Normalization of Cerebral Blood Flow, Neurochemicals, and White Matter Integrity after Kidney Transplantation : Journal of the American Society of Nephrology

Journal Logo

Clinical Research

Normalization of Cerebral Blood Flow, Neurochemicals, and White Matter Integrity after Kidney Transplantation

Lepping, Rebecca J.1,2; Montgomery, Robert N.3; Sharma, Palash3; Mahnken, Jonathan D.2,3; Vidoni, Eric D.2,4; Choi, In-Young1,4; Sarnak, Mark J.5; Brooks, William M.1,2,4,6; Burns, Jeffrey M.2,4,6; Gupta, Aditi2,7,8

Author Information
JASN 32(1):p 177-187, January 2021. | DOI: 10.1681/ASN.2020050584
  • Free
  • Infographic
  • SDC


CKD is associated with increased risk of dementia, stroke, and mortality.1 For patients with ESKD, dialysis is lifesaving, but initiation of dialysis increases strokes and mortality.2 These deleterious clinical findings in CKD can be explained by the structural and physiologic brain abnormalities of increased elevated cerebral blood flow (CBF), altered cerebral neurochemical concentrations, and decreased white matter integrity.3–15

Patients with CKD have elevated CBF,7,8 likely secondary to disrupted cerebral autoregulation from dysfunctional blood-brain barrier due to endothelial inflammation.16,17 In addition, hemodialysis causes acute changes in the CBF, associated with ultrafiltration volume and change in hematocrit—major determinants of blood viscosity.18 These changes in cerebral hemodynamics can contribute to increase in stroke.5,19,20 Patients with CKD also have dysregulation of cerebral neurochemicals. Cerebral osmolytes, such as myo-inositol (mI) and choline-containing compounds (Cho), are elevated in CKD.45678,9–11 Although causality is unproven, lower eGFR is correlated with higher concentrations of some of these osmolytes.21 Increase in cerebral osmolytes can alter cerebral osmotic pressure and impair cellular structure and function.22 In addition, patients with CKD have decreased white matter integrity measured by diffusion tensor imaging (DTI),12–15 a magnetic resonance imaging (MRI) method to measure the diffusion of water molecules along nerve tracts.23

Each of these abnormalities is associated with cognitive impairment in the non-CKD population24,25 and may be responsible for the high prevalence of cognitive impairment in ESKD.26 Although many conditions associated with cognitive impairment, such as Alzheimer disease and traumatic brain injury, have irreversible brain abnormalities,27–30 due to the unique nature of brain disease in CKD, it is possible that some of these abnormalities are reversible. Because cognition improves after kidney transplantation (KT),31 we hypothesize that brain abnormalities will also improve with KT. We used sophisticated noninvasive MRI methods to characterize the structural and physiologic brain abnormalities in ESKD and determine their reversibility with KT. We examined CBF, cerebral neurochemical concentrations, and white matter integrity longitudinally in patients before and after KT. This study builds upon the results of our prior study31 and reports a comprehensive evaluation of brain abnormalities pre-KT to post-KT.


This is a single-center, prospective, longitudinal, observational cohort study on the effect of KT on brain health of patients with ESKD on the KT waiting list in comparison with healthy non-CKD controls. The study was approved by the institutional review board and is registered in the US National Library of Medicine (; NCT01883349). Patients wait listed for KT were enrolled and followed for 12 months post-KT. Brain MRI was performed at baseline, 3 months post-KT, and 12 months post-KT. If patients did not receive KT within 1 year of the baseline visit, a second pre-KT brain MRI was performed to keep the time period between the last pre-KT MRI and KT to <1 year. Controls underwent brain MRI once.


Adult participants between the ages of 30 and 70 who were listed for KT and expected to receive a KT within a year were enrolled. These included patients scheduled for a living donor KT and patients wait listed for at least 2 years. Exclusion criteria included multiorgan listing, MRI contraindications (including claustrophobia), recent stroke, uncontrolled psychosis, active seizure disorder, or current use of antipsychotics or antiepileptics. Controls were age-matched healthy persons without CKD or stroke and without contraindications to MRI. All participants signed informed consent prior to initiating study procedures.

Demographics and Clinical Data

Demographic and clinical data were obtained from patients’ medical records and interviews. Demographic data included age, race, sex, and education. Clinical data included assessment of comorbidities, specifically history of coronary artery disease (defined as history of myocardial infarction, coronary angioplasty, or coronary artery bypass grafting), diabetes (defined as past or current use of oral hypoglycemics or insulin), hypertension (defined as past or current use of antihypertensives), stroke, depression, smoking (in the last 100 days), use of anticoagulants, atrial fibrillation, primary cause of ESKD, and dialysis modality. The type of KT, time on dialysis before KT, kidney function at the time of KT for preemptive KTs, induction, panel reactive antibody, delayed graft function, and episodes of rejection were also recorded for patients who received a KT during the study period.

