Mini-Review Systematic Review of Structural and

Mini-Review

Systematic Review of Structural and Functional Neuroimaging Findings in Children and Adults with CKD Divya G. Moodalbail,* Kathryn A. Reiser,* John A. Detre,†‡ Robert T. Schultz,* John D. Herrington,* Christos Davatzikos,‡ Jimit J. Doshi,‡ Guray Erus,‡ Hua-Shan Liu,†‡ Jerilynn Radcliffe,* Susan L. Furth,*§ and Stephen R. Hooper|

Summary CKD has been linked with cognitive deficits and affective disorders in multiple studies. Analysis of structural and functional neuroimaging in adults and children with kidney disease may provide additional important insights into the pathobiology of this relationship. This paper comprehensively reviews neuroimaging studies in both children and adults. Major databases (PsychLit, MEDLINE, WorldCat, ArticleFirst, PubMed, Ovid MEDLINE) were searched using consistent search terms, and studies published between 1975 and 2012 were included if their samples focused on CKD as the primary disease process. Exclusion criteria included case reports, chapters, and review articles. This systematic process yielded 43 studies for inclusion (30 in adults, 13 in children). Findings from this review identified several clear trends: (1) presence of cerebral atrophy and cerebral density changes in patients with CKD; (2) cerebral vascular disease, including deep white matter hyperintensities, white matter lesions, cerebral microbleeds, silent cerebral infarction, and cortical infarction, in patients with CKD; and (3) similarities in regional cerebral blood flow between patients with CKD and those with affective disorders. These findings document the importance of neuroimaging procedures in understanding the effect of CKD on brain structure, function, and associated behaviors. Results provide a developmental linkage between childhood and adulthood, with respect to the effect of CKD on brain functioning across the lifespan, with strong implications for a cerebrovascular mechanism contributing to this developmental linkage. Use of neuroimaging methods to corroborate manifest neuropsychological deficits or perhaps to indicate preventive actions may prove useful to individuals with CKD. Clin J Am Soc Nephrol 8: 1429–1448, 2013. doi: 10.2215/CJN.11601112

Introduction Kidney function may affect brain function on many levels, ranging from developmental alterations and vascular injury to disorders of metabolism. There is increasing interest in understanding the neurologic basis of cognitive and affective dysfunction in children and adults with kidney disease that is evident in neurobehavioral findings across the lifespan (e.g., increased rates of neurocognitive impairments in children; increased rates of dementia in adults). Neuroimaging methods provide a noninvasive means of examining brain structure and function (1) and represent an important approach for further elucidating the kidney-brain connection in individuals with CKD. Sixteen years ago, Butler et al. (2) touted the benefits of using magnetic resonance angiography as a screening tool for cerebral aneurysms and subarachnoid hemorrhage in patients with autosomal dominant polycystic kidney disease (ADPKD). Similarly, Burn and Bates (3) reviewed neurologic presentations in patients with a wide range of kidney diseases (e.g., Von Hippel-Lindau disease) from genetic predisposition to tumors to direct and indirect effects of kidney disease on brain functioning. www.cjasn.org Vol 8 August, 2013

To date, several comprehensive reviews have examined the neuroimaging literature in individuals with CKD. Aǧildere et al. (4) provided one of the first reviews of the neurologic complications of ESRD in patients undergoing hemodialysis (HD) assessed by use of magnetic resonance imaging (MRI). These investigators highlighted the importance of MRI, including T2-weighted, fluid-attenuated inversion recovery, and proton density images, as a tool for clinical investigation of ESRD-associated neurologic symptoms. They reported that patients with ESRD, particularly those receiving HD, have high rates of neurologic complications compared with healthy controls. They surveyed the use of MRI in combination with new and conventional imaging methods for diagnosing white matter changes, cerebral atrophy, osmotic demyelination syndrome, dialysis encephalopathy, hypertensive encephalopathy, intracranial hemorrhage, cerebral ischemia and infarction, infection, and sinus thrombosis. Brouns and De Deyn (5) also reviewed the causes and presentations of neurologic complications in uremic patients with ESRD. They noted that dialysis and renal transplantation themselves may precipitate

*Department of Pediatrics, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; † Department of Neurology, ‡ Department of Radiology, and § Department of Epidemiology, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania; and | Department of Psychiatry and Carolina Institute for Developmental Disabilities, University of North Carolina School of Medicine, Chapel Hill, North Carolina Correspondence: Dr. Divya G. Moodalbail, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Boulevard, Philadelphia, PA 19104. Email: divya. moodalbail@gmail. com

Copyright © 2013 by the American Society of Nephrology

1429

1430

Clinical Journal of the American Society of Nephrology

neurologic symptoms and that treatment and maintenance for patients with ESRD should address both renal and neurologic components. More recently, Lakadamyali and Ergün (6) reviewed brain MRI findings for diagnosis of acute neurologic complications in patients with ESRD undergoing HD. They noted that symptoms may be caused directly by CKD itself, HD, or a combination of the two. Cerebral ischemia and infarction, intracerebral hemorrhage, posterior reversible encephalopathy syndrome, osmotic demyelination syndrome, cerebral infection, gadolinium accumulation, sinus vein thrombosis, and dialysis disequilibrium syndrome are all reported as being detected via MRI. Vogels et al. (7) reviewed the association between CKD and brain lesions as noted on MRI and computed tomography (CT). They reported association of CKD with brain atrophy, silent cerebral infarction, and white matter lesions (WMLs). Taken together, these earlier reviews describe the neurologic complications that can stem from CKD. Importantly, however, none of these reviews provide findings pertaining to children. In addition, the earlier reviews focused on specific neurologic complications from CKD and did not provide a broad overview of the neuroimaging literature related to CKD. This paper provides an updated review of neuroimaging in kidney disease, focusing on quantitative findings from both structural and functional neuroimaging studies in children and adults. The primary objective is to provide a contemporary summary of the available studies. Analysis of structural and functional neuroimaging in adults and children with kidney disease will provide additional important insights into the pathobiology of CKD.

Methods A systematic review of major search databases was conducted to compile a comprehensive list of studies that used neuroimaging techniques in individuals with CKD and were published between 1975 and 2012. Three reviewers systematically reviewed databases, including PsychLit, Ovid MEDLINE, MEDLINE, WorldCat, PubMed, and ArticleFirst. Designated search terms included MRI AND chronic kidney disease, computed tomography AND hemofiltration, polycystic kidney AND brain AND magnetic resonance imaging, dialysis AND single-photon emission computed tomography (SPECT), dialysis AND MRI, dialysis AND regional cerebral blood flow (rCBF), rCBF AND CKD, CKD AND CT, dialysis AND transcranial Doppler, dialysis AND cerebral magnetic resonance imaging, neurologic impairment in CKD, neuroimaging AND CKD, and diffusion tensor imaging (DTI) AND CKD. Studies were included in the review if (1) the sample focused on CKD (defined by an estimated GFR [eGFR] ,60 ml/min per 1.73 m2 for $3 months) as the primary disease process and (2) the study produced quantitative statistics. Exclusion criteria included case reports, book chapters, and previous review articles. Use of the search terms in this systematic process produced 1462 possible articles, and an additional 9 records were identified through other sources (e.g., citations included in the ascertained articles). Removal of the duplicate articles resulted in a pool of 1264 potential articles for inclusion. Application of the inclusion and exclusion

criteria produced 43 articles for inclusion in the review. These studies comprised 30 studies in adults (20 structural, 10 functional) and 13 studies in children (13 structural). The selection and identification process for candidate studies can be seen in Figure 1.

