Cerebral Magnetic Resonance T2 High Intensities in End-Stage Renal Disease
Background and Purpose The pathophysiological mechanisms that cause cerebral MR T2 high intensities in end-stage renal disease (ESRD) are unclear. We evaluated the incidence and the risk factors of T2-weighted MR brain high intensities in patients with ESRD.
Methods We examined the degree of T2-weighted MR brain high intensities (high intensity score) and determined the variables that had an independent association with the occurrence of high intensities in 38 patients with ESRD before chronic dialysis treatment, 173 patients with essential hypertension, and 72 normotensive control subjects.
Results The whole brain high intensity score was significantly higher in patients with ESRD than in the control subjects, but there was no significant difference in high intensity score between the ESRD and the hypertensive groups. Age, hypertension, and smoking were significant independent predictors of high intensities in a multiple logistic regression model. The distribution pattern of high intensities in ESRD patients was very similar to that obtained from hypertensive patients; the high intensity score was highest in the corona radiata and was lowest in the cerebellum.
Conclusions T2 high intensities on MR images of ESRD may reflect subcortical small-vessel alterations induced by hypertension.
Patients undergoing dialysis have a high prevalence of cerebrovascular disease, which is one of the major causes of their death.1 Prevention of cerebrovascular diseases is required to improve the prognosis and the quality of life of patients with end-stage renal disease (ESRD). In a study of dialyzed patients, CT showed cerebral atrophy in approximately half of the cases, and a correlation was found between cortical atrophy and duration of disease.2 MRI is much more sensitive than CT for detecting abnormal signal intensity of the brain.3 Fazekas et al4 reported the presence of MR T2 high intensities (HIs) in 33% of patients receiving regular hemodialysis.
The underlying process for the renal disease and/or dialysis could cause brain damage and structural lesions. There is, however, limited information about the brain imaging findings in ESRD. It is unclear whether structural alterations of brain have already begun before the initiation of chronic dialysis therapy. There is no previously published article that specifically discusses the characteristic abnormal signal intensity on MR images of the brain in ESRD patients. We compared T2-weighted MR images with the clinical and laboratory parameters of ESRD patients and normotensive and hypertensive subjects without evidence of neurological and renal abnormalities. We evaluated the distribution and the magnitude of T2 HI regionally to define morphological changes of the brain.
Subjects and Methods
Between April 1995 and January 1997, 15 men and 23 women with ESRD (mean serum creatinine, 724 μmol/L; range, 344 to 1564 μmol/L) with a mean age of 61.8 years (range, 27 to 82 years) who had not been started on chronic dialysis therapy were studied. They were either hospitalized for initiation of dialysis therapy or expected to be on chronic dialysis in next few months. Diagnoses included diabetic nephropathy (n=14), chronic glomerulonephritis (n=11), polycystic kidney (n=3), glomerulosclerosis (n=4), lupus nephritis (n=1), and ESRD of unknown cause (n=5). Patients with previous focal neurological deficits were excluded from the study. A complete clinical history was recorded, and neurological examinations and standard laboratory tests were done when MR images were performed. Table 1⇓ shows clinical characteristics of ESRD patients in this study. A diagnosis of hypertension was made if blood pressure was more than 160/90 mm Hg or if diastolic pressure was more than 95 mm Hg, as measured with conventional mercury sphygmomanometer with the subjects seated. All of the patients with ESRD in this study were hypertensive (systolic blood pressure, 181±25 mm Hg; diastolic blood pressure, 94±15 mm Hg). They had been treated with antihypertensive agents (angiotensin-converting enzyme inhibitors, n=4; calcium channel blockers, n=14; diuretics, n=4; the combination of drugs, n=14) and/or diet therapy. Hypercholesterolemia, diabetes mellitus, and hyperuricemia were defined as a total cholesterol concentration of more than 6.2 mmol/L, a fasting blood glucose concentration of more than 7.8 mmol/L or HbA1c of more than 6.5%, and a uric acid concentration of more than 506 μmol/L, respectively. Subjects receiving the definite diagnosis of diabetes mellitus previously were also included in the diabetic category. A diagnosis of ischemic heart disease was made by cardiologists using 12-lead ECG, thallium-201 tomography, and/or coronary angiography. Three ESRD patients had previous history of myocardial infarction. A body mass index of more than 25 kg/m2 was defined as obese.