BP, heart rate, and body mass index were also measured. At each visit, laboratory results, including hemoglobin, serum creatinine, and tacrolimus levels, were recorded from the latest values within the last 3 months. Serum hemoglobin and creatinine for controls were obtained from their primary care provider or tested specifically for the study to avoid enrollment of controls with undiagnosed CKD. Controls with eGFR of <60 ml/min per 1.73 m2 were excluded from the study.

MRI Data Acquisition

All MRI scans were performed using a 3-Tesla whole-body MRI scanner with a 20-channel head/neck receiver coil (Siemens Skyra, Erlangen, Germany). CBF was measured using pulsed arterial spin labeling (ASL), cerebral neurochemical concentrations were measured with magnetic resonance spectroscopic imaging (MRSI), and white matter integrity was measured by DTI. Neurochemicals measured were N-acetylaspartate (NAA), Cho, glutamate and glutamine (Glx), mI, and total creatine (Cr). Acquisition parameters for each modality were designed to collect data from the whole brain while maintaining good data quality, high signal-noise ratio, and acceptable spatial resolution. Detailed protocols for acquisition of CBF, cerebral neurochemicals, and white matter integrity and MRI data analysis are presented in Supplemental Material.

Research team members performing the MRI analysis were not blinded to the visit number of the participant, However, because of the design of the study with a follow-up pre-KT MRI in 1 year with patients who were not transplanted, the analysis team did not the know the exact transplant status of a participant.

Primary and Secondary Outcomes

Our primary outcomes were the effects of KT on whole-brain gray matter CBF, cerebral neurochemicals (NAA/Cr, Cho/Cr, Glx/Cr, and mI/Cr), and DTI-measured whole-brain white matter fractional anisotropy (FA) and mean diffusivity (MD). As secondary outcomes, to better understand possible anatomic specificity of changes, we measured regional CBF, FA, and MD in atlas-defined anatomic regions distributed across the entire brain.

Statistical Analyses

Baseline characteristics between patients and controls were analyzed using descriptive statistics. For categorical variables, differences in frequencies were measured using a nonparametric Fisher exact test. For continuous variables, mean differences were measured using a two-sample t test. Assumptions of normality for the t test were inspected by the quantile-quantile plot and histogram. For clinical measurements, unadjusted one-way repeated ANOVA was used to measure the continuous outcomes between patient pre-KT and patient post-KT groups. Residuals were assessed to evaluate the fit of the underlying model assumptions.

Each imaging modality (ASL, MRSI, and DTI) was analyzed separately. Imaging data in controls, as well as pre- and post-KT groups, were adjusted for covariates of age,32 race,33 sex,34 and education level35,36 that can confound results. Because the exact time for KT is unpredictable, a linear mixed model approach was used to account for repeated observations on participants. A linear mixed effects model was fitted using the PROC GLIMMIX procedure by incorporating correlations among the responses. To model these correlations, we included random effects in the linear predictor. Specifically, we used a random intercept for each subject, which provides a compound symmetry covariance structure. This linear mixed model is more flexible than repeated measures ANOVA as it allows for variation in time between scans and in the number of scans per participant and allows the inclusion of participants who have missing data at one or more time points without requiring an imputation method.37,38 Each participant’s individual intercept was estimated in the model using the random intercept term. Age was calculated at every time point in order to both control for the effect of age and function as the time variable. All available imaging data were included in the model to detect any group differences between patients pre-KT, patients post-KT, and controls. The pre-KT group includes both pre-KT scans, if available; the post-KT group includes both the 3- and 12-month post-KT scans, if available. The effects of group (control, pre-KT, and post-KT) and covariates (age, race, sex, and education level) were included as fixed effects. Residual plots were assessed to evaluate the model fit. We performed pairwise testing of the three groups using F and t tests as appropriate. We constructed scatterplots mapping each participant’s observed measures over time (chronological age), showing both the overall change with aging and the effect of KT. As a test for whether there would be any change associated with time alone, for those patients who had repeated pre-KT MRI, the pre-KT scans were compared using the Wilcoxon nonparametric signed rank test. Results with P<0.05 were considered statistically significant. All statistical analyses were performed using R Studio (version 3.6.3) and SAS (version 9.4; The SAS Institute, Cary, NC).