Results Adult Neuroimaging Studies Thirty studies included adults with CKD in their neuroimaging protocols: 20 structural and 10 functional. These studies span a time frame from 1987 (8) to 2011 (9–11). The study designs included retrospective, cross-sectional, casecontrol, and prospective approaches to data collection. Sample sizes ranged widely, and the composition of the CKD samples reflected significant heterogeneity across studies. Other studies examined AKI (e.g., Ronco et al. 12); however, our review focuses on CKD. For the structural neuroimaging studies, the primary technique used was MRI, with several studies using CT. For the functional neuroimaging studies, the primary techniques used were SPECT, transcranial Doppler ultrasonography, and rCBF using 133Xe inhalation. Findings in Adult Structural Neuroimaging Studies In adults, more severe kidney disease has been associated with abnormal structural findings. One study has documented the presence of intracranial arachnoid cysts in ADPKD (13), but more studies examined MRI findings in patients undergoing various renal replacement therapies (RRTs) and have reported the presence of cerebral atrophy and decreased cerebral density in both white and gray matter (9,14–21). Walters et al. (18) showed that five HD patients had postdialysis increases in cerebral volume compared with five controls. Savazzi et al. (20) found high correlations between atrophy on CT and MRI and predialytic BP. Yoshimitsu et al. (19) also studied MRI in HD patients and reported significantly more lacunae and more advanced periventricular hyperintensity than matched controls. In one of the largest studies conducted to date (187 kidney transplant and 29 liver transplant cases), Aǧildere et al. (16) found that roughly one third of adult renal transplant recipients had neuroradiologic findings on MRI, including nearly 20% of patients with hemispheric sulcal and ventricular dilatation. Similar findings have been uncovered in patients not receiving RRT. In neuroimaging studies conducted to date, with sample sizes ranging from 52 to 625 adult cases with CKD, high rates of structural abnormalities were detected with MRI, including WMLs (22). In a study by Wada et al. (23), elderly patients with CKD (defined by low eGFR [,60 ml/min per 1.73 m2] or a urinary albumin-to-creatinine ratio .30 mg/g) evidenced more lacunar infarcts and higher grades of WMLs. After adjustment for hypertension and diabetes, CKD continued to remain an independent risk factor for cerebrovascular disease–related lesions. Several other investigators have reported similar findings (24– 27). Using DTI, Kim et al. (9) showed subtle WMLs in a small sample of patients receiving peritoneal dialysis (PD). Cerebral small vessel disease has also been studied in the adult CKD population. Kobayashi et al. (28) found that decreased kidney function was a risk factor for silent lacunar infarcts. The prevalence of silent cerebral infarction

Clin J Am Soc Nephrol 8: 1429–1448, August, 2013

Neuroimaging in CKD, Moodalbail et al.

1431

Figure 1. | Systematic review of neuroimaging data in CKD. *Studies where CKD was not the primary disease process of interest, but other associated disease processes (e.g., stroke) were screened for in large populations of patients with CKD. Reproduced with permission from Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 6: e1000097, 2009.

also has been reported to be significantly higher in HD patients than in controls (17). Kobayashi et al. (29) also found an independent association between silent cerebral infarction and decreased kidney function, and silent cerebral infarction independently predicted outcomes, including stroke and dementia. Other investigators have asserted that WMLs reflect possible ischemic brain damage and generalized vascular damage, and both are independently associated with vascular nephropathy in patients with CKD (22). In patients with CKD, high rates of cerebral microbleeds have also been documented, with rates ranging from about 26% (30) to 61% (29). These microbleeds tended to be more prevalent in male patients, those with hypertension, those with advanced age, and those with worsening kidney function as defined by eGFR, with no relation to the use of antiplatelet or antithrombotic therapy (15,30,31). Table 1 summarizes the adult structural neuroimaging studies in CKD. Findings in Adult Functional Neuroimaging Studies Functional neuroimaging studies use methods to detect structural brain abnormalities and localized neural activity

in reaction to executing certain specific sensory, motor, and cognitive tasks. These studies are sensitive to changes in rCBF and blood flow velocity in targeted brain regions. Seven of the 10 studies included in this review focused on patients receiving HD or PD. In one of the oldest imaging studies conducted with adults, which used 133Xe inhalation, Gottlieb et al. (8) found that predialysis rCBF values did not differ from those in age-matched controls; however, after HD, there was a mild but significant reduction in rCBF. This reduction was not associated with cognitive impairment or neurologic symptoms. Using transcranial Doppler ultrasonography, several studies have documented similar findings. Prohovnik et al. (32) showed that 19 HD patients had reduced rCBF between sessions. Other investigators have reported similar findings (33,34), with blood flow reductions noted in the middle cerebral artery and the basilar artery after HD (35). However, some studies report contrary findings, suggesting that cerebral blood flow does not change significantly during or after HD (36). Many functional neuroimaging studies also have used SPECT in their examination of blood flow velocities

Cross-sectional

Prospective cohort

Cross-sectional

Cross-sectional

Cross-sectional

n=4 patients with ESRD on PD $ 3 yr; 3 men and 1 woman; mean age, 39.75 yr n=6 controls; 4 men and 2 women; mean age, 44.17 yr

n=142 patients with CKD; 96 mean and 46 women; mean age, 64.7 yr

n=162 patients with CKD, stages 1–5; 85 with diabetes, 77 without diabetes; 92 men and 70 women; mean age, 64.9 yr n=24 controls; 14 men and 10 women; mean age, 59.7 yr

n=610 Japanese adults with CKD; 302 men and 308 women; mean age, 56.4 yr

n=335 patients with CKD n=40 patients with EHT and no CKD; not age-matched Overall: 226 men and 149 women; mean age, 63.5 yr

Kim et al. (2011)

Kobayashi et al. (2010)

Shima et al. (2010)

Yakushiji et al. (2010)


Cross-sectional

Study Design

n=179 HD patients; mean age, 58.2 yr (range, 23–86 yr) n=58 healthy controls; mean age, 55.9 yr (range, 41–77)

Sample Description

Naganuma et al. (2012)

Authors

Table 1. Adult structural neuroimaging studies in CKD (n=20)

MRI (T1-, T2-, and proton density–weighted images)

MRI (T1-, T2-, and FLAIR weighted images)

MRI (T1-, T2-, FLAIR, and T2*-weighted images)


DTI

MRI (T1-, T2-weighted and FLAIR)

Imaging Technique

The prevalence of WMHs was significantly higher in HD patients than healthy persons. In the HD patients, multiple logistic regression analysis showed that independent and significant factors associated with the presence of PVH were age, female sex, and systolic BP and those associated with the presence of DSWMH were age, female sex, systolic BP, and body mass index. DTI findings were investigated in patients with ESRD who showed no specific lesions on conventional brain MRI. DTI tractography showed that all 4 patients had .1 lesion. Patients with ESRD showed abnormalities on DTI that were associated with cognition; however, they did not show significant cognitive abnormalities. At baseline, 87 patients had SBI. During 2-yr follow-up, 43 patients (30.3%) developed the following primary outcomes: doubling of SCr level, dialysis therapy, and death from cardiovascular causes. SBI was an independent predictor of outcomes; eGFR decreased more in patients with SBI than in those without SBI. CMBs were found in 35 patients with CKD (25.6%) but not in controls. CMBs were more prevalent in men, those with higher BP, those with advanced age, and those with poor kidney function. There was an association between the prevalence of CMBs and advancing CKD stage (P,0.01). eGFR was associated with the prevalence of CMBs independent of age, sex, and hypertension; however, no relationship was seen between CMBs and diabetes mellitus and dyslipidemia. Cerebral atrophy was found in 25 (4.1%) cases. Age, male sex, hypertension, kidney function, WMH, and lacunae were associated with cerebral atrophy. eGFR , 60 ml/min per 1.73 m2 was associated with cerebral atrophy. Treatment of CKD may control age-related degenerative processes of the brain. SBIs were seen 56.5% of patients. Hypertensive nephrosclerosis and advancing eGFR state were associated with SBI. eGFR was related to SBI, in addition to age and BP (P=0.025). Patients with CKD should undergo active detection of SBI and more intensive preventive management, especially for hypertension.

Main Findings

1432 Clinical Journal of the American Society of Nephrology

Cross-sectional

Cross-sectional

Cross-sectional

Cross-sectional

Cross-sectional

n=484 patients in the Rotterdam Study; mean eGFR, 54.8 ml/min per 1.73 m2; 239 men and 245 women; mean age, 73.4 yr

n=625 community-based Japanese elderly; 359 men and 266 women; mean age, 62.7 yr

n=615 patients with CKD from Northern Manhattan Study; 246 men and 369 women; mean age, 70 yr

n=57 PD patients; 34 men and 23 women; mean age, 48.4 yr n=57 matched hypertensive controls with normal renal function; 34 men and 23 women; mean age, 51.9 yr

n=80 HD patients; 34 men and 46 women; mean age, 62.9 yr

Ikram et al. (2008)

Wada et al. (2008)

Khatri et al. (2007)

Kim et al. (2007)

Watanabe (2007)

Study Design

Sample Description

Authors

Table 1. (Continued)

MRI (T2-weighted images)

MRI (T1-, T2-, and FLAIR-weighted images)

MRI (FLAIR images)


MRI (T1-, T2and proton density–weighted images)

Imaging Technique Volumes of WML and presence of lacunar infarcts reflected cerebral small vessel disease. Persons with lower GFR had less deep white matter volume and WML. GFR was not associated with gray matter volume or lobar white matter volume. Impaired kidney function was associated with cerebral small vessel disease. Patients with lower eGFR showed more lacunar infarcts and higher grades of WML. The mean grades of WML or the mean numbers of lacunar infarction in the patients with albuminuria were greater than those in patients without albuminuria. After exclusion of patients with stage 2 hypertension or diabetes, CKD remained an independent risk for CVD-related lesions. Creatinine clearance 15–60 ml/min was associated with increased log-white matter hyperintensity volume as was eGFR 15–60 ml/min. Findings highlight the importance of CKD as a possible determinant of cerebrovascular disease and/or as a marker of microangiopathy. The prevalence of leukoaraiosis was significantly greater in patients on PD than controls (P,0.001). Higher T2 signal intensities were seen in the PD patients than in controls, particularly in the anterior circulation of the brain with relative sparing of the posterior fossa. ESRD, older age, and poor control of BP were significant predictors of leukoaraiosis. CMBs were found in 28 patients (35%). The duration of HD did not significantly affect the appearance of the CMBs. Old intracerebral hemorrhages were seen in 7 patients, 5 of whom (71%) had CMBs. The frequency of old intracerebral hemorrhages was significantly higher in the patients with CMBs than in those without CMBs (P=0.48), and the patients with old intracerebral hemorrhages had significantly more CMBs than the patients without them (P=0.0065). The high ratio of patients with CMBs was considered to have been caused not by maintenance HD but by other factors, such as hypertension.

