 4–6 (%) 28.7 44.7 40.4 53.5  
 Odds ratio (95% CI) 1.0 2.1 (1.2–3.5) 1.8 (1.0–3.0) 2.8 (1.6–4.8)  
LPTA No. of plates 102 102 91 106 Linear: <0.0001 
 Range of predictor 25–13 14–17 18–21 22–35  
 Cartilage morphology 2–3 (%) 13.7 16.7 26.4 16.0  
 Cartilage morphology 4–6 (%) 56.9 43.1 45.1 22.6  
 Odds ratio (95% CI) 1.0 0.7 (0.4–1.2) 0.8 (0.5–1.5) 0.3 (0.2–0.5)  
BO No. of plates 98 97 104 98 Linear: <0.0001 
 Range of predictor 38.46–54.55 54.76–60.42 60.47–66.66 66.67–100  
 Cartilage morphology 2–3 (%) 16.3 19.6 20.2 16.3  
 Cartilage morphology 4–6 (%) 31.6 30.9 39.4 65.3  
 Odds ratio (95% CI) 1.0 0.8 (0.5–1.4) 1.1 (0.7–2.0) 3.4 (1.9–6.1)  
 Lateral cartilage loss 
 Quartiles Lowest 2nd 3rd Highest P for trend 
PLR No. of plates 99 102 100 102 Linear: 0.0068 
 Range of predictor 0.66–0.87 0.88–0.98 0.98–1.12 1.13–1.71  
 Cartilage morphology 2–3 (%) 19.2 18.6 18.0 13.7  
 Cartilage morphology 4–6 (%) 29.3 33.3 48.0 55.9  
 Odds ratio (95% CI) 1.0 1.1 (0.7–1.9) 1.7 (1.0–2.9) 2.0 (1.2–3.6)  
SA No. of plates 101 103 99 101 Linear: 0.0009 
 Range of predictor 98–113 114–119 120–124 125–155  
 Cartilage morphology 2–3 (%) 15.8 20.4 19.2 16.8  
 Cartilage morphology 4–6 (%) 28.7 44.7 40.4 53.5  
 Odds ratio (95% CI) 1.0 2.1 (1.2–3.5) 1.8 (1.0–3.0) 2.8 (1.6–4.8)  
LPTA No. of plates 102 102 91 106 Linear: <0.0001 
 Range of predictor 25–13 14–17 18–21 22–35  
 Cartilage morphology 2–3 (%) 13.7 16.7 26.4 16.0  
 Cartilage morphology 4–6 (%) 56.9 43.1 45.1 22.6  
 Odds ratio (95% CI) 1.0 0.7 (0.4–1.2) 0.8 (0.5–1.5) 0.3 (0.2–0.5)  
BO No. of plates 98 97 104 98 Linear: <0.0001 
 Range of predictor 38.46–54.55 54.76–60.42 60.47–66.66 66.67–100  
 Cartilage morphology 2–3 (%) 16.3 19.6 20.2 16.3  
 Cartilage morphology 4–6 (%) 31.6 30.9 39.4 65.3  
 Odds ratio (95% CI) 1.0 0.8 (0.5–1.4) 1.1 (0.7–2.0) 3.4 (1.9–6.1)  

Adjusted for age, sex and BMI.

Table 3.

Association between patellar alignment (four groups) and adjusted means of cartilage loss in a medial compartment of PF joint