Demographic and Clinical Comparisons

A total of 48 participants (29 patients and 19 controls) were enrolled (Table 1). Of the 29 patients, 23 received a KT. Twenty-two patients completed the 3-month post-KT MRI, and 18 completed the 12-month post-KT MRI (Supplemental Figure 1). Four of the 29 patients were not transplanted within 1 year, and per study protocol, they underwent repeated baseline MRI at 1 year after enrollment. Patients and controls had no differences in age, race, sex, or education (all P>0.10) (Table 1). The mean serum creatinine and eGFR for controls were 0.85±0.13 mg/dl and 92.8±13.5 ml/min, respectively. Patients with ESKD had a greater incidence of diabetes (P=0.03) and hypertension (P<0.001) compared with controls. The most common causes of ESKD were diabetes (21%) and autosomal dominant polycystic kidney disease (28%). Six patients had not initiated dialysis at baseline. Supplemental Table 1 describes the patients with and without a post-KT MRI. Clinical characteristics pre- and post-KT are summarized in Table 2. Hematocrit and serum calcium increased and serum bicarbonate decreased after KT. Other parameters, such as BP, heart rate, and body mass index, did not change with KT (all P=0.10) (Table 2).

Table 1. - Baseline demographics and clinical characteristics of study participants
Characteristic Controls, n=19 Patients with ESKD, n=29 P Value
Age, yr, mean ± SD 48.78±8.38 53.43±11.38 0.11
Race, n (%) 0.65
 White 17 (89.5) 25 (86.2)
 Black 0 (0.0) 2 (6.9)
 Other 2 (10.5) 2 (6.9)
Sex, n (%) 0.23
 Men 9 (47.4) 20 (69.0)
 Women 10 (52.6) 9 (31.0)
Education, n (%) 0.32
 High school diploma 1 (5.3) 5 (17.2)
 Some college 3 (15.8) 9 (31.0)
 4-yr college degree 7 (36.8) 8 (27.6)
 Graduate school 8 (42.1) 7 (24.1)
Comorbid conditions, n (%)
 Coronary artery disease 0 (0.0) 3 (10.3) 0.27
 Diabetes 0 (0.0) 7 (24.1) 0.03
 Hypertension 5 (26.3) 25 (86.2) <0.001
 Stroke 0 (0.0) 1 (3.5) 0.99
 Depression 2 (10.5) 7 (24.1) 0.29
 Smoking 6 (31.6) 9 (31.0) 0.99
On anticoagulants, n (%) 2 (6.9)
Atrial fibrillation, n (%) 3 (10.3)
Cause of ESKD, n (%) NA NA
 Diabetes 6 (20.7)
 Hypertension 3 (10.3)
 ADPKD 8 (27.6)
 Other 12 (41.4)
Dialysis modality, n (%) NA NA
 In-center hemodialysis 11 (37.9)
 Home hemodialysis 3 (10.4)
 Peritoneal dialysis 9 (31.0)
 Not on dialysis 6 (20.7)
Continuous values are presented as mean ± SD. P values represent the unadjusted two-sample t test. Categorical variables are presented as frequency (percentage). P values represent the unadjusted nonparametric Fisher exact test. NA, not applicable; ADPKD, autosomal dominant polycystic kidney disease.

Table 2. - Pre- to post-KT changes in clinical measures
Clinical Measure 1-yr Pre-KT, n=4, Mean ± SD Pre-KT, n=29, Mean ± SD 3-mo Post-KT, n=22, Mean ± SD 12-mo Post-KT, n=18, Mean ± SD P Value
Hematocrit, % 33.9±4.6 33.3±4.1 39.7±4.2 43.6±4.2 <0.001
Serum calcium, mg/dl 9.5±0.4 9.5±0.5 9.8±0.5 9.7±0.4 0.004
Serum bicarbonate, mEq/L 23.3±2.2 25.5±3.5 24.0±1.8 23.3±2.7 0.02
Serum creatinine, mg/dl NA NA 1.4±0.5 1.3±0.5 NA
eGFR, ml/min per 1.73 m2 NA NA 55.4±13.7 55.7±13.8 NA
Tacrolimus level, ng/ml NA NA 11.4±12.4 7.8±3.2 NA
SBP, mm Hg 114.5±7.0 141.9±16.5 137.5±17.8 138.8±15.1 0.69
DBP, mm Hg 62.3±10.5 79.2±10.3 77.7±10.1 77.3±13.2 0.98
HR, beats per minute 75.5±7.7 75.9±14.5 75.4±10.8 74.4±16.1 0.67
BMI, kg/m2 31.0±3.8 30.0±4.8 30.0±5.1 35.0±13.2 0.20
All measures are presented as mean ± SD. P values represent unadjusted one-way repeated measures ANOVA for pre- to post-transplant comparison. NA, not applicable; SBP, systolic BP; DBP, diastolic BP; HR, heart rate; BMI, body mass index.

