Differences between X-ray and MRI-determined knee cartilage thickness in weight-bearing and non-weight-bearing conditions

OBJECTIVE

Determine the effect of loading upon MRI-based mean medial femorotibial cartilage thickness (mMFT_th) and radiograph-based minimum joint space width (mJSW), and determine loading's effect on the relationship between these measures.

METHODS

MRI and radiographs were analyzed of 25 knees in weight-bearing and non-weight-bearing conditions. Eight subjects had a Kellgren-Lawrence (KL) grade of 0, indicating no evidence of radiographic OA. The rest were KL = 2 or KL = 3, indicating mild to moderate OA. The change from unloaded to loaded conditions was calculated.

RESULTS

Joint space measures decreased from unloaded to loaded conditions for both radiographs (mJSW = 3.29 mm unloaded to 3.16 mm loaded, P < 0.05) and MRI (mMFT_th = 2.70 mm unloaded to 2.55 mm loaded P < 0.001). The mean absolute difference measured from radiographs was larger for the OA group than the control group, at -0.20 mm for OA vs +0.01 mm for control. Loaded X-ray and loaded MRI joint space values from our study were no better correlated to one another than loaded X-ray and unloaded MRI.

CONCLUSION

Knee loading does not add a very significant value to the study of joint space on healthy knees, but loading may play a role in the study of OA knees. Unloaded MRI assessments of cartilage thickness are as correlated to loaded JSW as to loaded MRI measurements. More study is necessary to determine whether loaded MRI adds significant value to the study of OA progression.

Association of MR relaxation and cartilage deformation in knee osteoarthritis

We assessed the relationship between cartilage MR relaxation times and biomechanical response of tibiofemoral articular cartilage to physiological loading in healthy subjects and patients with osteoarthritis (OA). Female subjects above 40 years of age with (N(1)  = 20) and without (N(2)  = 10) OA were imaged on a 3T MR scanner using a custom made loading device. MR images were acquired with the knee flexed at 20° with and without a compressive load of 50% of the subject's bodyweight. The subjects were categorized based on the clinical MRI scoring of medial and lateral cartilage surfaces. Data were stratified twice into two equal groups (low and high) at the median value of T(1ρ) and T(2) relaxation time. The change in contact area and cartilage deformation was measured within these groups. Paired Student's t-test (α = 0.05) was used to analyze the effect of loading on contact area and deformation. The average area of the contact region in the medial compartment was significantly higher in OA subjects compared with normal subjects in both unloaded (314 ± 112 mm(2) vs. 227 ± 106 mm(2), p = 0.023) and loaded (425 ± 128 mm(2) vs. 316 ± 107 mm(2), p = 0.01) conditions. The overall relative change of cartilage thickness in the medial compartment was significantly higher than the lateral compartment (-5.3 ± 9.9% vs. -1.9 ± 9.2%, p = 0.042). When cartilage was divided into deep and superficial layers, superficial layers showed higher changes in relaxation time (T(1ρ) and T(2)) than the changes in relaxation time of whole cartilage (Normal: 12.5% vs. 6.9%; OA: 10.9% vs. 4.6%). The average T(1ρ) and T(2) times, change in area of contact region, and change in cartilage thickness in subjects with OA were higher when compared to normal subjects. This study provides support for a relationship between the mechanical response of cartilage to physiological loading (cartilage-on-cartilage contact area and cartilage deformation) and MR relaxation times (T(1ρ) and T(2)) in both OA patients and normal subjects.

Clinical, radiographic, molecular and MRI-based predictors of cartilage loss in knee osteoarthritis

OBJECTIVE

To examine the relationship of baseline clinical, radiographic, molecular and MRI measures with structural progression (subregional MRI-based femorotibial cartilage loss) in knee osteoarthritis (OA).

METHODS

Single knees of 75 female participants with radiographic knee OA (and 77 healthy control participants) were examined over 24 months using MRI. Subregional femorotibial cartilage thickness was determined at baseline and follow-up. Baseline clinical, radiographic, molecular (n=16) and quantitative MRI-based measures of the meniscus and cartilage, including delayed gadolinium-enhanced MRI (dGEMRIC) and T2, were obtained. Differences in these baseline measures between radiographic osteoarthritic knees with longitudinal cartilage thinning (or thickening) and those with no significant change were evaluated by receiver operator characteristic analyses and Wilcoxon rank sum tests.

