Diffusion-weighted MRI reveals epiphyseal and metaphyseal abnormalities in Legg-Calvé-Perthes disease: a pilot study.

BACKGROUND: Legg-Calvé-Perthes disease (LCP) is thought to be associated with ischemic events in the femoral head. However, the types and patterns of reperfusion after these ischemic events are unclear. PURPOSES: We therefore determined whether (1) there would be any age-related diffusion changes; (2) diffusion-weighted MR imaging would reveal ischemic damage; and (3) diffusion changes are correlated with prognostic MR findings in patients with LCP. METHODS: We prospectively performed conventional, perfusion, and diffusion-weighted MR imaging studies in 17 children with unilateral LCP. We then measured the apparent diffusion coefficient (ADC) values in the epiphysis and the metaphysis, and compared them with those of the contralateral normal side. Based on perfusion MR imaging, we assessed reperfusion to the epiphysis as either periphyseal or transphyseal. We studied T2-signal intensity changes in the metaphysis and the presence of focal physeal irregularity. We correlated diffusion changes with reperfusion to the epiphysis, T2-signal intensity change, and focal physeal irregularity. RESULTS: Normal diffusion decreased with age. In LCP hips, epiphyseal diffusion increased early and remained elevated through the healing stage. Six of the 17 patients who had a metaphyseal ADC greater than 50% over the normal side had 13 times greater odds of having an association with transphyseal reperfusion to the epiphysis. The increase of metaphyseal ADC also was associated with an increased T2-signal intensity in the metaphysis and presence of focal physeal irregularity. CONCLUSIONS: Diffusion-weighted MR imaging can be used as a complimentary modality to evaluate ischemic tissue damage with a potential prognostic value in patients with LCP.

The real estate factor: quantifying the impact of infarct location on stroke severity.

BACKGROUND AND PURPOSE: The severity of the neurological deficit after ischemic stroke is moderately correlated with infarct volume. In the current study, we sought to quantify the impact of location on neurological deficit severity and to delineate this impact from that of volume. METHODS: We developed atlases consisting of location-weighted values indicating the relative importance in terms of neurological deficit severity for every voxel of the brain. These atlases were applied to 80 first-ever ischemic stroke patients to produce estimates of clinical deficit severity. Each patient had an MRI and National Institutes of Health Stroke Scale (NIHSS) examination just before or soon after hospital discharge. The correlation between the location-based deficit predictions and measured neurological deficit (NIHSS) scores were compared with the correlation obtained using volume alone to predict the neurological deficit. RESULTS: Volume-based estimates of neurological deficit severity were only moderately correlated with measured NIHSS scores (r=0.62). The combination of volume and location resulted in a significantly better correlation with clinical deficit severity (r=0.79, P=0.032). CONCLUSIONS: The atlas methodology is a feasible way of integrating infarct size and location to predict stroke severity. It can estimate stroke severity better than volume alone.

Conversion of ischemic brain tissue into infarction increases with age.

BACKGROUND AND PURPOSE: Brain regions normal on diffusion-weighted imaging (DWI) but abnormal on mean transit time (MTT) maps represent tissue at risk of infarction, yet the fate of these regions is quite variable. The imperfect correlation between tissue outcome and initial imaging parameters suggests that each patient's brain may have different susceptibility to ischemic stress. We hypothesize that age is a marker for tissue susceptibility to ischemia and thus plays a role in determining tissue outcome in human stroke. METHODS: Sixty patients with acute ischemic stroke and a region of DWI/MTT mismatch that was >20% of the DWI volume were included. All patients were scanned twice, within 12 hours of symptom onset and on day 5 or later. The percentage mismatch lost (PML) was calculated as percentage of initial DWI/MTT mismatch volume that was infarcted on the follow-up MRI. The statistical analysis explored relationships among the covariates age, Trial of Org 10172 in Acute Stroke Treatment (TOAST) subtypes, time-to-MRI, and initial DWI, MTT volume, mean arterial blood pressure and blood glucose level at admission, and previous history of hypertension and diabetes mellitus. RESULTS: Univariate comparisons showed that age (P=0.003), hypertension (P=0.009), and diabetes mellitus (P=0.0002) were significantly associated with PML. Regression analyses showed age to be a significant covariate (P=0.02). The regression model predicted a change in PML of approximately 0.65% per year. The adjusted proportion of variance (R2) in PML that could be explained by age alone was 14%. CONCLUSIONS: Age-dependent increase in conversion of ischemic tissue into infarction suggests that age is a biological marker for the variability in tissue outcome in acute human stroke.