All transplanted patients received induction and maintenance immunosuppression per institutional protocol. Supplemental Table 2 describes the clinical characteristics of the 22 patients who had a post-KT MRI. Per institutional policy, ABO-incompatible KTs were not performed at our center. Induction immunosuppression consisted of either thymoglobulin or basiliximab (depending on immunologic risk of the patient), steroids, and a mycophenolate compound. A calcineurin inhibitor (tacrolimus) was started 24 hours after the surgery. Maintenance immunosuppression consisted of a mycophenolate compound and tacrolimus with or without low-dose steroids. Post-KT, all patients were on tacrolimus and had a functional graft with a mean serum creatinine of 1.4±0.5 mg/dl (Table 2). Supplemental Table 3 shows all MRI measurements in controls and patients with ESKD. Individual imaging results are described below. Supplemental Table 4 shows the estimated effect of covariates for the MRI variables in the mixed model analysis.


Table 3 shows the comparisons of adjusted CBF between different groups over the total gray matter and in specific anatomic regions within the total gray matter mask. CBF in the total gray matter was higher in patients with ESKD pre-KT compared with controls (P=0.003) (Figure 1, Table 3) and normalized post-KT to values observed in controls. Figure 1 shows data for individual participants at pre- and post-KT for comparison. When CBF was analyzed regionally, the decrease in CBF with KT was consistent across all brain regions (Table 3).

Table 3. - Overall gray matter and regional comparisons of CBF in patients with ESKD pre-KT, patients with ESKD post-KT, and controls
Area of the Brain Analyzed P Value a Pairwise Comparisons Estimated Difference (P Value) b
Change Pre- to Post-KT Pre-KT versus Control Post-KT versus Control
Total gray matter <0.001 −7.24 (<0.001) 7.91 (0.003) 0.67 (0.78)
Regions analyzed
 ACC 0.07 −5.40 (0.02) 5.21 (0.20) −0.19 (0.96)
 Frontal mid 0.008 −7.37 (0.002) 7.00 (0.05) −0.37 (0.91)
 Hippocampus <0.001 −8.67 (<0.001) 10.89 (0.002) 2.21 (0.46)
 M1 0.002 −6.30 (0.001) 7.58 (0.02) 1.28 (0.65)
 PCC <0.001 −12.17 (<0.001) 9.39 (0.06) −2.78 (0.54)
 Precuneus <0.001 −10.40 (<0.001) 9.12 (0.02) −1.28 (0.71)
 Superior parietal 0.03 −6.77 (0.01) 4.45 (0.24) −2.32 (0.49)
 Temporal 0.001 −5.84 (0.001) 7.69 (0.01) 1.85 (0.47)
 Thalamus <0.001 −12.87 (<0.001) 13.15 (0.01) 0.28 (0.95)
 Pallidum 0.03 −6.11 (0.01) 5.77 (0.17) −0.35 (0.93)
 Putamen 0.001 −7.09 (0.001) 8.26 (0.02) 1.17 (0.71)
 Caudate 0.008 −7.13 (0.002) 6.63 (0.05) −0.49 (0.87)
 Frontal 0.003 −6.30 (0.001) 7.26 (0.01) 0.96 (0.71)
 Parietal <0.001 −8.36 (<0.001) 8.85 (0.01) 0.49 (0.87)
ACC, anterior cingulate cortex; frontal mid, middle frontal gyrus; M1, primary motor cortex; PCC, posterior cingulate cortex.
aP value for linear mixed model F test for any group differences between patients pre-KT, patients post-KT, and controls, adjusted for age, race, sex, and education level. Pre-KT includes both pre-KT scans, if available; post-KT includes both the 3- and 12-month post-KT scans, if available.
bP value for adjusted linear contrast of the pairwise estimated group differences t test. All units are in milliliters per minute per 100 g tissue.

Figure 1.:
CBF in total gray matter was elevated in patients pre-KT compared to controls, and normalized post-KT. Scatterplot displays individual participant data as a function of age and group (controls: black circles; pre-KT: blue triangles; post-KT: orange squares). Dashed lines represent individual participant trajectories over time. Distance between the solid lines represents the estimated group main effect (i.e., intercept) differences for fixed covariate values of race (White), sex (men), education (more than high school), and slice position (upper). The slope of the solid lines represents the overall effect of age. P values represent comparisons between the group main effects (i.e., intercepts).

Brain Neurochemicals

Table 4 shows the group comparisons of different neurochemicals in the adjusted linear mixed model. Figure 2 shows data for individual patients at different time points in the study. Of the four neurochemicals that were analyzed, the Cho/Cr (P=0.001) and mI/Cr (P<0.001) were higher in patients pre-KT compared with controls and normalized post-KT. NAA/Cr and Glx/Cr were not different between the patients pre-KT and controls and did not change post-KT.