MR data, clinical history, and laboratory data from age-matched normal renal function groups were used for comparison: 72 normotensive subjects who had nonspecific neurological complaints of headaches (n=41), paresthesias (n=18), or dizziness (n=12) (mean age, 58.5 years; range, 21 to 83) and the 173 patients with essential hypertension (mean age, 63.9 years; range, 29 to 92). All subjects gave their informed consent.
MR images were obtained with a 1.5-T superconducting unit (Magnetom H15, Siemens). Multiple spin-echo sequences were done with a repetition time of 4500 milliseconds and an echo time of 90 milliseconds to produce T2-weighted images. Transaxial images of 8-mm-thick sections of the brain with a 2-mm gap were obtained. Two-dimensional Fourier transformation of images and a 256×256 data-acquisition matrix were used. The three-dimensional time-of-flight images of MR angiography were acquired with a repetition time of 38 milliseconds, echo time of 7 milliseconds, flip angle of 20°, 15-cm field of view, partition of 64, 128×120 acquisition matrix, and one signal average for a total imaging time of 5 minutes and 14 seconds.
The MR images were visually interpreted by one observer (Y.I.) blinded to the clinical information of the patients based on the method of Erkinjuntti et al.5 For MR localization of the HI site, measures of the distinct zone of high attenuation (viz, size [diameter], number, and appearance [focal or confluent]) were graded on a scale of 0 to 5 (HI score): 0=absent; l=<5 small (<5 mm) and/or <2 large (≥5 mm) focal lesions; 2=5 to 12 small focal and/or 2 to 4 large focal lesions; 3=>12 small focal and/or >4 large focal lesions; 4=predominantly confluent lesions. There was high interrater agreement (κ=0.64) for evaluation of T2 HI.6 Focal HIs on MR images were qualitatively defined in the transaxial images for 12 brain regions from six tomographic slices, as described previously (Fig 1⇓).6 The defined brain areas were delineated from the supraventricular slice (centrum semiovale), high ventricular slice (corona radiata, anterior and posterior periventricular white matters at the level of the body of the lateral ventricle), midventricular slice (anterior and posterior periventricular white matters at the level of the body of the lateral ventricle and the genu of corpus callosum), low ventricular slices (putamen, thalamus, and anterior and posterior white matters at the level of the basal ganglia), mid-pons, and cerebellar hemisphere. They include internal watershed areas, basal ganglia, subcortical white matters, and periventricular areas. For each ventricular level, each white matter was assigned to anterior and posterior regions adjoining the frontal horn or the occipital horn of the lateral ventricle, respectively. The total score was obtained by summing the scores of all regions in both right and left hemispheres for the six brain levels (maximum score, 96).
Relative risk for the presence of T2 HI was analyzed with the use of odds ratio and χ2 statistics. Multiple logistic regression analysis was used to determine the independent effects of the predictor variables for the presence of T2 HI. Regional or whole brain HI score differences between the three groups were assessed with the Kruskal-Wallis test and subsequent Mann-Whitney U test. Data were presented as mean±SD. A level of P<.05 was accepted as statistically significant.
In comparison with patients with no T2 HI on MRI (n=111), patients with any T2 HI on MRI (n=172) significantly more often had hypertension, ischemic heart disease, ESRD, or hyperuricemia, and they less often had obesity (Table 2⇓). In this comparison, diabetes mellitus, smoking, hypercholesterolemia, sex, and alcohol drinking were not significantly different. The logistic regression technique was used for multivariate analysis with 11 variables comprising hypertension, ischemic heart disease, ESRD, diabetes mellitus, hyperuricemia, smoking, hypercholesterolemia, male sex, alcohol drinking, obesity, and age (years). Multiple logistic regression showed that age, hypertension, and smoking were significantly and independently correlated with the presence of T2 HI (Table 3⇓), while ESRD showed no significant association with T2 HI. MR angiography showed no occlusive or stenotic changes of extracranial and intracranial vessels in all subjects.