Femoral head deformation and repair following induction of ischemic necrosis: a histologic and magnetic resonance imaging study in the piglet.

BACKGROUND: Ischemic necrosis of the femoral head can be induced surgically in the piglet. We used this model to assess femoral head deformation and repair in vivo by sequential magnetic resonance imaging and by correlating end-stage findings with histologic assessments. METHODS: Ischemic necrosis of the femoral head was induced in ten three-week-old piglets by tying a silk ligature around the base of the femoral neck (intracapsular) and cutting the ligamentum teres. We used magnetic resonance imaging with the piglets under general anesthesia to study the hips at forty-eight hours and at one, two, four, and eight weeks. Measurements on magnetic resonance images in the midcoronal plane of the involved and control sides at each time documented the femoral head height, femoral head width, superior surface cartilage height, and femoral neck-shaft angle. Histologic assessments were done at the time of killing. RESULTS: Complete ischemia of the femoral head was identified in all involved femora by magnetic resonance imaging at forty-eight hours. Revascularization began at the periphery of the femoral head as early as one week and was underway in all by two weeks. At eight weeks, magnetic resonance imaging and histologic analysis showed deformation of the femoral head and variable tissue deposition. Tissue responses included (1) vascularized fibroblastic ingrowth with tissue resorption and cartilage, intramembranous bone, and mixed fibro-osseous or fibro-cartilaginous tissue synthesis and (2) resumption of endochondral bone growth. At eight weeks, the mean femoral head measurements (and standard error of the mean) for the control compared with the ligated femora were 10.4 +/- 0.4 and 4.8 +/- 0.4 mm, respectively, for height; 26.7 +/- 0.8 and 31.2 +/- 0.8 mm for diameter; 1.1 +/- 0.1 and 2.3 +/- 0.1 mm for cartilage thickness; and 151 degrees +/- 2 degrees and 135 degrees +/- 2 degrees for the femoral neck-shaft angle. Repeated-measures mixed-model analysis of variance revealed highly significant effects of ligation in each parameter (p < 0.0001). CONCLUSIONS: Magnetic resonance imaging allows for the assessment of individual hips at sequential time periods to follow deformation and repair. There was a variable tissue response, and histologic assessment at the time of killing was shown to correlate with the evolving and varying magnetic resonance imaging signal intensities. Femoral head height on the ischemic side from one week onward was always less than the initial control value and continually decreased with time, indicating collapse as well as slowed growth. Increased femoral head width occurred relatively late (four to eight weeks), indicating cartilage model overgrowth concentrated at the periphery.

Physiological noise in MR images: an indicator of the tissue response to ischemia?

PURPOSE: To determine whether measuring signal intensity (SI) fluctuations in MRI time series data from acute stroke patients would identify ischemic tissue. MATERIALS AND METHODS: Prebolus perfusion-weighted MRI data from 32 acute ischemic stroke patients (N = 32) was analyzed as a time series. Ischemic and normal tissue regions were outlined and compared. RESULTS: The magnitude of the measured SI fluctuations was significantly lower in ischemic regions relative to normal tissue. Spatial differences in these fluctuations occurred in a manner that was different than other perfusion-based metrics. CONCLUSION: Prior studies have shown that SI fluctuations in MRI time series data correspond to the presence of physiological "noise," which includes vasomotion, an autoregulatory phenomenon that affects the tissue response to ischemia. In this study, SI fluctuations were found to decrease in ischemia, consistent with the notion that small vessels will remain open (fluctuations in vessel diameter will decrease) when there is a challenge to flow. Spatial variation in SI fluctuations appeared to be different from spatial variation seen on other perfusion-based metrics, suggesting that a separate contrast mechanism is responsible, one that might be of diagnostic and prognostic value in acute stroke in which the ability of tissue to withstand ischemia is currently not well visualized.

Early ischemia in growing piglet skeleton: MR diffusion and perfusion imaging.