Main Findings

Clin J Am Soc Nephrol 8: 1429–1448, August, 2013 Neuroimaging in CKD, Moodalbail et al. 1433

Retrospective

Cross-sectional

n=187 kidney transplant cases; of cases with neuroradiologic findings; 36 men and 20 women; mean age, 39.3 yr n=29 liver transplant cases; 22 men and 7 women; mean age, 29.4 yr

n=52 patients with CKD stages 3 and 4 without diabetes; SCr, 2.03–7.91 mg/dl; 38 men and 14 women; mean age, 49 yr n=32 controls; 21 men and 11 women; mean age, 49 yr

Aǧildere et al. (2006)

Martinez-Vea et al. (2006)

Study Design

Sample Description

Authors



MRI (T-2 weighted and proton density– or FLAIR- and T1-weighted sequences)

Imaging Technique

Neuroradiologic findings were classified into 3 groups: Group 1 was related to transplantation. Overall, 30% of kidney transplant recipients had neuroradiologic findings: 3 patients (1.6%) had posterior reversible encephalopathy syndrome; 2 patients (1.1%) had tuberculosis granulomas; 1 patient (0.5%) had osmotic demyelination syndrome; 1 patient (0.5%) had a Nocardia abscess; and 1 (0.5%) had focal cerebritis. Group 2 was related to chronic parenchymal disease. Including kidney and liver transplant recipients, 38 patients (20.3%) had brain atrophy; 37 (20%), white matter changes; 3 (1.6%), sinus thrombosis; 8 (4.3%), lacunar infarct; 1 (0.5%), renal osteodystrophy in the cranial bones; and 4 (2.2%), intracranial hemorrhage secondary to ESRD. Group 3 was related to neither transplant nor chronic parenchymal disease: 3 patients (1.6%) had intracranial lipomas; 1 (0.5%), mesial temporal sclerosis; and 1 (0.5%), an anterior white matter aneurysm in renal transplant patients. WML were more prevalent in the CKD group than in controls (33% versus 6%; P=0.008). WML were more common in patients who were older; had renal disease caused by cardiovascular disease or vascular nephropathy; and had higher systolic BP, greater pulse pressure, larger left ventricular mass index, greater levels of C-reactive protein, and higher rates of antihypertensive drug prescriptions. Stage and duration of CKD were not related to the presence of WML. Only vascular nephropathy (P=0.03) predicted an increased risk for WML, suggesting that WML reflect ischemic brain damage caused by generalized vascular damage.

Main Findings


Cross-sectional

Cross-sectional

Cross-sectional

Cross-sectional

n=51 patients with CKD; mean SCr, 2.01 mg/dl; 29 men and 22 women; mean age, 52.7 yr n=80 control patients with EHT but no CKD; 47 men and 33 women; mean age, 58.8 yr

n=123 HD patients; 80 men and 43 women; mean age, 55.6 yr n=52 healthy controls; 29 men and 23 women; mean age, 51.7 yr

n=5 HD patients; 3 men and 2 women; mean age, 44.2 yr n=5 healthy controls; 2 men and 3 women; mean age, 67.4 yr

n=55 HD patients; 35 men and 20 women; mean age, 52 yr n=35 controls; 18 men and 17 women; mean age, 42 yr


Nakatani et al. (2003)

Walters et al. (2001)

Yoshimitsu et al. (2000)

Study Design

Sample Description

Authors


MRI

MRI (T1-weighted images)


MRI

Imaging Technique Lacunae prevalence was 25% in patients with a Ccr . 40 ml/min per 1.73 m2, 85% in patients with a Ccr , 40 ml/min per 1.73 m2, and 29% in patients with essential hypertension with normal renal function. Patients with lacunae had significantly lower hematocrits associated with increased fibrinogen and lipoprotein levels compared with those without lacunae. Plasma total homocysteine and insulin levels at 2 hr after a 75-g glucose tolerance test correlated with lacunae. Ischemic heart changes on echocardiography and thickened carotid intima-media thickness were more frequent in patients with lacunae. The best contributing factor for lacunar infarcts was decline in Ccr. Decreased renal function is a risk factor for silent lacunar infarcts. The prevalence of SCI was significantly higher in HD patients than in healthy controls (P,0.0001). For both groups, independent risk factors for SCI were chronic renal failure, hypertension, smoking, and age. In the HD group, age and smoking were independent risk factors of SCI (P,0.0001), whereas HD duration and hypertension were not. The findings indicate that chronic renal failure maintained by HD increases the prevalence of SCI and that age and smoking are related to SCI in HD patients. Potential cerebral volume change caused by HD was quantified by MRI immediately before and after HD. Patients had an increase in cerebral volume after HD, which averaged 32.8 ml (SEM, 7.4 ml). Controls averaged 1.4 ml (SEM, 0.6 ml). No patient had significant neurologic symptoms. HD patients showed more lacunae and more advanced PVH than controls. The VBR, the number of lacunae, and the severity of PVH tended to increase with age in HD. VBRs at all age groups were significantly higher in HD recipients than in controls (7.0% versus 3.7% at the fourth decade of life; 8.4% versus 5.9% at the fifth decade; 9.6% versus 5.4% at the sixth decade; and 11.6% versus 6.3% at the seventh decade). Both the number of lacunae and the severity of PVH were significantly correlated to VBR in HD recipients.

Main Findings


Cross-sectional

Retrospective

n=30 patients with CHD; 21 men and 9 women; mean age, 58 yr n=30 controls matched for age, sex, and major cerebrovascular risk factors; 23 men and 7 women; mean age, 60 yr

n=247 patients with ADPKD; 96 men and 151 women; mean age, 44 yr n=247 controls matched for age and sex

Fazekas et al. (1995)

Schievink et al. (1995)

MRI (180 cases) or CT (67 cases)

MRI (T1- and T2-weighted and axial mixed-intensity images)

CT (n=25) and MRI (n=17)

Imaging Technique

No significant correlation was found between the degree of atrophy and the uremia-altered laboratory measures of creatinine, hematocrit, cholesterol, triglyceride, albumin, PTH, calcium, and inorganic phosphate. No relationship between degree of atrophy and number of months the patients had been on HD or time that passed between the finding of Ccr , 30 ml/min per 1.73 m2 and the start of HD. Large correlations were found between CA and predialytic BP and between CA and the duration of hypertension. Hypertension seems to be related to cerebral parenchymal damage and should be corrected. The brains of patients with CHD showed significantly more atrophy on visual rating and semiquantitative morphometric measures. Multiple lacunes or confluent white matter hyperintensities predominated in 33% of patients. Marked cognitive impairment was associated with more extensive enlargement of the third ventricle (P,0.04) and the temporal horns (P,0.02) but not with the presence of cerebral ischemic lesions or any difference in laboratory data. These results call attention to a very high rate of cerebral damage in individuals undergoing CHD and suggest brain degeneration of toxic-metabolic cause that is associated with severe cognitive impairment. Unexpectedly high number of patients with ADPKD had intracranial arachnoid cysts. The cysts were found in 20 patients (8.1%) versus 2 controls (0.8%) (P,0.0001). Multiple intracranial arachnoid cysts were found in 2 patients. Pineal cysts were found in 2 patients (0.8%) and choroid plexus cysts were found in 3 patients (1.2%), but this was not different from controls. None of the intracranial cysts were symptomatic or treated surgically.