Table 4. - Comparisons of brain neurochemical concentrations normalized to creatine in patients with ESKD pre-KT, patients with ESKD post-KT, and controls
Neurochemical Analyzed P Value a Pairwise Comparisons Estimated Difference (P Value) b
Change Pre- to Post-KT Pre-KT versus Control Post-KT versus Control
NAA-Cr 0.04 0.02 (0.14) 0.06 (0.08) 0.09 (0.02)
Cho-Cr <0.001 −0.03 (<0.001) 0.04 (0.001) 0.008 (0.47)
Glx-Cr 0.23 0.05 (0.23) −0.08 (0.10) −0.04 (0.47)
mI-Cr <0.001 −0.06 (<0.001) 0.11 (<0.001) 0.05 (0.05)
aP value for linear mixed model F test for any group differences between patients pre-KT, patients post-KT, and controls, adjusted for age, race, sex, and education level. Pre-KT includes both pre-KT scans, if available; post-KT includes both the 3- and 12-month post-KT scans, if available.
bP value for adjusted linear contrast of the pairwise estimated group differences t test. Units are ratios to creatine.

Figure 2.:
MRSI-measured neurochemicals Cho and mI were elevated in patients pre-KT compared to controls, and normalized post-KT; neurochemicals NAA and Glx were not impacted by ESKD. Panels display (A) NAA-Cr ratio, (B) Cho-Cr ratio, (C) Glx-Cr ratio, and (D) mI-Cr ratio in patients pre-KT, patients post-KT, and controls. Scatterplots display individual participant data as a function of age and group (controls: black circles; pre-KT: blue triangles; post-KT: orange squares). Dashed lines represent individual participant trajectories over time. Distance between the solid lines represents the estimated group main effect (i.e., intercept) differences for fixed covariate values of race (White), sex (men), and education (more than high school). The slope of the solid lines represents the overall effect of age. P values represent comparisons between the group main effects (i.e., intercepts).

White Matter Integrity

Figure 3 and Table 5 show adjusted group comparisons of FA and MD for all tracts and for each of the regional tracts. FA increased pre- to post-KT (P=0.001). Neither pre-KT (P=0.23) nor post-KT (P=0.72) FA was different from controls. MD decreased post-KT (P<0.001). Similar to FA, neither pre-KT (P=0.33) nor post-KT (P=0.89) MD was different from controls.

Figure 3.:
Diffusion metrics (A) FA and (B) MD in whole brain changed pre- to post-KT. Scatterplots display individual participant data as a function of age and group (controls: black circles; pre-KT: blue triangles; post-KT: orange squares). Dashed lines represent individual participant trajectories over time. Distance between the solid lines represents the estimated group main effect (i.e., intercept) differences for fixed covariate values of race (White), sex (men), and education (more than high school). The slope of the solid lines represents the overall effect of age. P values represent comparisons between the group main effects (i.e., intercepts).
Table 5. - Whole-brain and regional comparisons of diffusion metrics FA and MD in patients with ESKD pre-KT, patients with ESKD post-KT, and controls
Regions and Metrics Analyzed P Value a Pairwise Comparisons Estimated Difference (P Value) b
Change Pre- to Post-KT Pre-KT versus Control Post-KT versus Control
All tracts FA 0.003 0.007 (0.001) −0.010 (0.23) −0.003 (0.72)
All tracts MD 0.001 −0.022 (<0.001) 0.019 (0.33) −0.003 (0.89)
Regional FA
 ATR 0.001 0.011 (<0.001) −0.018 (0.19) −0.007 (0.63)
 CG 0.71 0.004 (0.58) −0.024 (0.52) −0.020 (0.59)
 CH 0.51 0.011 (0.46) 0.017 (0.52) 0.028 (0.29)
 CST 0.06 0.009 (0.02) −0.001 (0.93) 0.008 (0.50)
 FMAJ 0.26 0.006 (0.14) −0.015 (0.40) −0.009 (0.62)
 FMIN 0.22 0.006 (0.16) 0.008 (0.45) 0.014 (0.19)
 IFOF 0.02 0.008 (0.01) −0.015 (0.20) −0.007 (0.56)
 ILF 0.005 0.009 (0.002) −0.015 (0.16) −0.006 (0.58)
 SLF 0.23 0.002 (0.54) −0.015 (0.09) −0.013 (0.14)
 SLFT 0.60 0.002 (0.52) −0.010 (0.40) −0.007 (0.52)
 UF 0.27 0.011 (0.15) −0.017 (0.31) −0.007 (0.70)
Regional MD
 ATR 0.008 −0.037 (0.002) 0.045 (0.47) 0.008 (0.90)
 CG 0.002 −0.016 (0.005) 0.020 (0.25) 0.004 (0.83)
 CH 0.33 −0.017 (0.50) −0.070 (0.25) −0.087 (0.15)
 CST 0.12 −0.011 (0.09) −0.013 (0.41) −0.024 (0.14)
 FMAJ 0.004 −0.042 (0.002) 0.049 (0.15) 0.007 (0.83)
 FMIN 0.08 −0.020 (0.03) 0.002 (0.92) −0.018 (0.35)
 IFOF <0.001 −0.019 (<0.001) 0.027 (0.11) 0.009 (0.60)
 ILF 0.003 −0.023 (0.001) 0.026 (0.35) 0.002 (0.93)
 SLF 0.003 −0.012 (0.001) 0.022 (0.13) 0.009 (0.52)
 SLFT <0.001 −0.010 (<0.001) 0.015 (0.12) 0.005 (0.59)
 UF 0.33 −0.020 (0.14) 0.009 (0.82) −0.011 (0.78)
ATR, anterior thalamic radiation; CG, cingulum in the cingulated cortex area; CH, cingulum in the hippocampal area; CST, corticospinal tract; FMAJ, forceps major; FMIN, forceps minor; IFOF, inferior fronto-occipital fasciculus; ILF, inferior longitudinal fasciculus; SLF, superior longitudinal fasciculus; SLFT, temporal projection of the SLF; UF, uncinate fasciculus.
aP value for linear mixed model F test for any group differences between patients pre-KT, patients post-KT, and controls, adjusted for age, race, sex, and education level. Pre-KT includes both pre-KT scans, if available; post-KT includes both the 3- and 12-month post-KT scans, if available.
bP value for adjusted linear contrast of the pairwise estimated between-group differences t test. Units are scalar (zero to one) for FA and 10−3 mm2/s for MD.