There were no significant differences between the right and left hemispheres in regional HI score for any region tested. Regional HI score was calculated as the average of the right hemisphere and left hemisphere measures (Table 4⇓). In comparison with normotensive subjects, ESRD patients and hypertensive patients had significantly higher HI scores in the whole brain, but there was no statistically significant difference between the ESRD and the hypertensive groups. Regional HI scores were widely distributed and included all degrees of severity. Significant differences were found among the mean scores of 12 brain regions: χ2=24.5, df=11, P=.01 for the normotensive group; χ2=318.5, df=11, P<.00001 for the hypertensive group; and χ2=81.5, df=11, P<.00001 for the ESRD group. The highest mean score was found in the corona radiata, and the lowest mean score was in the cerebellum in all three groups. In both ESRD and hypertensive patients, mean scores were significantly higher in the centrum semiovale, corona radiata, putamen, high ventricular anterior white matter, high ventricular posterior white matter, midventricular anterior white matter, midventricular posterior white matter, and low ventricular posterior white matter than in the normotensive subjects. The mean scores were not significantly different among the three groups in the thalamus, pons, cerebellum, and low ventricular anterior white matter.
We found that ESRD patients before chronic hemodialytic therapy had a greater severity of T2 HI than normotensive subjects, but there was no significant difference between the hypertensive group and the ESRD group in the severity of HI. The results suggest that ESRD patients already have substantial brain MR changes even when they do not receive hemodialysis, and hypertension is highly associated with an increased severity of HI.
Silent T2 HI on MR imaging has been defined as an “asymptomatic cerebrovascular disease” in the Classification of Cerebrovascular Diseases III from the National Institute of Neurological Disorders and Stroke.7 There are no specific pathological studies of HI in ESRD. Previous studies that compared MR images and pathology in patients with brain HI have reported that the increased signal on T2-weighted images reflects increased water content, and a variety of pathological changes may explain the different brain HIs.8 9 10 11 12 13 Periventricular HI has been found to represent subependymal gliosis and demyelination associated with discontinuity of the ependymal lining. Confluent areas of HI involve varying degrees of fiber loss, small cavitations, arteriolosclerosis, rarefaction of the white matter, and lacunar infarcts.
The strong association between whole brain HI score and age was consistent with prior reports in patients with normal renal function,14 15 16 17 18 suggesting that there is a progression of structural alterations in the aging brain of ESRD patients. In addition to age, the multiple logistic regression indicates that hypertension and smoking are independent predictors for the presence of T2 HI. Hypertension was associated with a high frequency of HI in both the ESRD and hypertensive groups. Although a χ2 test (Table 2⇑) showed a significant positive association between ESRD and the presence of HI, multiple logistic regression analysis (Table 3⇑) did not, probably because of the high frequency (100%) of hypertensive patients in the ESRD patients. There was no significant difference between prevalence of T2 HI in ESRD patients and that in hypertensive group (70% versus 68%). We consider that hypertension and ESRD might be interrelated risk factors. Moreover, it is noteworthy that the HI score of whole brain in the ESRD group was nearly equal to that in the age-matched hypertensive group, as shown in Table 4⇑ (12.3 versus 11.1). In other words, both age and hypertension are thought to be primary pathogenic factors in HI lesions in ESRD.
The distribution of HI scores varied widely from region to region of the ESRD brain. When the regional HI distribution in the ESRD group is analyzed, the highest value was obtained in the corona radiata, followed by those in the midventricular posterior white matter and low ventricular posterior white matter. The lowest values of HI score were found in the cerebellum, followed by that in the pons. The distribution patterns of HI were almost the same in the ESRD group and hypertensive group, suggesting that the pathophysiology of the HI lesions in the two groups is similar. The white matter is more vulnerable to ischemia than the cortical gray matter because each of the perforating arteries that supply the periventricular white matter is an end artery with minimal overlap and anastomosis to the territories of the different groups.19 20 Periventricular white matter of the corona radiata corresponds to the border zone between the deep and superficial territories of the middle cerebral artery.21 Thus, progressive arteriosclerosis of the lenticulostriate arteries is presumably implicated in the processes of the striking increase of the HIs in the corona radiata in the ESRD and hypertensive patients.
In conclusion, T2 HIs of ESRD likely result from small-vessel injury of the brain. They correlate with age, hypertension, and smoking. The presence of brain HI in ESRD may not be considered benign because there is an association between impaired cognitive or motor function and HI.22 Further studies are needed to determine whether the HI changes in ESRD are themselves predictors of future occurrence of cerebrovascular events.