 Medial cartilage loss 
 Quartiles Lowest 2nd 3rd Highest P for trend 
PLR No. of plates 100 101 100 102 Linear: 0.0109 
 Cartilage morphology 2–3 (%) 29.0 23.8 38.0 32.4  
 Cartilage morphology 4–6 (%) 27.0 37.6 40.0 44.1  
 Odds ratio (95% CI) 1.0 1.4 (0.8–2.4) 2.0 (1.2–3. 3) 2.0 (1.2–3.4)  
SA No. of plates 101 103 99 101 Linear: 0.0016 
 Cartilage morphology 2–3 (%) 31.7 31.1 27.3 34.7  
 Cartilage morphology 4–6 (%) 26.7 38.8 38.4 45.5  
 Odds ratio (95% CI) 1.0 1.7 (1.0–2.8) 1.6 (0.9–2.6) 2.4 (1.4–4.1)  
LPTA No. of plates 102 102 91 106 Linear: 0.1553 
 Cartilage morphology 2–3 (%) 37.3 29.4 30.8 25.5 U-shape: 0.3376 
 Cartilage morphology 4–6 (%) 35.3 43.1 40.7 30.2  
 Odds ratio (95% CI) 1.0 1.3 (0.8–2.2) 1.2 (0.7–2.0) 0.7 (0.4–1.2)  
BO No. of plates 98 97 104 98 Linear: 0.0756 
 Cartilage morphology 2–3 (%) 32.7 25.8 27.9 36.7  
 Cartilage morphology 4–6 (%) 29.6 35.1 42.3 41.8  
 Odds ratio (95% CI) 1.0 0.9 (0.5–1.6) 1.2 (0.7–2.1) 1.5 (0.9–2.6)  
 Medial cartilage loss 
 Quartiles Lowest 2nd 3rd Highest P for trend 
PLR No. of plates 100 101 100 102 Linear: 0.0109 
 Cartilage morphology 2–3 (%) 29.0 23.8 38.0 32.4  
 Cartilage morphology 4–6 (%) 27.0 37.6 40.0 44.1  
 Odds ratio (95% CI) 1.0 1.4 (0.8–2.4) 2.0 (1.2–3. 3) 2.0 (1.2–3.4)  
SA No. of plates 101 103 99 101 Linear: 0.0016 
 Cartilage morphology 2–3 (%) 31.7 31.1 27.3 34.7  
 Cartilage morphology 4–6 (%) 26.7 38.8 38.4 45.5  
 Odds ratio (95% CI) 1.0 1.7 (1.0–2.8) 1.6 (0.9–2.6) 2.4 (1.4–4.1)  
LPTA No. of plates 102 102 91 106 Linear: 0.1553 
 Cartilage morphology 2–3 (%) 37.3 29.4 30.8 25.5 U-shape: 0.3376 
 Cartilage morphology 4–6 (%) 35.3 43.1 40.7 30.2  
 Odds ratio (95% CI) 1.0 1.3 (0.8–2.2) 1.2 (0.7–2.0) 0.7 (0.4–1.2)  
BO No. of plates 98 97 104 98 Linear: 0.0756 
 Cartilage morphology 2–3 (%) 32.7 25.8 27.9 36.7  
 Cartilage morphology 4–6 (%) 29.6 35.1 42.3 41.8  
 Odds ratio (95% CI) 1.0 0.9 (0.5–1.6) 1.2 (0.7–2.1) 1.5 (0.9–2.6)  

Adjusted for age, sex and BMI.

For range of predictors see Table 2.

Table 4.

Association between patellar alignment (four groups) and adjusted means of BML in a lateral compartment of PF joint