RESULTS

The relatively strongest predictors of longitudinal cartilage thinning were reduced baseline cartilage thickness in the medial femur (area under the curve (AUC)=0.81), varus malalignment (AUC=0.77), reduced minimum joint space width and a greater radiographic joint space narrowing (JSN) score (both AUC=0.74). These remained significant after adjusting for multiple comparisons using false discovery rates. Reduced bone resorption (C-terminal telopeptide of type I collagen; AUC=0.65) and a low dGEMRIC index (reflecting low proteoglycan content) in the medial tibia (AUC=0.68) were associated with longitudinal cartilage thinning, but failed to reach statistical significance after correction for multiple testing in this (small) sample.

CONCLUSIONS

This exploratory study indicates that baseline molecular or MRI cartilage compositional markers may not provide better discrimination between knees with cartilage thinning and those without longitudinal change than simple radiographic measures, such as greater JSN score.

Using ordered values of subregional cartilage thickness change increases sensitivity in detecting risk factors for osteoarthritis progression

OBJECTIVE

To examine whether ordered values of (sub)regional femorotibial cartilage thickness change are superior to region-based approaches in detecting risk factors for cartilage loss in osteoarthritis (OA).

METHODS

58 women with knee OA had 3 Tesla MR images acquired at baseline and 24 months. Changes in cartilage thickness (∆ThCtAB) were determined in eight medial femorotibial subregions. An ascending sort of individual ∆ThCtAB measurements was done to create "ordered values". Risk factors for cartilage loss considered were: age, BMI, anatomical knee axis (AAA), minimal (medial) joint space width (mJSW), and percent of medial tibial plateau covered by the meniscus (percent cover). All change metrics were tested for association with the risk factors using Kendall's τ and relative sensitivity of multiple tests of subregions and ordered values were compared with single metrics of change from plate and compartment summaries and the first ordered value.

RESULTS

The associations between subregion ∆ThCtAB and AAA (P=0.0002), mJSW (P=0.016), and age (P=0.011) were significant, but only AAA (at α=0.05) and age (at α=0.1) remained significant after adjusting for multiple subregions. In contrast, cMFTC had P-values<0.05 for AAA (P=0.0001), mJSW (P=0.016), and meniscus subluxation (0.04). The first ordered value had significant associations with AAA (P=0.0004), mJSW (P=0.003), meniscus subluxation (P=0.02) and percent cover (P=0.031) all of which were significant at α=0.05 after adjusting for tests on multiple risk factors.

CONCLUSION

Ordered values of ∆ThCtAB were more sensitive in detecting risk factors of cartilage loss than subregional ∆ThCtAB. Sensitivity was further enhanced by considering the minimum ordered value as a single test, thus not requiring adjustment for multiple tests. Using ordered values there was a significant association between ∆ThCtAB and baseline AAA, mJSW, meniscus subluxation and meniscus percent cover. This study provides an important step in validating ordered values of cartilage change.

Comparison of 1-year vs 2-year change in regional cartilage thickness in osteoarthritis results from 346 participants from the Osteoarthritis Initiative

OBJECTIVE

To compare femorotibial cartilage thickness changes over a 2- vs a 1-year observation period in knees with radiographic knee osteoarthritis (OA).

METHODS

One knee of 346 Osteoarthritis Initiative (OAI) participants was studied at three time points [baseline (BL), year-1 (Y1), year-2 (Y2) follow-up]: 239 using coronal fast low angle shot (FLASH) and 107 using sagittal double echo at steady state (DESS) MR imaging. Changes in cartilage thickness were assessed in femorotibial cartilage plates and subregions, after manual segmentation with blinding to time-point.

RESULTS

The standardized response mean (SRM) of total joint cartilage thickness over 2 years was modestly higher than over 1 year (FLASH: -0.44 vs -0.32/-0.28 [first/second year]; DESS: -0.42 vs -0.39/-0.18). For the subregion showing the largest change per knee (OV1), the 2-year SRM was similar or lower (FLASH: -1.20 vs -1.22/-1.61; DESS: -1.38 vs -1.64/-1.51) than the 1-year SRM. The changes in total joint cartilage thickness were not significantly different in the first and second year (FLASH: -0.8% vs -0.7%; DESS: -1.3% vs -0.8%) and were negatively correlated. Analysis of smallest detectable changes (SDCs) revealed that only few participants displayed significant progression in both consecutive periods. The location of the subregion contributing to OV1 in each knee was highly inconsistent between the first and second year observation period.

CONCLUSIONS

The SRM of region-based cartilage thickness change in OA is modestly larger following a 2-year vs a 1-year observation period, while it is relatively similar when an OV-approach is chosen. Structural progression displays strong temporal and spatial heterogeneity at an individual knee level that should be considered when planning clinical trials.