Brain Changes without KT

Of the 29 patients with ESKD in the study, four patients had a second pre-KT MRI while awaiting KT. There was no change detected in CBF, cerebral neurochemicals, FA, or MD over 1 year in these patients (all P>0.10) (Supplemental Table 5).


We found that structural and physiologic brain abnormalities in ESKD reversed post-KT. CBF, which was elevated pre-KT, decreased post-KT to levels in controls, both globally and in each anatomic region of gray matter analyzed. Cho/Cr and mI/Cr, which were also elevated pre-KT, normalized post-KT. Finally, with an increase in FA and decrease in MD, white matter integrity also improved post-KT. We also noted the expected age-related changes in brain measurements that were controlled for in our linear mixed model.

The reversibility of brain abnormalities in CKD might have important mechanistic and therapeutic implications. Potential reversibility underscores an opportunity to develop improved management strategies other than KT for patients who cannot be transplanted. Better dialysis techniques that include targeting dialysis-related ischemia to reduce metabolic derangements could help prevent and perhaps mitigate some of these abnormalities. Preservation of residual renal function to retain tubular secretion may also positively affect these brain abnormalities. The overarching effect of our observations would be to identify patients with reversible versus irreversible brain abnormalities and to devise strategies to reverse brain abnormalities with the goal of improving cognition. Cognitive impairment is common in patients with ESKD undergoing KT evaluation39 and affects eligibility for KT.40 However, given that cognitive impairment and brain abnormalities might improve with KT, more of these patients who are currently rejected for KT might get transplanted.

Elevated CBF seen in our patients pre-KT is consistent with other studies in adult8,17 and pediatric41 patients with CKD. Disruption of cerebral autoregulation due to inflammation and endothelial dysfunction affecting the blood-brain barrier may play a role.16,17 Alternatively, elevated CBF could be secondary to increased metabolic demand. Our results that CBF decreased pre- to post-KT is consistent with prior cross-sectional42 and longitudinal studies,43 although some other studies found conflicting results,44,45 possibly due to differences in the methodology of assessing CBF. ASL has the advantage of repeatability and ability to quantify global and regional CBF.46 Because hemodialysis preferentially affects the watershed areas of the brain,47 we also explored regional changes in the brain in addition to global changes. All regions showed a decrease in CBF (Table 3), pointing to systemic etiologies, such as reduced inflammation and improvement of cerebral autoregulation. It is also possible that CBF decreased due to the vasoconstrictive effects of calcineurin inhibitors. Distinct from the KT-related change in CBF, we observed an age-related decrease in CBF in all participants, a well-established phenomenon seen in the general population.48

Similar to other published studies, we found that patients pre-KT had higher cerebral Cho/Cr and mI/Cr compared with controls.45678,9–11 Both Cho/Cr and mI/Cr decreased post-KT. Cho is a phospholipid cell membrane precursor and represents breakdown products, such as phosphocholine, phosphatidylcholine, glycerophosphorylcholine, and phosphorylcholine.22 Importantly, Cho is also a precursor of trimethylamine N-oxide, a marker of cardiovascular disease and mortality.49,50 mI, primarily found in glial cells, is involved in the phosphoinositide-mediated signal transduction and is a marker of inflammation or gliosis.22 High plasma mI is also associated with the high prevalence of peripheral polyneuropathy in ESKD.51 With their relative small sizes (mol wt 105 and 180 g/mol, respectively), both compounds are readily filtered by the glomeruli and eliminated with tubular secretion.52,53 Both plasma51,54 and intracerebral Cho and mI concentrations are elevated in CKD21 and decrease only slightly with dialysis.9 Cho and mI are both major cerebral osmolytes, and high concentrations can increase osmotic pressure in the brain, cause cellular edema, and alter cellular structure and function. Both compounds are transported from the plasma to the brain through the blood-brain barrier by diffusion and carrier mediated transporter systems.55,56 Because of the blood-brain barrier, changes in plasma concentrations of neurochemicals do not correlate with cerebral concentrations in healthy individuals.57 However, with the disruption of the blood-brain barrier in CKD, higher serum concentrations could theoretically affect cerebral neurochemical concentrations. Our study indicates normalization of Cho and mI post-KT, perhaps by better elimination with tubular secretion,52,53 which cannot be restored with dialysis or measured by creatinine clearance or serum creatinine–based eGFR. Improvement in these neurochemicals after KT could restore cerebral osmotic regulation, lower cellular edema, and improve cell function and cognition.