Main Findings

HD, hemodialysis: MRI, magnetic resonance imaging; FLAIR, fluid-attenuated inversion recovery; WMH, white matter hyperintensities; PVH, periventricular hyperintensity; DSWMH, deep and subcortical white matter hyperintensity; PD, peritoneal dialysis; DTI, diffusion-tensor imaging; SBI, silent brain infarct; eGFR, estimated GFR; CMB, cerebral microbleeds; EHT, essential hypertension; WML, white matter lesions; HD, hemodialysis; SCr, serum creatinine; Ccr, creatinine clearance; SCI, silent cerebral infarct; VBR, ventricular-brain ratio; CT, computed tomography; PTH, parathyroid hormone; CA, cerebral atrophy; CHD, coronary heart disease; ADPKD, autosomal dominant polycystic kidney disease.

Retrospective

n=25 regular HD treatment patients; 19 men and 6 women; mean age, 50.3 yr

Savazzi et al. (1999)

Study Design

Sample Description

Authors



n=14 patients with stage 5 CKD, eGFR , 15 ml/min per 1.73 m2 or permanent renal replacement therapy; 10 men and 4 women; mean age, 54 yr

n=23 nondialyzed patients with CKD stages 4 and 5; 15 men and 8 women; mean age, 48.4 yr n=15 dialyzed patients; 10 men and 5 women; mean age, 52.5 yr n=23 matched healthy controls; 9 men and 14 women; mean age, 52.5 yr

Tryc et al. (2011)

Sample Description

Nam et al. (2011)

Authors

Table 2. Adult functional neuroimaging studies in CKD (n=10)

Cross-sectional

Prospective

Study Design

1

Tc ECD SPECT

H-MRS

99m

Imaging Technique HDRS and SPECT were administered before starting dialysis and 6 mo after dialysis initiation. rCBF was compared between patients whose HDRS score decreased by .25% and those whose HDRS scores decreased by ,25% or those with an increase in HDRS score. A decrease of .25% in HDRS was correlated with higher perfusion in the left middle temporal gyrus (P=0.05) and higher rCBF in the right parahippocampal gyrus (P=0.04). Less reduction in HDRS was correlated with decreased rCBF in the left superior frontal gyrus (P=0.03) and right orbitofrontal cortex (P=0.05). MRS alterations were predominantly found in the white matter. Concentrations of creatinine-containing compounds were decreased in dialyzed and nondialyzed patients. Choline concentration (and combined N-acetylaspartate and N-acetylaspartylglutamate concentration) were reduced in dialyzed patients. Disturbance in memory/learning and attention were observed in both patient groups, but attention deficits were more severe in dialyzed patients. Patients with CKD without clinical signs of uremic encephalopathy showed metabolic disturbances in distinct brain regions and cognitive impairments. HD was accompanied by more severe cognitive dysfunction and metabolic alterations than CKD alone. A negative effect of HD on cognitive function must be considered.

Main Findings


n=27 patients with CKD, predialytic stage 4–5; mean GFR, 9.1 ml/min per 1.73 m2; 16 men and 11 women; mean age, 51.9 yr

n=19 HD patients; all men; mean age, 63 yr n=5 PD patients; all men; mean age, 63 yr n=14 controls; 11 men and 3 women; mean age, 61 yr

n=12 HD patients; 6 men and 6 women; mean age, 44 yr

Prohovnik et al. (2007)

Skinner et al. (2005)

Sample Description

Kim et al. (2008)

Authors


Case-control

Prospective

Prospective

Study Design

Transcranial Doppler ultrasonography

MP-RAGE, ASL, and T1-MRI sequences plus ICA Doppler ultrasonography and cerebral oximetry

Tc-99 ECD SPECT

Imaging Technique

rCBF and depressive mood were negatively correlated in the right insula, posterior cingulated gyrus, and left superior temporal gyrus and positively correlated in the left fusiform gyrus. rCBF and insomnia were negatively correlated in the right middle frontal gyrus, bilateral cingulated gyri, right insula, right putamen, and right inferior parietal lobule and positively correlated in left fusiform gyrus and bilateral cerebellar tonsils. rCBF and anxiety were negatively correlated in the left inferior frontal gyrus, right superior frontal gyrus, right middle temporal gyrus, right superior temporal gyrus, and left superior frontal gyrus and positively correlated in the right lingual gyrus and right parahippocampal gyrus. The patterns were similar to those with MD without CKD. HD patients showed bilateral cerebralatrophy in the caudate nucleus and midbrain associated with longer duration of dialysis. Cerebral oxygenation and carotid blood flow were low before dialysis compared with controls. Cerebral oxygenation improved only slightly after HD, while carotid blood flow normalized. PD patients showed values between those of HD patients and controls. Duration of HD was associated with global gray matter volume, change of blood flow during dialysis, and baseline rSO2. HD patients appear to have low CBF between sessions. MCA FV, dynamic pressure autoregulation, and CRCO2 were measured before and after HD. MCA FV decreased from 57cm z s21 before to 46 cm z s21 after HD (P,0.01). The THRR was 1.29 (0.13) before HD and did not change significantly after HD. CRCO2 was 21.7 kPa21 before HD and remained unchanged afterward (20.9 kPa21). MCA FV decreased significantly after HD. Dynamic pressure autoregulation and CRCO2 remain normal in patients with CRF and are not altered significantly by HD.

Main Findings


n=18 HD patients; 10 men and 8 women; mean age, 62 yr

n=18 HD patients; 13 men and 5 women; age range, 46–82 yr

n=17 patients with ESRD or undergoing dialysis; 13 men and 4 women; mean age, 60 yr

Metry et al. (2002)

Lass et al. (1999)

Sample Description

Steafnidis et al. (2005)

Authors


Prospective

Cross-sectional

Cross-sectional

Study Design

Tc-99 HMPAO SPECT



Imaging Technique Before HD, the MFV in the MCA was within normal range and dependent on patient’s age, but during HD it decreased significantly. This decrease in MFV was related to ultrafiltration volume, blood volume, and percent change of the hematocrit of the arterial blood oxygen content and of the plasma fibrinogen levels. The MFV in the MCA decreased continuously during HD. The effect of hemodynamic and rheologic changes on MFV was examined during HD through continuous online monitoring. After HD, hematocrit and blood and plasma viscosity increased significantly. The change in MFV was not significantly different from zero and correlated significantly with change in hematocrit. During HD, the mean arterial pressure in 15 patients moved into the normal range, whereas 3 patients developed hypotension and their MAP decreased. Change in MFV was not significant. CBF does not appear to be diminished significantly during HD. Regional cerebral perfusion imaging was performed, and activity in the frontal and temporal lobes was compared by quantification against the ipsilateral and contralateral cerebellum using three-dimensional surface mapping. Discrete cortical defects consistent with infarcts were seen in 14 patients (82%). The defects were found in all lobules. The mean right and left frontal-to-cerebellar ratios were not significantly different from the right and left temporal-to-cerebellar ratios, and both were within normally accepted ranges, suggesting no evidence of Alzheimer-type dementia.

Main Findings


n=9 patients undergoing long-term HD; 4 men and 5 women; mean age, 50 yr n=44 mentally and neurologically healthy controls; age range, 49–79 yr

Gottlieb et al. (1987) Cross-sectional

Case-control

Study Design

Xe inhalation technique to measure rCBF

133


Imaging Technique Blood FVs were obtained from the MCA and the basilar artery immediately before and after HD. There was a significant reduction in the MFV in the MCA (P,0.01) and the basilar artery (P,0.01) after HD. A negative correlation was observed between the relative change in MFV and the loss of weight after HD, the amount of fluid removed, and the increase in hematocrit in the middle cerebral artery and the basilar artery. HD and the associated physiologic changes can significantly affect cerebral circulation. Transcranial Doppler ultrasonography can effectively monitor rapid changes in the cerebral circulation during HD. rCBF was measured before and after HD. Predialysis rCBF values did not differ from those in controls. After HD, rCBF decreased by a mean of 7% (P=0.02). Post-HD rCBF reduction was not associated with neurologic or cognitive dysfunction. Causes of HD-induced rCBF decreases are unknown, but increased blood viscosity and biochemical changes (e.g., urea reduction, blood alkalinization) are suspected.