The effects of acute loading on T1rho and T2 relaxation times of tibiofemoral articular cartilage

OBJECTIVE

To evaluate the effect of acute loading on healthy and osteoarthritic knee cartilage T(1ρ) and T(2) relaxation times.

DESIGN

Twenty subjects with radiographic evidence of osteoarthritis (OA) and 10 age-matched controls were enrolled. Magnetic resonance imaging (MRI) acquisition, including T(1ρ) and T(2) map sequences were performed unloaded and loaded at 50% body mass. Cartilage masks were segmented semi-automatically on registered high-resolution spoiled gradient-echo (SPGR) images for each compartment (medial and lateral). Cartilage lesions were identified using a modified Whole Organ Magnetic Resonance Imaging Score (WORMS) score. Statistical differences were explored using separate two-way (group×loading condition) Analysis of Variance (ANOVA) using age as a covariate to evaluate the effects of loading on T(1ρ) and T(2) relaxation times.

RESULTS

A significant decrease in T(1ρ) (44.5±3.8 vs 40.2±4.8ms for unloaded and loaded, respectively; P<0.001) and T(2) (31.8±3.8 vs 30.5±4.8ms for unloaded and loaded, respectively; P<0.001) relaxation times was observed in the medial compartment with loading while no differences were observed in the lateral compartment. This behavior occurred independent of WORMS score. Cartilage compartments with small focal lesions experienced greater T(1ρ) change scores with loading when compared to cartilage without lesions or cartilage with larger defects (P=0.05).

CONCLUSIONS

Acute loading resulted in a significant decrease in T(1ρ) and T(2) relaxation times of the medial compartment, with greater change scores observed in cartilage regions with small focal lesions. These data suggest that changes of T(1ρ) values with loading may be related to cartilage biomechanical properties (i.e., tissue elasticity) and may be a valuable tool for the scientist and clinician at identifying early cartilage disease.

Relation of regional articular cartilage morphometry and meniscal position by MRI to joint space width in knee radiographs

OBJECTIVE

To ascertain the contribution of articular cartilage morphometry and meniscal position on MRI to joint space width (JSW) measured in the Lyon schuss radiograph of the knee.

DESIGN

62 obese women with knee OA and 99 non-obese female controls (mean age 56.6 years) were imaged using 3T MRI and coronal water excitation spoiled gradient echo sequences. Segmentation of femorotibial cartilage morphology and regional morphometric analysis was performed using custom software. Meniscal position was measured quantitatively in sagittal and coronal planes. Minimum space width (mJSW) was measured in the Lyon Schuss knee radiograph; Kellgren and Lawrence grades (KLG) were assigned on standing anteroposterior knee films. The relative contribution of regional cartilage thickness and meniscal position to mJSW was assessed initially in univariate models and subsequently with multivariable modelling.

RESULTS

65% of the variation in mJSW was explained by regional cartilage thickness measures, different KLG and meniscal coverage. Of these measures the medial tibia cartilage thickness measures and central region of the central medial femur (ccMF) play a consistent role in variations in mJSW observed across all KLG. Further ccMF and the addition of percent meniscal coverage to this model explains the remaining differences in mean mJSW found between those subjects with definite joint space narrowing (KLG3) and those without OA.

CONCLUSION

The variation in radiographic mJSW is best described by five regional cartilage thickness measures and percent meniscal coverage. The magnitude of each measures contribution differs according to radiographic severity with more variability explained by cartilage thickness of ccMF cartilage thickness and percent meniscal coverage with more severe disease.

Subregional femorotibial cartilage morphology in women--comparison between healthy controls and participants with different grades of radiographic knee osteoarthritis

OBJECTIVE

To identify subregional differences in femorotibial cartilage morphology between healthy controls and women with different grades of radiographic knee osteoarthritis (OA).

DESIGN

158 women aged > or =40 years were studied. Weight-bearing extended anterior-posterior (AP) and Lyon schuss radiographs were obtained and the Kellgren Lawrence grade (KLG) determined. 97 women had a body mass index (BMI)< or =28, no symptoms, and were AP KLG0. 61 women had a BMI> or =30, symptoms in the target knee, and mild (KLG2=31) to moderate (KLG3=30) medial femorotibial radiographic OA in the AP views. Coronal spoiled gradient echo water excitation sequences were acquired at 3.0 Tesla. Total plate and regional measures of cartilage morphology of the weight-bearing femorotibial joint were quantified.

RESULTS

KLG2 participants displayed, on average, thicker cartilage than healthy controls in the medial femorotibial compartment (particularly anterior subregion of the medial tibia (MT) and peripheral [external, internal] subregions of the medial femur), and in the lateral femur. KLG3 participants displayed significantly thinner cartilage than KLG0 participants in the medial weight-bearing femur (central subregion), in the external subregion of the MT, and in the internal subregion of the lateral tibia. These differences were generally unaffected when possible effects of demographic covariates were considered.