 Lateral BML 
 Quartiles P for trend 
PLR No. of plates 102 102 100 103 Linear: 0.003 
 Frequency of BML (%) 10.8 9.8 9.0 28.2 U-shape: 0.007 
 Odds ratio (95% CI) 1.0 0.9 (0.4–2.4) 0.6 (0.3–1.7) 2.5 (1.1–5.4)  
SA No. of plates 102 104 99 102 Linear: 0.007 
 Frequency of BML (%) 5.9 22.1 10.1 19.6 U-shape: 0.92 
 Odds ratio (95% CI) 1.0 4.6 (1.8–2.1) 1.8 (0.6–5.2) 3.6 (1.4–9.6)  
LPTA No. of plates 104 103 90 108 Linear: <0.0001 
 Frequency of BML (%) 31.7 13.6 7.8 4.6  
 Odds ratio (95% CI) 1.0 0.4 (0.2–0.8) 0.2 (0.1–0.5) 0.1 (0.05–0.3)  
BO No. of plates 100 99 102 100 Linear: <0.0001 
 Frequency of BML (%) 8.00 8.1 15.7 27.0  
 Odds ratio (95% CI) 1.0 0.9 (0.3–2.5) 1.7 (0.7–4.3) 3.2 (1.3–7.7)  
 Lateral BML 
 Quartiles P for trend 
PLR No. of plates 102 102 100 103 Linear: 0.003 
 Frequency of BML (%) 10.8 9.8 9.0 28.2 U-shape: 0.007 
 Odds ratio (95% CI) 1.0 0.9 (0.4–2.4) 0.6 (0.3–1.7) 2.5 (1.1–5.4)  
SA No. of plates 102 104 99 102 Linear: 0.007 
 Frequency of BML (%) 5.9 22.1 10.1 19.6 U-shape: 0.92 
 Odds ratio (95% CI) 1.0 4.6 (1.8–2.1) 1.8 (0.6–5.2) 3.6 (1.4–9.6)  
LPTA No. of plates 104 103 90 108 Linear: <0.0001 
 Frequency of BML (%) 31.7 13.6 7.8 4.6  
 Odds ratio (95% CI) 1.0 0.4 (0.2–0.8) 0.2 (0.1–0.5) 0.1 (0.05–0.3)  
BO No. of plates 100 99 102 100 Linear: <0.0001 
 Frequency of BML (%) 8.00 8.1 15.7 27.0  
 Odds ratio (95% CI) 1.0 0.9 (0.3–2.5) 1.7 (0.7–4.3) 3.2 (1.3–7.7)  

Adjusted for age, sex and BMI.

For range of predictors see Table 2.

Table 5.

Association between patellar alignment (four groups) and adjusted means of BML in a medial compartment of PF joint

 Medial BML 
 Quartiles P for trend 
PLR No. of plates 102 101 100 103 Linear: 0.52 
 Frequency of BML (%) 12.8 13.9 13.0 16.5  
 Odds ratio (95% CI) 1.0 1.05 (0.5–2.4) 0.9 (0.4–2.1) 1.1 (0.5–2.5)  
SA No. of plates 101 104 99 102 Linear: 0.06 
 Frequency of BML (%) 9.9 15.4 12.1 18.6 U-shape: 0.91 
 Odds ratio (95% CI) 1.0 1.6 (0.7–3.8) 1.2 (0.5-3.1) 2.0 (0.9–4.7)  
LPTA No. of plates 104 103 90 107 Linear: 0.80 
 Frequency of BML (%) 15.4 15.5 10.0 14.0 U-shape: 0.72 
 Odds ratio (95% CI) 1.0 1.1 (0.5–2.5) 0.6 (0.3–1.5) 1.0 (0.5–2.3)  
BO No. of plates 99 99 102 100 Linear: 0.13 
 Frequency of BML (%) 14.1 18.2 11.8 12.0  
 Odds ratio (95% CI) 1.00 1.1 (0.5–2.4) 0.6 (0.3–1.4) 0.6 (0.3–1.4)  
 Medial BML 
 Quartiles P for trend 
PLR No. of plates 102 101 100 103 Linear: 0.52 
 Frequency of BML (%) 12.8 13.9 13.0 16.5  
 Odds ratio (95% CI) 1.0 1.05 (0.5–2.4) 0.9 (0.4–2.1) 1.1 (0.5–2.5)  
SA No. of plates 101 104 99 102 Linear: 0.06 
 Frequency of BML (%) 9.9 15.4 12.1 18.6 U-shape: 0.91 
 Odds ratio (95% CI) 1.0 1.6 (0.7–3.8) 1.2 (0.5-3.1) 2.0 (0.9–4.7)  
LPTA No. of plates 104 103 90 107 Linear: 0.80 
 Frequency of BML (%) 15.4 15.5 10.0 14.0 U-shape: 0.72 
 Odds ratio (95% CI) 1.0 1.1 (0.5–2.5) 0.6 (0.3–1.5) 1.0 (0.5–2.3)  
BO No. of plates 99 99 102 100 Linear: 0.13 
 Frequency of BML (%) 14.1 18.2 11.8 12.0  
 Odds ratio (95% CI) 1.00 1.1 (0.5–2.4) 0.6 (0.3–1.4) 0.6 (0.3–1.4)  

Adjusted for age, sex and BMI.