Main Findings

eGFR, estimated GFR; ECD, ethylcysteinate dimmer; SPECT, single-photon emission computed tomography; HDRS, Hamilton Depression Rating Scale; rCBF, regional cerebral blood flow; MRS, magnetic resonance spectroscopy; HD, hemodialysis; MD, major depression; PD, peritoneal dialysis; MP-RAGE, magnetization prepared rapid gradient echo; ASL, arterial spin labeling; MRI, magnetic resonance imaging; ICA, internal carotid artery; rSO2, regional oxygen saturation; MCA, middle cerebral artery; FV, flow velocity; CRCO2, carbon dioxide reactivity; THRR, transient hyperemic response ratio; MFV, mean flow velocity; HMPAO, hexamethylpropyleneamine oxime.

n=27 HD patients; 13 men and 14 women; mean age, 52.4 yr

Sample Description

Hata et al. (1994)

Authors



Prospective study

Prospective cohort study

n=33 school-age children who received renal transplant for ESRD when age ,5 yr; 22 boys and 11 girls; 7 children attending special school, 6 received remedial instruction, 20 in regular instruction

n=15 patients who developed CKD at birth or during first year of life; 13 boys and 2 girls; mean follow-up period, 50 mo

Qvist et al. (2002)

Elzouiki (1994)

Cross-sectional study

n=33 renal transplant recipients who received a renal allograft before age 5 yr; 22 boys and 11 girls; age range, 6–11 yr

Valanne et al. (2004)

Retrospective cohort study

Study Design

n=20 renal patients; 13 boys and 7 girls; mean age, 8.7 yr n=10 kidney transplant patients (7 with idiopathic nephritic syndrome, 2 with acute poststreptococcal GN, 1 with diffuse mesangial sclerosis)

Sample Description

Ishikura et al. (2006)

Authors

Table 3. Pediatric structural neuroimaging studies in CKD (n=13)

CT ; EEG; nerve conduction velocity; ABER

Pretransplant CT ; post-transplant MRI

CT and MRI

CT and MRI

Imaging Technique Used In most patients, radiologic abnormalities extended to the gray matter (17 of 20 patients), frontal and temporal lobes, and the cerebellum (16 patients). Posterior reversible encephalopathy syndrome should be suspected in pediatric kidney transplant recipients and patients with kidney disease if they have a sudden episode of neurologic symptoms, even if imaging findings are not restricted to the subcortical white matter of the occipital region. Brain MRI was performed after transplantation. Pretransplant CT scans of 26 patients were available for comparison; 18 patients (54%) had ischemic lesions in the vascular border zones. Mild lesions were seen in 10 patients, moderate in 6, and severe in 2. Other findings were rare. Brain lesions of the patients did not correlate with underlying disease but were related to adverse clinical events before transplant. Prevalence of cerebral atrophy after transplant was 15%; 55% of sample had watershed ischemic lesions (5 of the 7 in special school, 5 of 6 in remedial teaching, 8 of 20 receiving regular instruction). Brain infarcts were seen in 6 of 7 children who were attending special school. Children attending special school had a higher rate of prematurity, hypertensive crises (P=0.002), and seizures (P=0.03), with many of these events occurring during dialysis. All patients had a normal neurologic examination apart from hypotonia. Five had microcephaly. There was a correlation between malnutrition in the first 2 years of life and microcephaly. Developmental delay was present in 3 patients with microcephaly. Brain atrophy was seen in only 3 of 13 patients. EEG results were abnormal in 6 patients. Only 1 patient had diminished nerve conduction velocity; all patients had a normal ABER. None of the patients developed progressive encephalopathy or recurrent seizures. A policy of no oral aluminum therapy and early nutritional support leads to better neurologic outcome in children with early CKD.

Main Findings


n=3 patients with Lowe syndrome

n=10 patients with cystinosis; mean age, 14.2 yr n=10 comparison cases with primary renal disease but no cerebral abnormalities; mean age, 11.8 yr

n=22 patients with CKD (6 with stable CKD, 14 on dialysis, 2 post-transplant); age range, 2–18 yr

n=22 HD patients

n=15 patients with ESRD, none with SLE; 5 boys and 10 girls; age range, 3.5–20 yr; dialysis (n=14) and/or renal transplant (n=13)

Cochat et al. (1986)

Steinberg et al. (1985)

Kretzschmar et al. (1983)

Schnaper et al. (1983)

Sample Description

Pueshel et al. (1991)

Authors


Prospective study

Prospective study


Case-control study


Study Design

CT

CT

CT

CT

MRI

Imaging Technique Used

The study emphasized the CNS and renal pathology of Lowe syndrome. MRI showed diffuse high T2 signals in the periventricular region, indicating significant white matter destruction, which may contribute to the intellectual disability, seizure disorder, hypotonia, and areflexia seen in Lowe’s syndrome. Both groups included dialyzed children or transplant recipients without significant difference in length of ESRD management. Cranial CT showed brain atrophy in cystinotic patients characterized by enlargement of cortical sulci in 80% of cases versus only 20% from the comparison group. Brain atrophy was present in 13 patients (59%), particularly for patients receiving HD. Two patients had cortical infarcts, and 1 had a hypodense area in the basal ganglia. Metabolic derangements and/or the accumulation of toxic substances due to uremic state may be related to brain atrophy in young patients with CKD. Recurrent osmotic changes of the brain during HD may contribute to the presence of brain atrophy. Cranial CT was performed before and after HD in all patients and in half of those patients 8 mo later. In half of cases, brain atrophy was recognizable, with decreased absorption of cerebral parenchyma after HD in 73% of the cases. This finding is partly explained by cerebral edema or hydration. Cortical atrophy was seen in 8 cases and ventricular enlargement in an additional 2. Cortical atrophy was not associated with clinical signs or symptoms. The mean length of time the patients received long-term HD was twice as long in patients with atrophy (30.8 mo) as in those with normal scans (14.6 mo). The length of time that patients received dialysis contributed to the cortical atrophy.

Main Findings


Case-control study

Case-control study

Case-control study

n=30 patients on HD or PD; 16 boys and 14 girls; age range, 10–59 yr n=10 normal controls

n=25 patients with ESRD; 14 boys and 11 girls; mean age, 50 yr n=45 HD patients; 21 boys and 24 girls; mean age, 46 yr n=20 healthy controls; 10 boys and 10 girls; mean age, 55 yr

n=33 renal patients; mean SCr, 11.6 mg/dl; age range, 4–72 yr Healthy controls were age matched to renal group and randomly selected from radiologic files

La Greca (1982)

Papageorgiou et al. (1982)

Passer (1977)

CT

CT

CT

CT

Imaging Technique Used No morphologic modifications were observed. Significant changes in density were demonstrated after a dialysis session in the population treated intermittently. Patients with ESRD who were undergoing CAPD always had normal density values. Cerebral density changes related only to intermittent dialytic treatment. CT of the head was conducted before, immediately after, and 6 hr after dialysis. Density did not decrease in healthy persons or in patients on continuous PD. The absence of density variations in CAPD was explained by a gradual stable delivery of fluid columns and solutes. None of the patients had clinical signs of uremic encephalopathy. Ventricles were significantly enlarged in both groups compared with the controls, with the greatest enlargement in the patients undergoing HD. Ventricular enlargement expresses the degree of brain atrophy, which in ESRD may be due to the metabolic factors of renal disease and in HD to aluminum toxicity. Patients underwent CT as part of their predialysis evaluation. Normal controls were scanned for headaches or traumatic injury to the head. Cerebral atrophy was present in end-stage uremia. Degree of atrophy correlated with the age at onset of end-stage uremia and was most marked during the younger ages of cerebral development.

Main Findings

CT, computed tomography; MRI, magnetic resonance imaging; EEG, electroencephalography; ABER auditory brainstem evoked response; CNS, central nervous system; HD, hemodialysis; SLE, systemic lupus erythematosus; CAPD, continuous ambulatory peritoneal dialysis; PD, peritoneal dialysis; SCr, serum creatinine.

Case-control study

Study Design

n=48 patients with ESRD; 28 boys and 35 girls; age range, 6–71 yr; 38 receiving dialysis n=15 patients with CKD

Sample Description

Dettori et al. (1982)

Authors



1444


Table 4. Summary of common findings across the neuroimaging studies

CKD Populations Studied Common Theme (Total n=43) Non-ESRD Cerebral atrophy and cerebral density changes Fazekas et al. (1995) [A] Savazzi et al. (1999) [A] Yoshimutu et al. (2000) [A] Walters et al. (2001) [A] Aǧildere et al. (2006) [A] Prohovnik et al. (2007) [A] Yakushiji (2010) [A] Passer (1977) [P] Dettori et al. (1982) [P] La Greca et al. (1982) [P] Papageorgiou et al. (1982) [P] Kretzschmar et al. (1983) [P] Schnaper et al. (1983) [P] Steinberg et al. (1985) [P] Cochat et al. (1986) [P] Elzouiki et al. (1994) [P] Cerebral vascular abnormalities White matter lesions/abnormalities Pueshel et al. (1991) [A] Aǧildere et al. (2006) [A] Martinez-Vea et al. (2006) [A] Khatri et al. (2007) [A] Kim et al. (2007) [A] Watanabe et al. (2007) [A] Ikram et al. (2008) [A] Wada et al. (2008) [A] Shima et al. (2010) [A] Kim et al. (2011) [A] Naganuma et al. (2011) [A] Nam et al. (2011) [A] Tryc et al. (2011) [A] Naganuma et al. (2012) [A] Ishikura et al. (2006) [P] Silent cerebral infarction Nakatani et al. (2003) [A] Kobayashi et al. (2009) [A] Kobayashi et al. (2010) [A] Cortical infarcts Lass et al. (1999) [A] Qvist et al. (2002) [A] Kobayashi et al. (2004) [A] Valanne et al. (2004) [A] Effect of hemodialysis on cerebral circulation/oxygenation Gottlieb et al. (1987) [A] Hata et al. (1994) [A] Metry et al. (2002) [A] Skinner et al. (2005) [A] Steafnidis et al. (2005) [A] Prohovnik et al. (2007) [A] Depression and kidney disease Kim t al. (2007) [A] Kim et al. (2008) [A] Nam et al. (2011) [A] Other Intracranial cysts Schievink et al. (1995) [A] A, adult study; P, pediatric study.