PURPOSE: To determine whether diffusion changes with ischemia of increasing duration, whether diffusion magnetic resonance (MR) imaging provides different information than does gadolinium-enhanced imaging, and which structural and/or biochemical changes are potentially responsible for any changes in diffusion. MATERIALS AND METHODS: Ischemia was surgically induced in one hip of each piglet (n=8) after approval from the Subcommittee on Research Animal Care; the other hip served as a control. Piglets were imaged at approximately 48 hours and 1, 2, 4, and 8 weeks after surgery at 1.5 T by using line-scan diffusion and dynamic gadolinium-enhanced MR imaging. Apparent diffusion coefficients (ADCs) and enhancement ratios (ERs) were calculated. Significant differences in ADC and ER values over time were evaluated by using the Student t test (P<.05). At 8 weeks, piglets were sacrificed for histologic evaluation. RESULTS: MR images of ischemic hips showed essentially no flow 48 hours after surgery. Spontaneous partial reperfusion was observed 1-4 weeks after surgery (ischemic ER/control ER=66%+/-35 [standard deviation]), and the ER of the ischemic hips was well above that of the control hips at 8 weeks. The ADC of ischemic hips was elevated above that of control hips before reperfusion 1 week after surgery by 47%+/-12 and remained elevated despite flow restoration. Gross structural abnormalities on MR images appeared to coincide with reperfusion. Histologic findings revealed abnormal epiphyseal cartilage thickening, cartilaginous islands within ossified tissue, and less fatty marrow in ischemic hips than in control hips; all of these factors could explain elevated ADC. CONCLUSION: Diffusion is sensitive to early ischemia and follows a different time course than that of changes observed with gadolinium enhancement. ADC remained elevated in this model of severe, prolonged ischemia despite the spontaneous partial restoration of blood flow seen on gadolinium-enhanced images.

Gadolinium-enhanced MR images of the growing piglet skeleton: ionic versus nonionic contrast agent.

PURPOSE: To determine whether there are differences in the distribution of ionic and nonionic gadolinium-based contrast agents by evaluating contrast enhancement of the physis, epiphyseal cartilage, secondary ossification center, and metaphysis in the knees of normal piglets. MATERIALS AND METHODS: Following approval from the Subcommittee on Research Animal Care, knees of 12 3-week-old piglets were imaged at 3-T magnetic resonance (MR) imaging after intravenous injection of gadoteridol (nonionic contrast agent; n = 6) or gadopentetate dimeglumine (ionic contrast agent; n = 6). Early enhancement evaluation with gradient-echo MR imaging was quantified and compared (Student t test) by means of enhancement ratios. Distribution of contrast material was assessed and compared (Student t test) by means of T1 measurements obtained before and at three 15-minute intervals after contrast agent administration. The relative visibility of the physis, epiphyseal cartilage, secondary ossification center, and metaphysis was qualitatively assessed by two observers and compared (Wilcoxon signed rank test). Differences in matrix content and cellularity that might explain the imaging findings were studied at histologic evaluation. RESULTS: Enhancement ratios were significantly higher for gadoteridol than for gadopentetate dimeglumine in the physis, epiphyseal cartilage, and secondary ossification center (P < .05). After contrast agent administration, T1 values decreased sharply for both agents-but more so for gadoteridol. Additionally, there was less variability in T1 values across structures with this contrast agent. Gadoteridol resulted in greater visibility of the physis, while gadopentetate dimeglumine resulted in greater contrast between the physis and metaphysis (P < .05). CONCLUSION: The results suggest different roles for the two gadolinium-based contrast agents: The nonionic contrast medium is better suited for evaluating perfusion and anatomic definition in the immature skeleton, while the ionic contrast medium is better for evaluating cartilage fixed-charge density.

Automated perfusion-weighted MRI using localized arterial input functions.

PURPOSE: To investigate the utility of an automated perfusion-weighted MRI (PWI) method for estimating cerebral blood flow (CBF) based on localized arterial input functions (AIFs) as compared to the standard method of manual global AIF selection, which is prone to deconvolution errors due to the effects of delay and dispersion of the contrast bolus. MATERIALS AND METHODS: Analysis was performed on spin- and gradient-echo EPI images from 36 stroke patients. A local AIF algorithm created an AIF for every voxel in the brain by searching out voxels with the lowest delay and dispersion, and then interpolating and spatially smoothing them for continuity. A generalized linear model (GLM) for predicting tissue outcome, and MTT lesion volumes were used to quantify the performance of the localized AIF method in comparison with global methods using ipsilateral and contralateral AIFs. RESULTS: The algorithm found local AIFs in each case without error and generated a higher area under the receiver operating characteristic (ROC) curve compared to both global-AIF methods. Similarly, the local MTT lesion volumes had the least mean squared error (MSE). CONCLUSION: Automated CBF calculation using local AIFs is feasible and appears to produce more useful CBF maps.