CONCLUSIONS

The results indicate that in femorotibial OA regional cartilage thickening and thinning may occur, dependent on the (radiographic) disease status of the joint. These changes appear to display a heterogeneous spatial pattern, where certain subregions are more strongly affected than others.

Change in regional cartilage morphology and joint space width in osteoarthritis participants versus healthy controls: a multicentre study using 3.0 Tesla MRI and Lyon–Schuss radiography

Objective:

Cartilage morphology displays sensitivity to change in osteoarthritis (OA) with quantitative MRI (qMRI). However, (sub)regional cartilage thickness change at 3.0 Tesla (T) has not been directly compared with radiographic progression of joint space narrowing in OA participants and non-arthritic controls.

Methods:

A total of 145 women were imaged at 7 clinical centres: 86 were non-obese and asymptomatic without radiographic OA and 55 were obese with symptomatic and radiographic OA (27 Kellgren–Lawrence grade (KLG)2 and 28 KLG3). Lyon–Schuss (LS) and fixed flexion (FF) radiographs were obtained at baseline, 12 and 24 months, and coronal spoiled gradient echo MRI sequences at 3.0 T at baseline, 6, 12 and 24 months. (Sub)regional, femorotibial cartilage thickness and minimum joint space width (mJSW) in the medial femorotibial compartment were measured and the standardised response means (SRMs) determined.

Results:

At 6 months, qMRI demonstrated a −3.7% “annualised” change in cartilage thickness (SRM −0.33) in the central medial femorotibial compartment (cMFTC) of KLG3 subjects, but no change in KLG2 subjects. The SRM for mJSW in 12-month LS/FF radiographs of KLG3 participants was −0.68/−0.13 and at 24 months was −0.62/−0.20. The SRM for cMFTC changes measured with qMRI was −0.32 (12 months; −2.0%) and −0.48 (24 months; −2.2%), respectively.

Conclusions:

qMRI and LS radiography detected significant change in KLG3 participants at high risk of progression, but not in KLG2 participants, and only small changes in controls. At 12 and 24 months, LS displayed greater, and FF less, sensitivity to change in KLG3 participants than qMRI.

Predictors of systolic augmentation from left ventricular preexcitation in patients with dilated cardiomyopathy and intraventricular conduction delay

BACKGROUND

VDD pacing can enhance systolic function in patients with dilated cardiomyopathy and discoordinate contraction; however, identification of patients likely to benefit is unclear. We tested predictors of systolic responsiveness on the basis of global parameters as well as directly assessed mechanical dyssynchrony.

METHODS AND RESULTS

Twenty-two DCM patients with conduction delay were studied by cardiac catheterization with a dual-sensor micromanometer to measure LV and aortic pressures during sinus rhythm and LV free-wall pacing. Pacing enhanced isovolumetric (dP/dt(max)) and ejection-phase (pulse pressure, PP) systolic function by 35+/-21% and 16.4+/-11%, respectively, and these changes correlated directly (r=0.7, P=0.001). %DeltadP/dt(max) was weakly predicted by baseline QRS (r=0.6, P<0.02), more strongly by baseline dP/dt(max) (r=0.7, P=0.001), and best by bidiscriminate analysis combining baseline dP/dt(max) < or =700 mm Hg/s and QRS > or =155 ms to predict %DeltadP/dt(max) > or =25% and %DeltaPP > or =10% (P<0.0005, chi(2)), with no false-positives. Benefit could not be predicted by %DeltaQRS. To test whether basal mechanical dyssynchrony predicted responsiveness to LV pacing, circumferential strains were determined at approximately 80 sites throughout the LV by tagged MRI in 8 DCM patients and 7 additional control subjects. Strain variance at time of maximal shortening indexed dyssynchrony, averaging 28.0+/-7.1% in normal subjects versus 201.4+/-84.3% in DCM patients (P=0.001). Mechanical dyssynchrony also correlated directly with %DeltadP/dt(max) (r=0.85, P=0.008). Conclusions-These results show that although mechanical dyssynchrony is a key predictor for pacing efficacy in DCM patients with conduction delay, combining information about QRS and basal dP/dt(max) provides an excellent tool to identify maximal responders.

Mapping of regional myocardial strain and work during ventricular pacing: experimental study using magnetic resonance imaging tagging

OBJECTIVES

The purpose of this study was to determine the spatial distribution of myocardial function (myofiber shortening and work) within the left ventricular (LV) wall during ventricular pacing.