Dialysis

Post-Transplant

X X X X X X X X X X X

X X X X X X X

X X

X X X X X X X X X X X X

X X X X X

X

X X

X X X

X

X X

X X X X X X X X X X X X

X


in adults with CKD. Lass et al. (37) documented discrete cortical defects consistent with infarcts in 82% of their heterogeneous patient sample (patients with CKD and those undergoing dialysis), with the defects being observable in each of the lobules. Although magnetic resonance spectroscopy does not assess rCBF, it provides indices that help elucidate neurochemical functions within the brain. To date one study has assessed magnetic resonance spectroscopy in adults with CKD. Using magnetic resonance spectroscopy in HD patients, patients with stage 4 and 5 CKD, and matched healthy controls, Tryc et al. (11) reported reduced brain activity, as defined by targeted neurometabolites related to healthy brain functioning (i.e., choline, N-acetylaspartate, and N-acetylaspartylglutamate), in HD patients. Additionally, memory/learning and attention impairments were observed in both nondialyzed and dialyzed patients with CKD, but more severe cognitive dysfunction and neurometabolic alterations were seen in HD patients than in those with CKD stages 4 and 5. Finally, the relationship between CKD and depression in adults also has received attention in the neuroimaging literature. Kim et al. (14) demonstrated that in 27 patients with CKD, rCBF patterns correlated with symptom clusters of depressive mood. These findings were similar to those in individuals with major depressive disorder without CKD (38) and patients with Parkinson disease and depression (39). More recently, Nam et al. (10) studied 14 patients with CKD before and 6 months after HD and found correlations between reduced levels of depression, higher perfusion in the left middle temporal gyrus, and higher rCBF in the right parahippocampal gyrus. These latter findings implicate the importance of the kidneybrain connection not only in the neurologic aspect of cognition and learning but also in the affective presentation of many individuals with CKD, particularly with respect to the appearance of affective disorders. Table 2 summarizes the adult functional neuroimaging studies in CKD. Pediatric Neuroimaging Studies Thirteen studies included children with CKD in their neuroimaging protocols. Unlike the adult studies, all of the pediatric studies to date have been of a structural nature. These studies span from 1977 (40) to 2006 (41). The study designs included retrospective, cross-sectional, case-control, and prospective approaches to data collection. Sample sizes ranged widely, with the composition of the CKD samples reflecting significant heterogeneity across studies. For the structural neuroimaging studies, all of the pediatric investigations used CT, and several studies also used MRI. Findings in Pediatric Structural Neuroimaging Studies As with adults, cerebral atrophy has been noted in children with ESRD, and this appears to be more prevalent in the HD population. Passer (40) was one of the first few to publish neuroimaging findings in ESRD. Thirty-three patients with ESRD secondary to various causes underwent head CT as part of their predialysis evaluation. This group of patients ranged widely in age and included


1445

children. Cerebral atrophy was noted in ESRD, with the degree of atrophy correlating with the age of onset. Similar findings have been reported by other investigators (42–44) using solely pediatric samples; Qvist et al. (45) documented the presence of cerebral atrophy even after transplant in 15% of their cases. Papageorgiou et al. (46) conducted a similar comparative study of brain atrophy in patients with CKD and those undergoing long-term HD age 16–70 years compared with age-matched controls. None of the patients had clinical signs of uremic encephalopathy. Compared with their age-matched controls, the renal patients had significantly enlarged ventricles, with the greatest enlargement noted in the HD group. Using a small sample of pediatric patients who developed CKD at birth, Elzouki and colleagues (47) found that approximately 23% showed cerebral atrophy. Enlarged ventricles, a correlate of brain atrophy, have also been reported by investigators studying primarily pediatric samples; estimates were significantly higher and ranged from 59% (44) to 67% (43). In addition, enlarged cortical sulci have been reported on cranial CT in up to 80% of pediatric cases with cystinosis (48). Widening of cerebrospinal fluid spaces secondary to brain atrophy (42), with lower cerebral density, has been typically associated with HD (43,49) in comparison with patients receiving continuous ambulatory PD (50). In general, the literature suggested that although brain atrophy was evident in patients with stable CKD and those receiving continuous ambulatory PD, it was definitely more frequent in HD patients. In addition to cerebral atrophy, brain lesions have also been documented using both CT and MRI. Specifically, Ishikura et al. (41) studied 20 cases of posterior reversible encephalopathy syndrome. In most patients, radiologic abnormalities extended to the gray matter (85%), particularly in the frontal and temporal lobes, and the cerebellum. Moreover, ischemic lesions have been documented in the vascular border zones at a rate of about 54% in children who received a renal transplant before 5 years of age (45,51). Cortical infarcts have been documented after transplant as well (44,45). These findings, however, also showed the presence of various neurologic deficits (e.g., brain infarcts, cerebral atrophy) and other factors (e.g., history of hemodynamic crisis, prematurity, congenital nephrosis), including hypercoagulability related to long-term PD before transplantation, suggesting a priori brain insults during the first 5 years of life (51). Table 3 summarizes the pediatric structural neuroimaging studies in CKD.

Discussion This review of the neuroimaging literature in CKD provides a contemporary examination of one important facet of the kidney-brain connection: namely, structural and functional neuroimaging studies in CKD. This review complements earlier reviews conducted on this topic but to our knowledge is the first to include both pediatric and adult studies and to focus on the structural and functional findings in the brain. This review focused on understanding potential linkages between kidney disease and brain structural and functional

1446


integrity across the lifespan. As can be seen in Table 4, several clear findings emerged. In addition to clustering studies under four major themes, Table 4 shows the sample composition for these studies. The three substantive themes are presented below. Cerebral Atrophy and Cerebral Density Changes Are Found in Children and Adults With CKD Both cerebral atrophy and cerebral density changes have been noted in CKD, particularly with more severe kidney disease. These findings were seen more frequently in both child and adult patients on intermittent dialysis treatments (long-term HD, intermittent PD) than in patients with CKD (who were not receiving dialysis) or patients undergoing continuous veno-venous hemofiltration (CVVH). These changes were present in adults but were more marked during the periods of cerebral development in children. These findings were not necessarily associated with the presence of neurologic signs or symptoms, but up to 59% of these cases were noted to have brain atrophy (44). Recurrent osmotic changes of the brain during HD and aluminum toxicity (in older studies) were associated with a greater risk for brain atrophy in HD patients. The degree of brain atrophy also has been related to predialytic BP and the duration of hypertension. Even after transplantation, the prevalence of cerebral atrophy is at least 15%–20%. Similarly, significant changes in cerebral density have been discovered in both children and adults with CKD, with changes reported after HD but not after CVVH. These findings favor continuous RRTs, such as CVVH, to avoid this effect; however, this is not a viable treatment option for long-term outpatient dialysis therapy. In general, these findings support the ongoing study of more frequent and/ or intensive dialysis regimens, in place of HD three times a week (52). Signs of Cerebral Vascular Disease Are Seen in Children and Adults with CKD Both kidney and brain are highly vascularized organs and may share sensitivity to vascular disease. Accordingly, the brains of patients with CKD tend to show confluent deep white matter hyperintensities, WMLs, cerebral microbleeds, and territorial cerebral infarcts. ESRD, chronologic age, and hypertension appear to be significant predictors of the presence of abnormal subcortical white matter. These findings indicate how kidney disease and its sequelae lead to cerebrovascular disease through microangiopathy (23). Further, vascular nephropathy was found to be the most important factor related to the presence of these lesions, suggesting that WMLs may reflect ischemic brain damage secondary to generalized vascular damage. Conversely, stage and duration of CKD have not been linked to the presence of WMLs. CKD has been related to small vessel disease–related lesions. Lower eGFR levels and albuminuria were associated with more lacunar infarcts and higher grades of WMLs and less white matter volume. These findings were present after adjustment for age, sex, and cardiovascular factors and when individuals with stage 2 hypertension, cardiovascular risk factors, or diabetes were excluded from examination (23,25). Similarly, the incidence of