Similar to a prior cross-sectional study,9 we did not see any differences in the levels of Glx/Cr and NAA/Cr between controls and patients pre-KT, suggesting that CKD does not affect these neurochemicals as much as Cho and mI. Consistent with this hypothesis, there was no change in Glx/Cr and NAA/Cr pre- to post-KT. Glutamine is mainly found in glia and is a precursor to glutamate,22 the most abundant excitatory neurotransmitter in the neurons. NAA is also an osmolyte located in the cell bodies, axons, and dendrites of neurons and is a marker of neuronal number, density, and integrity.58 NAA/Cr is often reduced in neurodegenerative diseases and may not be affected in CKD.22

Similar to our prior study,31 KT resulted in an increase in FA and a decrease in MD. A recent study demonstrated similar findings.43 These changes in FA and MD indicate decreased free dispersion and greater axial movement of water molecules along the white matter tracts. A decrease in cerebral osmolytes and improvement in cerebral edema as described above can explain these DTI results. Pre-KT, even with maintenance dialysis, the clearance of substances that are eliminated by tubular secretion is reduced and can result in (often subclinical) cerebral edema. Although we have previously shown that patients on dialysis have lower FA and higher MD compared with healthy controls,59 we did not find FA and MD values to be significantly different in patients pre-KT and controls in this study. This could be because of the small sample size and single MRI measurements in controls leading to a larger SD.

The evidence that dialysis has short- and long-term deleterious effects on brain health is rapidly accumulating.4,5,6–15161718,19,20 Our MRSI and DTI results indicate normalization of cerebral edema post-KT, a finding that is not observed with dialysis. In fact, cerebral edema increases with longer dialysis vintage and higher ultrafiltration volumes.60 Similarly, change in CBF with dialysis can be detrimental.2 We demonstrate reversibility in abnormalities in CBF, neurochemical concentrations, and white matter integrity in ESKD. Because our study protocol included repeated pre-KT MRI in patients who did not receive KT within 1 year, we were able to compare pre-KT MRIs. We did not observe any change in brain abnormalities. Although we do not have the power to make definite conclusions in this study, our findings are consistent with our previous work.59 Normalized brain abnormalities even after 12 months post-KT indicate that the post-KT changes were not due to acute effects of KT, such as steroids, but a phenomenon that persists beyond the immediate post-KT period.

Despite the modest sample size, our longitudinal approach, linear mixed model analysis, and a uniform change in measures among patients strengthen our findings. For example, all patients had a decrease in CBF after KT. Unlike other studies, our CBF calculations were corrected for hematocrit, which affects blood viscosity.61 Because hematocrit changes following KT, correction for hematocrit is important for accurate CBF measurements in ESKD. Our models were adjusted for age,32 race,33 sex,34 and level of education.35,36 Age in particular is associated with changes in brain volume and function, and age-related changes observed over many years could be larger than changes observed in the 12 months post-KT. Another strength was the measurement of regional CBF, FA, and MD changes in addition to global changes.

In summary, abnormalities in CBF, neurochemical concentrations, and white matter integrity in CKD are normalized with KT. This reversibility in brain abnormalities has important implications in appreciating and managing the risk of dementia and stroke in our CKD population. More studies are needed to understand the mechanisms underlying these brain abnormalities, to understand the role of residual kidney function in preserving brain health on dialysis, and to explore innovations in RRTs to mitigate these abnormalities even in patients who cannot be transplanted.


A. Gupta has a consultancy agreement with Novartis Pharmaceuticals, has funding support from Novartis and Veloxis Pharmaceuticals, and is on the regional medical advisory board for the National Kidney Foundation, outside the submitted work. M. Sarnak is on the steering committee of a trial funded by Akebia, attended an advisory board for Bayer, and is a consultant for Cardurian, outside the submitted work. E. Vidoni is a stakeholder in a provisional patent (18KU028M-02) pending for measuring CBF. All remaining authors have nothing to disclose.