BACKGROUND

Asynchronous electrical activation, as induced by ventricular pacing, causes various abnormalities in LV function, perfusion and structure. These derangements may be caused by abnormalities in regional contraction patterns. However, insight into these patterns during pacing is as yet limited.

METHODS

In seven anesthetized dogs, high spatial and temporal resolution magnetic resonance-tagged images were acquired in three orthogonal planes. Three-dimensional deformation data and LV cavity pressure and volume were used to determine midwall circumferential strain and external and total mechanical work at 192 sites around the left ventricle.

RESULTS

During ventricular pacing, systolic fiber strain and external work were approximately zero in regions near the pacing site, and gradually increased to more than twice the normal value in the most remote regions. Total mechanical work, normalized to the value during right atrial pacing, was 38 +/- 13% (right ventricular apex [RVapex] pacing) and 61 +/- 23% (left ventricular base [LVbase] pacing) close to the pacing site, and 125 +/- 48% and 171 +/- 60% in remote regions, respectively (p < 0.05 between RVapex and LVbase pacing). The number of regions with reduced work was significantly larger during RVapex than during LVbase pacing. This was associated with a reduction of global LV pump function during RVapex pacing.

CONCLUSIONS

Ventricular pacing causes a threefold difference in myofiber work within the LV wall. This difference appears large enough to regard local myocardial function as an important determinant for abnormalities in perfusion, metabolism, structure and pump function during asynchronous electrical activation. Pacing at sites that cause more synchronous activation may limit the occurrence of such derangements.

Mapping propagation of mechanical activation in the paced heart with MRI tagging

The temporal evolution of three-dimensional (3-D) strain maps derived from magnetic resonance imaging (MRI) tagging were used to noninvasively evaluate mechanical activation in the left ventricle (LV) while seven canine hearts were paced in situ from three different sites: the base of the LV free wall (LVb), the right ventricular apex (RVa), and the right atrium (RA). Strain maps plotted against time showed the evolution of shortening over the entire LV midwall and were used to generate mechanical activation maps showing the onset of circumferential shortening. RA pacing showed rapid synchronous shortening; LVb pacing showed a wave front of mechanical activation propagating slowly and steadily from the pacing site, whereas RVa pacing showed regions of rapid and slower propagation. The mechanical (M) activation times correlated linearly with the electrical (E) activation (M = 1.06E + 8.4 ms, R = 0.95). The time for 90% activation of the LV was 63.1 +/- 24.3 ms for RA pacing, 130.2 +/- 9.8 ms for LVb pacing, and 121.3 +/- 17.9 ms for RVa pacing. The velocity of mechanical activation was calculated for LVb and RVa pacing and was similar to values reported for electrical conduction in myocardium. The propagation of mechanical activation for RVa pacing showed regional variations, whereas LVb pacing did not.

Imaging asynchronous mechanical activation of the paced heart with tagged MRI

A method for imaging the rapid temporal-spatial evolution of myocardial deformations in the paced heart is proposed. High time resolution-tagged MR images were obtained after stimulation of the myocardium with an MR-compatible pacing system. The images were analyzed to reconstruct dynamic models of local 3D strains over the entire left ventricle during systole. Normal canine hearts were studied in vivo with pacing sites on the right atrium, left ventricular free wall and right ventricular apex. This method clearly resolved local variations in myocardial contraction patterns caused by ventricular pacing. Potential applications are noninvasive determination of electrical conduction abnormalities and the evaluation of new pacing therapies.

The representation of concurrent vowels in the cat anesthetized ventral cochlear nucleus: evidence for a periodicity-tagged spectral representation

Chopper units of the ventral cochlear nucleus (VCN) provide a rare representation of stimulus spectrum and a temporal representation of fundamental frequency (F0). This dual representation may be useful in segregating competing speech sounds, where differences in F0 are a cue. Responses to the vowel portion of concurrently presented pairs of syllables /bV integral/ with different F0's (88, 98, and 112 Hz) were studied in the VCN of anesthetized cats; 11 English vowels were used for V. Vowels were chosen so that one had a formant frequency just above the unit's best frequency (BF) and the other had a formant just below BF. By changing the stimulus sampling rate, formant peaks were shifted relative to the unit's BF, producing a range of stimuli, varying in the relative power of the two vowels within the unit's tuning curve. Results show that units' discharge rates reflect the energy within their tuning curves and the relative synchronization of units' responses to the two F0's favors the dominant vowel. A method of segregating two vowels is provided in which relative synchronization to the F0's is used to apportion discharge rate between the vowels. Best results were obtained in chopper units, although primarylike units showed similar behavior.