cerebral microbleeds appears to be high in patients receiving maintenance HD, with rates hovering around 25%– 35% (31). These findings point to poor renal function as a significant risk factor for cerebral microbleeds as well as for future cerebrovascular events. It is interesting to know that although no relationship between anticoagulants and cerebral microbleeds has been uncovered in adults, this relationship has not been studied in pediatric patients. Further evidence from these studies points to the presence of cerebral infarctions as being present in CKD, serving as yet another marker of possible cerebrovascular disease. The prevalence of silent cerebral infarctions has been shown to be at least 50% in patients with ESRD (40). Among causes of CKD, hypertensive nephrosclerosis had a strong association with silent brain infarctions, and lower/declining eGFR, chronologic age, and hypertension were independently associated with a higher prevalence of silent brain infarction. Taken together, these neuroimaging findings point to CKD as an independent risk factor for these infarctions and provide a strong linkage to the risk for cerebrovascular disease. Finally, in adults receiving HD, blood flow velocities in the middle cerebral artery and the basilar artery immediately before and after HD are significantly reduced after dialysis treatment (32). Functional imaging techniques, such as transcranial Doppler ultrasonography, should be considered to monitor rapid changes in cerebral circulation during HD treatments. To date, no studies have investigated blood flow velocities in children with CKD, although such studies may provide early clues for the later appearance of cerebrovascular disease. Functional Brain Imaging Techniques Show Circulatory Patterns Consistent with Affective Disorders For adults, the available neuroimaging studies suggest that rCBF patterns in CKD are similar to those seen in patients with major depressive disorder without CKD. Conversely, improvement in depressive symptoms has been associated with higher perfusion in the left middle temporal gyrus and higher rCBF in the right parahippocampal gyrus (10,14). To date, no such studies have been conducted with children and adolescents with CKD.

Summary This comprehensive review has analyzed structural and functional neuroimaging in adults and children with kidney disease. This review points to several key evidencebased findings supporting the ongoing study of the kidneybrain connection, with noteworthy findings indicating the presence of cortical atrophy and other structural differences in both children and adults with CKD, particularly with increasing disease burden and the need for RRT, and while controlling for other disease-related factors. The findings also provided evidence suggesting developmental continuity of cerebrovascular involvement in both children and adults with CKD. These latter studies point to cerebrovascular disease as one primary mechanism for the appearance of abnormal brain structure, impaired brain functions, and perhaps increasing cognitive impairments across the lifespan.


Despite the overarching nature of the available studies, it is important to note that our systematic review also revealed several limitations that moderate the strength of the findings. The corpus of studies spans approximately 35 years, with 30 of the 43 studies involving adults. The number of pediatric studies is increasing, but to date no published functional imaging studies have directly examined brain function in children and adolescents with CKD. Further, many of the earlier pediatric studies actually include both adults and children in their samples, thus obscuring the effects of CKD in child brain structure and function. Study designs ranged widely but largely used cross-sectional approaches to patient ascertainment; this raises issues of how CKD, and its changing degree of severity, interacts with neurodevelopmental aspect of brain structure and function over time. Additionally, very few longitudinal studies have been conducted with adults or children. Only a small subset of the included studies focused on functional neuroimaging. Finally, most of the studies were conducted with small and/or heterogeneous samples of individuals with CKD and RRT, although there were some notable exceptions in the structural neuroimaging literature for adults. In the future, it will be of interest to expand studies to link complications of CKD, such as hypertension and bone mineral disorder, with neuroimaging. One such study of interest is the ongoing, National Institutes of Health– sponsored SPRINT-MIND (Systolic Blood Pressure Intervention Trial-Memory and Cognition in Decreased Hypertension) study, which will determine whether a lower systolic BP goal will further reduce the risk of cardiovascular and kidney disease or age-related cognitive decline. Future pediatric study designs should aim at recruiting pure CKD population (with no history of prematurity, underlying central nervous system abnormalities, or genetic disorders) because that will permit study of the kidney-brain connection with perhaps more careful consideration. The use of neuroimaging procedures as part of routine clinical care requires ongoing investigation, but these findings do point to the use of such procedures for cases showing neurologic symptoms. Their use from a preventive stance for young patients with CKD also holds promise with respect to managing their cerebrovascular health and associated neurologic morbidity across the lifespan. Acknowledgments This study was supported by the Pennsylvania Department of Health: Commonwealth Universal Research Enhancement (CURE) Program, Health Research Formula Grant Award SFY 2010-14 awarded to The Children’s Hospital of Philadelphia. Disclosures S.R.H. reported research and travel support received from Eli Lilly and The Children’s Hospital of Philadelphia. The other authors had no conflicts to report. References 1. Gerson AC, Butler R, Moxey-Mims M, Wentz A, Shinnar S, Lande MB, Mendley SR, Warady BA, Furth SL, Hooper SR: Neurocognitive outcomes in children with chronic kidney disease: Current findings and contemporary endeavors. Ment Retard Dev Disabil Res Rev 12: 208–215, 2006


1447

2. Butler WE, Barker FGI 2nd, Crowell RM: Patients with polycystic kidney disease would benefit from routine magnetic resonance angiographic screening for intracerebral aneurysms: A decision analysis. Neurosurgery 38: 506–515, discussion 515–516, 1996 3. Burn DJ, Bates D: Neurology and the kidney. J Neurol Neurosurg Psychiatry 65: 810–821, 1998 4. Agildere AM, Kurt A, Yıldırım T, Benli S, Altino¨rs N: MRI of neurologic complications in end-stage renal failure patients on hemodialysis: pictorial review. Eur Radiol 11: 1063–1069, 2001 5. Brouns R, De Deyn PP: Neurological complications in renal failure: A review. Clin Neurol Neurosurg 107: 1–16, 2004 6. Lakadamyali H, Ergu¨n T: MRI for acute neurologic complications in end-stage renal disease patients on hemodialysis. Diagn Interv Radiol 17: 112–117, 2011 7. Vogels SCM, Emmelot-Vonk MH, Verhaar HJJ, Koek HL: The association of chronic kidney disease with brain lesions on MRI or CT: A systematic review. Maturitas 71: 331–336, 2012 8. Gottlieb D, Mildworf B, Rubinger D, Melamed E: The regional cerebral blood flow in patients under chronic hemodialytic treatment. J Cereb Blood Flow Metab 7: 659–661, 1987 9. Kim HS, Park JW, Bai DS, Jeong JY, Hong JH, Son SM, Jang SH: Diffusion tensor imaging findings in neurologically asymptomatic patients with end stage renal disease. NeuroRehabilitation 29: 111–116, 2011 10. Nam H-Y, Song SH, Kim S-J, Kwak IS, Kim IJ, Lee SB, Lee DW, Kim BS, Pak K, Kim Y-K, Yun HS: Effect of dialysis on cerebral blood flow in depressive end-stage renal disease patients. Ann Nucl Med 25: 165–171, 2011 11. Tryc AB, Alwan G, Bokemeyer M, Goldbecker A, Hecker H, Haubitz M, Weissenborn K: Cerebral metabolic alterations and cognitive dysfunction in chronic kidney disease. Nephrol Dial Transplant 26: 2635–2641, 2011 12. Ronco C, Bellomo R, Brendolan A, Pinna V, La Greca G: Brain density changes during renal replacement in critically ill patients with acute renal failure. Continuous hemofiltration versus intermittent hemodialysis. J Nephrol 12: 173–178, 1999 13. Schievink WI, Huston J 3rd, Torres VE, Marsh WR: Intracranial cysts in autosomal dominant polycystic kidney disease. J Neurosurg 83: 1004–1007, 1995 14. Kim S-J, Song SH, Kim JH, Kwak IS: Statistical parametric mapping analysis of the relationship between regional cerebral blood flow and symptom clusters of the depressive mood in patients with pre-dialytic chronic kidney disease. Ann Nucl Med 22: 201–206, 2008 15. Watanabe A: Cerebral microbleeds and intracerebral hemorrhages in patients on maintenance hemodialysis. J Stroke Cerebrovasc Dis 16: 30–33, 2007 ¸ akır B, Ozgu¨l E, Kural F, Haberal M: 16. Aǧıldere AM, Basxaran C, C Evaluation of neurologic complications by brain MRI in kidney and liver transplant recipients. Transplant Proc 38: 611–618, 2006 17. Nakatani T, Naganuma T, Uchida J, Masuda C, Wada S, Sugimura T, Sugimura K: Silent cerebral infarction in hemodialysis patients. Am J Nephrol 23: 86–90, 2003 18. Walters RJL, Fox NC, Crum WR, Taube D, Thomas DJ: Haemodialysis and cerebral oedema. Nephron 87: 143–147, 2001 19. Yoshimitsu T, Hirakata H, Fujii K, Kanai H, Hirakata E, Higashi H, Kubo M, Tanaka H, Shinozaki M, Katafuchi R, Yokomizo Y, Oh Y, Tomooka S, Fujimi S, Fujishima M: Cerebral ischemia as a causative mechanism for rapid progression of brain atrophy in chronic hemodialysis patients. Clin Nephrol 53: 445–451, 2000 20. Savazzi GM, Cusmano F, Bergamaschi E, Vinci S, Allegri L, Garini G: Hypertension as an etiopathological factor in the development of cerebral atrophy in hemodialyzed patients. Nephron 81: 17–24, 1999 21. Fazekas G, Fazekas F, Schmidt R, Kapeller P, Offenbacher H, Krejs GJ: Brain MRI findings and cognitive impairment in patients undergoing chronic hemodialysis treatment. J Neurol Sci 134: 83–88, 1995 22. Martinez-Vea A, Salvado´ E, Bardajı´ A, Gutierrez C, Ramos A, Garcıá C, Compte T, Peralta C, Broch M, Pastor R, Angelet P, Marcas L, Saurı´ A, Oliver JA: Silent cerebral white matter lesions and their relationship with vascular risk factors in middle-aged predialysis patients with CKD. Am J Kidney Dis 47: 241–250, 2006