This study was supported by a grant from the Japanese Ministry of Health and Welfare.
Reprint requests to Akira Wada, MD, Department of Internal Medicine, Osaka National Hospital, Hoenzaka, 2–1-14, Chuo-ku, Osaka 540, Japan.
- Received July 28, 1997.
- Revision received September 9, 1997.
- Accepted September 15, 1997.
- Copyright © 1997 by American Heart Association
Brunner FP, Wing AJ, Dykes SR, Brynger HOA, Fassbinder W, Selwood NH. International review of replacement therapy: strategies and results. In: Maher JF, ed. Replacement of Renal Function by Dialysis. 3rd ed. Norwell, Mass: Kluwer Academic Publishers; 1989:697–719.
Awad IA, Spetzler RF, Hodak JA, Awad CA, Garey R. Incidental subcortical lesions identified on magnetic resonance imaging in the elderly, I: correlation with age and cerebrovascular risk factors. Stroke. 1986;17:1084–1089.
Isaka Y, Nagano K, Narita M, Ashida K, Imaizumi M. High signal intensity on T2-weighted magnetic resonance imaging and cerebral hemodynamic reserve in carotid occlusive disease. Stroke. 1997;28:354–357.
Special report from the National Institute of Neurological Disorders and Stroke: classification of cerebrovascular disease III. Stroke. 1990;21:637–676.
Awad IA, Johnson PC, Spetzler RF, Hodak JA. Incidental subcortical lesions identified on magnetic resonance imaging in the elderly, II. postmortem pathological correlations. Stroke. 1986;17:1090–1097.
Kirkpatrick JB, Hayman LA. White-matter lesions in MR imaging of clinically healthy brains of elderly subjects: possible pathologic basis. Radiology. 1987;509–511:.
Braffman BH, Zimmerman RA, Trojanowski JQ, Gonatas NK, Hickey WF, Schlaepfer WW. Brain MR: pathologic correlation with gross and histopathology, 1: lacunar infarction and Virchow-Robin spaces. AJNR Am J Neuroradiol. 1988;9:621–628.
Van Swieten JC, Van Den Hout JHW, Van Ketel BA, Hijdra A, Wokke JHJ, Gijn JV. Periventricular lesions in the white matter on magnetic resonance imaging in the elderly: a morphometric correlation with arteriosclerosis and dilated periventricular spaces. Brain. 1991;114:761–764.
Fazekas F, Kleinert R, Offenbacher H, Schmidt R, Kleinert G, Payer F, Radner H, Lechner H. Pathologic correlates of incidental MRI white matter signal hyperintensities. Neurology. 1993;43:1683–1689.
Fazekas F, Niederkorn K, Schmidt R, Offenbacher H, Horner S, Bertha G, Lechner H. White matter signal abnormalities in normal individuals: correlation with carotid ultrasonography, cerebral blood flow measurements, and cerebrovascular risk factors. Stroke. 1988;19:1285–1288.
Schmidt R, Fazekas F, Kleinert G, Offenbacher H, Gindl K, Payer F, Freidl W, Niederkorn K, Lechner H. Magnetic resonance imaging signal hyperintensities in the deep and subcortical white matter: a comparative study between stroke patients and normal volunteers. Arch Neurol. 1992;49:825–827.
Manolio TA, Kronmal RA, Burke GL, Poirier V, O’Leary DH, Gardin JM, Fried LP, Steinberg EP, Bryan RN. Magnetic resonance abnormalities and cardiovascular disease in older adults: the Cardiovascular Health Study. Stroke. 1994;25:318–327.
Moody DM, Bell MA, Challa VR. Features of the cerebral vascular pattern that predict vulnerability to perfusion or oxygenation deficiency: an anatomic study. AJNR Am J Neuroradiol. 1990;11:431–439.
Bogousslavsky J, Regli F. Centrum ovale infarcts: Subcortical infarction in the superficial territory of the middle cerebral artery. Neurology. 1992;42:1992–1998.
Salloway S, Malloy P, Kohn R, Gillard E, Duffy R, Rogg J, Tung G, Richardson E, Thomas C, Westlake R. MRI and neuropsychological differences in early- and late-life-onset geriatric depression. Neurology. 1996;46:1567–1574.