For range of predictors see Table 2.

All studied predictors were statistically significantly associated with cartilage loss and BML in the lateral compartment of the PF joint (Tables 2 and 4). PLR and SA were also significantly associated with cartilage loss in medial compartment (Table 3) and no significant association was found between all patellar alignment indices and medial BML (Table 5).

With increasing PLR there was an increased lateral and medial cartilage loss and lateral BML. Increasing SA, as well, was positively associated with increased lateral and medial cartilage loss and lateral BML. The lowest quartile of LPTA (25°–13°) was associated with the greatest lateral cartilage loss whereas range 22°–35° (the highest quartile) with the lowest one. The same direction of association was found between LPTA and lateral BML. LPTA was also associated with medial cartilage loss but not with medial BML. More laterally displaced patella (higher BO) was associated with increased lateral cartilage loss and lateral BML but not with changes in medial compartment of patellofemoral joint (PFJ).

Discussion

In this cross-sectional study, we have found significant associations between indices of patellar alignment using standard knee MRI and measures of PF OA, such as cartilage morphology and BMLs on MRI.

PLR is an index of the vertical position of the patella that is measured on the lateral X-ray view and was originally proposed by Insall and Salvati [12]. Insall and Salvati measured the PLR on the lateral X-ray of the knee at the 30° flexion. The normal ratio was suggested to be ∼1 and a ratio of 0.8 was considered to show patella baja, whilst a ratio of 1.2 is indicative of patella alta. Shabshin et al. [26] used MRIs of extended knees to measure the PLR and suggested that PLR of >1.50 or <0.74 defined patella alta and baja, respectively. A high-riding patella (patella alta) can be associated with lateral patellar dislocation and subluxation, chondromalacia patellae, patellar ligament rupture and Sinding-Larsen–Johansson disease, patellar and quadriceps tendonitis and Osgood–Schlatter disease [12, 13, 27–31]. A low-riding patella (patella baja) has been associated with quadriceps tendon rupture, neuromuscular disorders, achondroplasia and after surgical advancement of the tibial tuberosity [32, 33]. Our study demonstrated that increasing PLR (higher riding patella) was significantly associated with increasing cartilage loss in both medial and lateral compartments of PF joint and with BML in the lateral compartment. Higher riding patellae are typically located against a much shallower femoral sulcus, that can lead to instability (the evidence for this is a higher rate of patellar dislocations associated with patella alta), and therefore to accelerated cartilage loss and BML. From a biomechanical point of view, the patella increases the mechanical advantage of extensor muscles by transmitting forces across the knee at a greater distance (moment) from the axis of rotation thus increasing the functional lever arm of the quadriceps as well as changing the direction of pull of quadriceps mechanism. A relatively longer patellar tendon would decrease the mechanical advantage afforded by the patella to the functional lever arm of the quadriceps. This would increase compression in the PF joint and predispose to excessive cartilage loss and BMLs.

As one can note (Table 4), the association between PLR and lateral BML showed a statistically significant P for trend both for linear and U-shaped models. The most protective range of PLR in relation to lateral BML in our study was 0.98–1.12. Both too high (PLR: 1.13–1.71) and too low (PLR: 0.66–0.87) located patellae were associated with lateral BML.