1448


23. Wada M, Nagasawa H, Iseki C, Takahashi Y, Sato H, Arawaka S, Kawanami T, Kurita K, Daimon M, Kato T: Cerebral small vessel disease and chronic kidney disease (CKD): Results of a crosssectional study in community-based Japanese elderly. J Neurol Sci 272: 36–42, 2008 24. Khatri M, Wright CB, Nickolas TL, Yoshita M, Paik MC, Kranwinkel G, Sacco RL, DeCarli C: Chronic kidney disease is associated with white matter hyperintensity volume: The Northern Manhattan Study (NOMAS). Stroke 38: 3121–3126, 2007 25. Ikram MA, Vernooij MW, Hofman A, Niessen WJ, van der Lugt A, Breteler MMB: Kidney function is related to cerebral small vessel disease. Stroke 39: 55–61, 2008 26. Yakushiji Y, Nanri Y, Hirotsu T, Nishihara M, Hara M, Nakajima J, Eriguchi M, Nishiyama M, Hara H, Node K: Marked cerebral atrophy is correlated with kidney dysfunction in nondisabled adults. Hypertens Res 33: 1232–1237, 2010 27. Naganuma T, Takemoto Y, Shoji T, Shima H, Ishimura E, Okamura M, Nakatani T: Factors associated with cerebral white matter hyperintensities in haemodialysis patients. Nephrology (Carlton) 17: 561–568, 2012 28. Kobayashi S, Ikeda T, Moriya H, Ohtake T, Kumagai H: Asymptomatic cerebral lacunae in patients with chronic kidney disease. Am J Kidney Dis 44: 35–41, 2004 29. Kobayashi M, Hirawa N, Yatsu K, Kobayashi Y, Yamamoto Y, Saka S, Andoh D, Toya Y, Yasuda G, Umemura S: Relationship between silent brain infarction and chronic kidney disease. Nephrol Dial Transplant 24: 201–207, 2009 30. Shima H, Ishimura E, Naganuma T, Yamazaki T, Kobayashi I, Shidara K, Mori K, Takemoto Y, Shoji T, Inaba M, Okamura M, Nakatani T, Nishizawa Y: Cerebral microbleeds in predialysis patients with chronic kidney disease. Nephrol Dial Transplant 25: 1554–1559, 2010 31. Naganuma T, Takemoto Y, Yamasaki T, Shima H, Shoji T, Ishimura E, Nishizawa Y, Morino M, Okamura M, Nakatani T: Factors associated with silent cerebral microbleeds in hemodialysis patients. Clin Nephrol 75: 346–355, 2011 32. Prohovnik I, Post J, Uribarri J, Lee H, Sandu O, Langhoff E: Cerebrovascular effects of hemodialysis in chronic kidney disease. J Cereb Blood Flow Metab 27: 1861–1869, 2007 33. Skinner H, Mackaness C, Bedforth N, Mahajan R: Cerebral haemodynamics in patients with chronic renal failure: Effects of haemodialysis. Br J Anaesth 94: 203–205, 2005 34. Stefanidis I, Bach R, Mertens PR, Liakopoulos V, Liapi G, Mann H, Heintz B: Influence of hemodialysis on the mean blood flow velocity in the middle cerebral artery. Clin Nephrol 64: 129–137, 2005 35. Hata R, Matsumoto M, Handa N, Terakawa H, Sugitani Y, Kamada T: Effects of hemodialysis on cerebral circulation evaluated by transcranial Doppler ultrasonography. Stroke 25: 408– 412, 1994 36. Metry G, Spittle M, Rahmati S, Giller C, Giller A, Kaufman A, Schneditz D, Manno E, Brener Z, Boniece I, Ronco F, Ronco C, Levin NW: Online monitoring of cerebral hemodynamics during hemodialysis. Am J Kidney Dis 40: 996–1004, 2002 37. Lass P, Buscombe JR, Harber M, Davenport A, Hilson AJ: Cognitive impairment in patients with renal failure is associated

38.

39.

40. 41.

42. 43. 44. 45.

46.

47. 48. 49. 50. 51. 52.

with multiple-infarct dementia. Clin Nucl Med 24: 561–565, 1999 Bench CJ, Friston KJ, Brown RG, Frackowiak RS, Dolan RJ: Regional cerebral blood flow in depression measured by positron emission tomography: The relationship with clinical dimensions. Psychol Med 23: 579–590, 1993 Wu H, Lou C, Huang Z, Shi G: SPECT imaging of dopamine transporters with (99m)Tc-TRODAT-1 in major depression and Parkinson’s disease. J Neuropsychiatry Clin Neurosci 23: 63–67, 2011 Passer JA: Cerebral atrophy in end-stage uremia. Proc Clin Dial Transplant Forum 7: 91–94, 1977 Ishikura K, Ikeda M, Hamasaki Y, Hataya H, Shishido S, Asanuma H, Nishimura G, Hiramoto R, Honda M: Posterior reversible encephalopathy syndrome in children: Its high prevalence and more extensive imaging findings. Am J Kidney Dis 48: 231–238, 2006 Kretzschmar K, Nix W, Zschiedrich H, Philipp T: Morphologic cerebral changes in patients undergoing dialysis for renal failure. AJNR Am J Neuroradiol 4: 439–441, 1983 Schnaper HW, Cole BR, Hodges FJ, Robson AM: Cerebral cortical atrophy in pediatric patients with end-stage renal disease. Am J Kidney Dis 2: 645–650, 1983 Steinberg A, Efrat R, Pomeranz A, Drukker A: Computerized tomography of the brain in children with chronic renal failure. Int J Pediatr Nephrol 6: 121–126, 1985 Qvist E, Pihko H, Fagerudd P, Valanne L, Lamminranta S, Karikoski J, Sainio K, Ro¨nnholm K, Jalanko H, Holmberg C: Neurodevelopmental outcome in high-risk patients after renal transplantation in early childhood. Pediatr Transplant 6: 53–62, 2002 Papageorgiou C, Ziroyannis P, Vathylakis J, Grigoriadis A, Hatzikonstantinou V, Capsalakis Z: A comparative study of brain atrophy by computerized tomography in chronic renal failure and chronic hemodialysis. Acta Neurol Scand 66: 378–385, 1982 Elzouki A, Carroll J, Butinar D, Moosa A: Improved neurological outcome in children with chronic renal disease from infancy. Pediatr Nephrol 8: 205–210, 1994 Cochat P, Drachman R, Gagnadoux M-F, Pariente D, Broyer M: Cerebral atrophy and nephropathic cystinosis. Arch Dis Child 61: 401–403, 1986 Dettori P, La Greca G, Biasioli S, Chiaramonte S, Fabris A, Feriani M, Pinna V, Pisani E, Ronco C: Changes of cerebral density in dialyzed patients. Neuroradiology 23: 95–99, 1982 La Greca G, Biasioli S, Chiaramonte S, Dettori P, Fabris A, Feriani M, Pinna V, Pisani E, Ronco C: Studies on brain density in hemodialysis and peritoneal dialysis. Nephron 31: 146–150, 1982 Valanne L, Qvist E, Jalanko H, Holmberg C, Pihko H: Neuroradiologic findings in children with renal transplantation under 5 years of age. Pediatr Transplant 8: 44–51, 2004 Fischbach M, Fothergill H, Zaloszyc A, Menouer S, Terzic J: Intensified daily dialysis: the best chronic dialysis option for children? Semin Dial 24: 640–644, 2011

Published online ahead of print. Publication date available at www. cjasn.org.