T2 and T1rho MRI in articular cartilage systems.

T2 and T1rho have potential to nondestructively detect cartilage degeneration. However, reports in the literature regarding their diagnostic interpretation are conflicting. In this study, T2 and T1rho were measured at 8.5 T in several systems: 1) Molecular suspensions of collagen and GAG (pure concentration effects): T2 and T1rho demonstrated an exponential decrease with increasing [collagen] and [GAG], with [collagen] dominating. T2 varied from 90 to 35 ms and T1rho from 125 to 55 ms in the range of 15-20% [collagen], indicating that hydration may be a more important contributor to these parameters than previously appreciated. 2) Macromolecules in an unoriented matrix (young bovine cartilage): In collagen matrices (trypsinized cartilage) T2 and T1rho values were consistent with the expected [collagen], suggesting that the matrix per se does not dominate relaxation effects. Collagen/GAG matrices (native cartilage) had 13% lower T2 and 17% lower T1rho than collagen matrices, consistent with their higher macromolecular concentration. Complex matrix degradation (interleukin-1 treatment) showed lower T2 and unchanged T1rho relative to native tissue, consistent with competing effects of concentration and molecular-level changes. In addition, the heterogeneous GAG profile in these samples was not reflected in T2 or T1rho. 3) Macromolecules in an oriented matrix (mature human tissue): An oriented collagen matrix (GAG-depleted human cartilage) showed T2 and T(1rho) variation with depth consistent with 16-21% [collagen] and/or fibril orientation (magic angle effects) seen on polarized light microscopy, suggesting that both hydration and structure comprise important factors. In other human cartilage regions, T2 and T1rho abnormalities were observed unrelated to GAG or collagen orientation differences, demonstrating that hydration and/or molecular-level changes are important. Overall, these studies illustrate that T2 and T1rho are sensitive to biologically meaningful changes in cartilage. However, contrary to some previous reports, they are not specific to any one inherent tissue parameter.

T2 of articular cartilage in the presence of Gd-DTPA2-.

T(2) information and delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) are both used to characterize articular cartilage. They are currently obtained in separate studies because Gd-DTPA(2-) (which is needed for dGEMRIC) affects the inherent T(2) information. In this study, T(2) was simulated and then measured at 8.45 T in 20 sections from two human osteochondral samples equilibrated with and without Gd-DTPA(2-). Both the simulations and data demonstrated that Gd-DTPA(2-) provides a non-negligible mechanism for relaxation, especially with higher (1 mM) equilibrating Gd-DTPA(2-) concentrations, and in areas of tissue with high T(2) (due to weak inherent T(2) mechanisms) and high tissue Gd-DTPA(2-) (due to a low glycosaminoglycan concentration). Nonetheless, T(2)-weighted images of cartilage equilibrated in 1 mM Gd-DTPA(2-) showed similar T(2) contrast with and without Gd-DTPA(2-), demonstrating that the impact on T(2) was not great enough to affect identification of T(2) lesions. However, T(2) maps of the same samples showed loss of conspicuity of T(2) abnormalities. We back-calculated inherent T(2)'s (T(2,bc)) using a T(2)-relaxivity value from a 20% protein phantom (r(2) = 9.27 +/- 0.09 mM(-1)s(-1)) and the Gd-DTPA(2-) concentration calculated from T(1,Gd). The back-calculation restored the inherent T(2) conspicuity, and a correlation between T(2) and T(2,bc) of r = 0.934 (P < 0.0001) was found for 80 regions of interest (ROIs) in the sections. Back-calculation of T(2) is therefore a viable technique for obtaining T(2) maps at high equilibrating Gd-DTPA(2-) concentrations. With T(2)-weighted images and/or low equilibrating Gd-DTPA(2-) concentrations, it may be feasible to obtain both T(2) and dGEMRIC information in the presence of Gd-DTPA(2-) without such corrections. These conditions can be designed into ex vivo studies of cartilage. They appear to be applicable for clinical T(2) studies, since pilot clinical data at 1.5 T from three volunteers demonstrated that calculated T(2) maps are comparable before and after "double dose" Gd-DTPA(2-) (as utilized in clinical dGEMRIC studies). Therefore, it may be possible to perform a comprehensive clinical examination of dGEMRIC, T(2), and cartilage volume in one scanning session without T(2) data correction.