SA is an indicator of femoral trochlear dysplasia. Femoral trochlear dysplasia (a shallower SA) is associated with patellar instability [34]. The range (98°–155°) of SA in our study was in accordance with previously published values [19, 34]. We found that a greater SA (shallower angle) was associated with greater cartilage loss (both medially and laterally) and with greater lateral BML. We hypothesize that the patellar instability that would be associated with a greater SA (shallower trochlear sulcus) can cause excessive traction and compression forces on the both sides of the PF joint and facilitate cartilage loss. The effect for BML is seen on the lateral side presumably as a result of the directional pull of the quadriceps mechanism.

In the present study, we used two indices of the PF relationship: LPTA and BO. LPTA represents the angle of patellar inclination that indicates the tightness or looseness of the lateral stabilizing mechanism of patella. BO indicates the lateral displacement of patella in relation to deepest part of the femoral sulcus. Both indices were significantly associated with cartilage loss and BML in the lateral compartment of the PF joint. LPTA also demonstrated a statistically significant P for linear and U-shape trend for medial cartilage loss. Results of BO measurements showed that in extended knees the central position of patella is an optimal situation for the PF joint. If the patella tends more laterally, it increases the risk for cartilage loss and BML in the lateral PF compartment.

The MRIs in our study were taken in a supine position with fully extended knees and relaxed quadriceps. Laterally displaced patellae (higher BO) and/or lateral border of patella too close to the lateral femoral condyle (decreased LPTA) on these images likely indicates tightness of the structures that hold the patella in a lateral position (lateral retinaculum, vastus lateralis). In this situation, during movement of the knee greater load and therefore greater shear force will be placed on the lateral PF compartment, in comparison to the situation where the patella is located directly against the trochlear sulcus and the load forces are divided equally between the lateral and medial PF compartments. Excessive load and shear forces in combination with movement can lead to attrition of cartilage and BML.

There were numerous limitations of the present study that need to be recognized. First, the MRI images were performed in a supine position of the patients and not in a weight bearing. This limitation is likely to have reduced our opportunity to measure dynamic changes in patellar position with weight bearing, and thus underscore that our findings are likely to be conservative for measures that could potentially change with weight bearing such as bisect offset and the LPTA. Another limitation of our study was that the MRI was obtained in a fully extended knee. This position, as mentioned before, is common in clinical practice, but in extended knee the patella is not positioned against trochlear sulcus and it makes the measurement of their congruence less precise. Whilst the observer reliability was good to excellent the ability to replicate the same methods at different sites was not assessed and could be impacted by such factors as the coil and subject positioning (possibly vendor and technologist specific), the subjects body habitus (most of our subjects were obese) and pain that could alter patellofemoral alignment.

Based upon the results of this study it does appear that non-weight bearing, full extension assessment of patellar alignment does increase our understanding of the reasons for cartilage loss and BML in PF joints. Further consideration of the importance of PF alignment needs to occur, preferably in more functional positions than supine and non-weight bearing. Further investigations in longitudinal studies are also needed to establish if measurements of patellar alignment on standard MRI can be used for the prediction of cartilage loss and/or BML. The findings of our study could possibly be used for the development of evaluation tools for individuals at risk for cartilage loss and BML that are part of the clinical picture of PF OA.

In conclusion, results of our study clearly suggest that patellar alignment is associated with cartilage thickness loss and BML of PF joints. We believe that those indices need to be included in a list of routine evaluations of PF MRI and evaluated further in longitudinal studies. Additional studies are required to establish the normal and abnormal ranges of patellar alignment indices.

Acknowledgements

The authors would like to thank the participants and staff of the Boston Osteoarthritis Knee Study. Supported by NIH AR47785, Osteoarthritis Biomarkers Grant from the Arthritis Foundation and by an Arthritis Foundation Clinical Sciences Grant.

The authors have declared no conflicts of interest.

References

1
Felson
DT
McAlindon
TE
Anderson
JJ
, et al.  . 
Defining radiographic osteoarthritis for the whole knee
Osteoarthritis Cartilage
 , 
1997
, vol. 
5
 (pg. 
241
-
50
)
2
Woolf
AD
Pfleger
B
Burden of major musculoskeletal conditions
Bulletin of the World Health Organization
 , 
2003
, vol. 
81
 (pg. 
646
-
56
)
3
Guermazi
A
Zaim
S
Taouli
B
Miaux
Y
Peterfy
CG
Genant
HG
MR findings in knee osteoarthritis
Eur Radiol
 , 
2003
, vol. 
13
 (pg. 
1370
-
86
)
4
Harrison
MM
Cooke
TD
Fisher
SB
Griffin
MP
Patterns of knee arthrosis and patellar subluxation
Clin Orthop Relat Res
 , 
1994
, vol. 
309
 (pg. 
56
-
63
)
5
Iwano
T
Kurosawa
H
Tokuyama
H
Hoshikawa
Y
Roentgenographic and clinical findings of patellofemoral osteoarthrosis. With special reference to its relationship to femorotibial osteoarthrosis and etiologic factors
Clin Orthop Relat Res
 , 
1990
, vol. 
252
 (pg. 
190
-
7
)
6
Fujikawa
K
Seedhom
BB
Wright
V
Biomechanics of the patello-femoral joint. Part II: a study of the effect of simulated femoro-tibial varus deformity on the congruity of the patello-femoral compartment and movement of the patella
Engg Med
 , 
1983
, vol. 
12
 (pg. 
13
-
21
)
7
Laurin
CA
Dussault
R
Levesque
HP
The tangential x-ray investigation of the patellofemoral joint: X-ray technique, diagnostic criteria and their interpretation
Clin Orthop Relat Res
 , 
1979
, vol. 
144
 (pg. 
16
-
26
)
8
Aglietti
P
Insall
JN
Cerulli
G
Patellar pain and incongruence. I: measurements of incongruence
Clin Orthop Relat Res
 , 
1983
, vol. 
176
 (pg. 
217
-
24
)
9
Insall
JN
Aglietti
P
Tria
AJ
Patellar pain and incongruence. II: Clinical application
Clin Orthop Relat Res
 , 
1983
, vol. 
176
 (pg. 
225
-
32
)
10
Murray
TF
Dupont
JY
Fulkerson
JP
Axial and lateral radiographs in evaluating patellofemoral malalignment
Am J Sports Med
 , 
1999
, vol. 
27
 (pg. 
580
-
4
)
11
Davies
AP
Costa
ML
Shepstone
L
Glasgow
MM
Donell
S
The sulcus angle and malalignment of the extensor mechanism of the knee
J Bone Joint Surg Br
 , 
2000
, vol. 
82
 (pg. 
1162
-
6
)
12
Insall
J
Salvati
E
Patella position in the normal knee joint
Radiology
 , 
1971
, vol. 
101
 (pg. 
101
-
4
)
13
Grelsamer
RP
Meadows
S
The modified Insall-Salvati ratio for assessment of patellar height
Clin Orthop
 , 
1992
, vol. 
282
 (pg. 
170
-
6
)
14
Ficat
RP
Hungerford
DS
Disorders of the patello-femoral Joint.
 , 
1977
Baltimore
The Williams and Wilkins Co.
15
Bull
AMJ
Katchburian
MV
Shih
YF
Amis
AA
Standardisation of the description of patellofemoral motion and comparison between different techniques
Knee Surg Sports Traumatol Arthrosc
 , 
2002
, vol. 
10
 (pg. 
184
-
93
)
16
Merchant
AC
Mercer
RL
Jacobsen
RH
Cool
CR
Roentgenographic analysis of patellofemoral congruence
J Bone Joint Surg Am
 , 
1974
, vol. 
56
 (pg. 
1391
-
6
)
17
Beaconsfield
T
Pintore
E
Maffulli
N
Petri
J
Radiological measurements in patellofemoral disorders. A review
Clin Orthop
 , 
1994
, vol. 
308
 (pg. 
18
-
28
)
18
Felson
DT
Chaisson
CE
Hill
CL
, et al.  . 
The association of bone marrow lesions with pain in knee osteoarthritis
Ann Intern Med
 , 
2001
, vol. 
134
 (pg. 
541
-
9
)
19
Muellner
T
Funovics
M
Nikolic
A
Metz
V
Schabus
R
Vecsei
V
Patellar alignment evaluated by MRI
Acta Orthop Scand
 , 
1998
, vol. 
69
 (pg. 
489
-
92
)
20
Carrillon
Y
Abidi
H
Dejour
D
Fantino
O
Moyen
B
Tran-Minh
VA
Patellar instability: assessment on MR images by measuring the lateral trochlear inclination-initial experience
Radiology
 , 
2000
, vol. 
216
 (pg. 
582
-
5
)
21
Kobayashi
T
Fujikawa
K
Nemoto
K
Yamazaki
M
Obara
M
Sato
S
Evaluation of patello-femoral alignment using MRI
Knee
 , 
2005
, vol. 
12
 (pg. 
447
-
53
)
22
Felson
DT
McLaughlin
S
Goggins
J
, et al.  . 
Bone marrow edema and its relation to progression of knee osteoarthritis
Ann Intern Med
 , 
2003
, vol. 
139
 (pg. 
330
-
6
)
23
Altman
R
Asch
E
Bloch
D
, et al.  . 
Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association
Arthritis Rheum
 , 
1986
, vol. 
29
 (pg. 
1039
-
49
)
24
Peterfy
CG
Guermazi
A
Zaim
S
, et al.  . 
Whole-Organ Magnetic Resonance Imaging Score (WORMS) of the knee in osteoarthritis
Osteoarthritis Cartilage
 , 
2004
, vol. 
12
 (pg. 
177
-
90
)
25
Amin
S
LaValley
MP
Guermazi
A
, et al.  . 
The relationship between cartilage loss on magnetic resonance imaging and radiographic progression in men and women with knee osteoarthritis
Arthritis Rheum
 , 
2005
, vol. 
52
 (pg. 
3152
-
9
)
26
Shabshin
N
Schweitzer
ME
Morrison
WB
Parker
L
MRI criteria for patella alta and baja
Skeletal Radiol
 , 
2004
, vol. 
33
 (pg. 
445
-
50
)
27
Von Gumppenberg
S
Jacob
RP
Engelhardt
P
Does Osgood-Schlatter disease modify the position of the patella?
Z Orthop Inhre Brenzgeb
 , 
1984
, vol. 
122
 (pg. 
798
-
802
)
28
Grana
WA
Krieghauser
LA
Scientific basis of extensor mechanism disorders
Clin Sports Med
 , 
1985
, vol. 
4
 (pg. 
247
-
57
)
29
Kannus
PA
Long patellar tendon: radiographic sign of patellofemoral pain syndrome
Radiology
 , 
1992
, vol. 
185
 (pg. 
859
-
63
)
30
Miller
TT
Staron
RB
Feldman
F
Patellar height on sagittal MR imaging of the knee
AJR
 , 
1996
, vol. 
167
 (pg. 
339
-
41
)
31
Aparicio
G
Abril
JC
Calvo
E
Alvarez
L
Radiographic study of patellar height in Osgood-Schlatter disease
J Pediatr Orthop
 , 
1997
, vol. 
17
 (pg. 
63
-
6
)
32
Noyes
FR
Wojtys
EM
Marshall
MT
The early diagnosis and treatment of developmental patella infera syndrome
Clin Orthop
 , 
1991
, vol. 
265
 (pg. 
241
-
52
)
33
Tria
AJ
JR
Alicea
JA
Cody
RP
Patella baja in anterior crutiate ligament reconstruction of the knee
Clin Orthop
 , 
1994
, vol. 
299
 (pg. 
229
-
34
)
34
Brattström
H
Shape of the intercondylar groove normally and in recurrent dislocation of the patella
Acta Scand
 , 
1964
, vol. 
68
 
Suppl
(pg. 
1
-
148
)

Comments

0 Comments