This work was supported by a gift from Forrest and Sally Hoglund (to Hoglund Biomedical Imaging Center), National Institutes of Health grant S10 RR29577 (to W. Brooks), National Institutes of Health grant K23 AG055666 (to A. Gupta), National Institutes of Health grant R21 AG061549 (to E. Vidoni), National Institutes of Health grant P30 AG035982 (to the University of Kansas Alzheimer’s Disease Center), National Institutes of Health Clinical and Translational Science Award grant UL1 TR002366 (to the University of Kansas Medical Center), National Institutes of Health grant P30 DK106912 (to the University of Kansas Medical Center Jared Grantham Kidney Institute), and the University of Kansas Medical Center Jared Grantham Kidney InstitutePilot Grant (to A. Gupta).

Published online ahead of print. Publication date available at

The authors acknowledge the support of the transplant coordinators and the kidney transplant team at the University of Kansas Hospital. W. Brooks, J. Burns, A. Gupta, and M. Sarnak designed the study; I.-Y. Choi, R. Lepping, and E. Vidoni analyzed the MRI data; J. Mahnken, R. Montgomery, and P. Sharma performed the statistical analysis; A. Gupta and R. Lepping drafted the manuscript; W. Brooks, J. Burns, J. Mahnken, R. Montgomery, M. Sarnak, and E. Vidoni reviewed and edited the manuscript; and all authors approved the final version of the manuscript.

Supplemental Material

This article contains the following supplemental material online at

Supplemental Figure 1. Flowchart of patients with ESKD undergoing MRI assessment at baseline, 3 months post-transplant, and 12 months post-transplant.

Supplemental Material. MRI data acquisition.

Supplemental Table 1. Characteristics of patients with and without a post-KT MRI.

Supplemental Table 2. Transplantation-related clinical characteristics of 22 patients with ESKD who received KT.

Supplemental Table 3. Unadjusted MRI measurements in controls and patients with ESKD.

Supplemental Table 4. Estimated effect of covariates for MRI variables in the mixed model analysis.

Supplemental Table 5. Comparison of repeated MRI measurements in patients pre-KT; no changes were observed in 1 year.


1. Murray AM, Seliger S, Lakshminarayan K, Herzog CA, Solid CA: Incidence of stroke before and after dialysis initiation in older patients. J Am Soc Nephrol 24: 1166–1173, 201323620399
2. Wachterman MW, O’Hare AM, Rahman O-K, Lorenz KA, Marcantonio ER, Alicante GK, et al.: One-year mortality after dialysis initiation among older adults. JAMA Intern Med 179: 987–990, 201931009039
3. Hsieh PC, Lin JH, Lee CL: Mixing of dissolved organic matter from distinct sources: Using fluorescent pyrene as a probe. J Environ Sci Health Part A Tox Hazard Subst Environ Eng 44: 170–178, 200919123097
4. Tryc AB, Alwan G, Bokemeyer M, Goldbecker A, Hecker H, Haubitz M, et al.: Cerebral metabolic alterations and cognitive dysfunction in chronic kidney disease. Nephrol Dial Transplant 26: 2635–2641, 201121216887
5. 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, 200717406658
6. Drew DA, Bhadelia R, Tighiouart H, Novak V, Scott TM, Lou KV, et al.: Anatomic brain disease in hemodialysis patients: A cross-sectional study. Am J Kidney Dis 61: 271–278, 201323040011
7. Ikram MA, Vernooij MW, Hofman A, Niessen WJ, van der Lugt A, Breteler MM: Kidney function is related to cerebral small vessel disease. Stroke 39: 55–61, 200818048865
8. Tamura MK, Pajewski NM, Bryan RN, Weiner DE, Diamond M, Van Buren P, et al.; SPRINT Study Research Group: Chronic kidney disease, cerebral blood flow, and white matter volume in hypertensive adults. Neurology 86: 1208–1216, 201626920359
9. Michaelis T, Videen JS, Linsey MS, Ross BD: Dialysis and transplantation affect cerebral abnormalities of end-stage renal disease. J Magn Reson Imaging 6: 341–347, 19969132100
10. Geissler A, Fründ R, Kohler S, Eichhorn HM, Krämer BK, Feuerbach S: Cerebral metabolite patterns in dialysis patients: Evaluation with H-1 MR spectroscopy. Radiology 194: 693–697, 19957862964
11. Menon DK, Sandford RN, Cassidy MJ, Bell JD, Baudouin CJ, Sargentoni J: Proton magnetic resonance spectroscopy in chronic renal failure. Lancet 337: 244–245, 19911670880
12. Hsieh TJ, Chang JM, Chuang HY, Ko CH, Hsieh ML, Liu GC, et al.: End-stage renal disease: In vivo diffusion-tensor imaging of silent white matter damage. Radiology 252: 518–525, 200919528357
13. Eldehni MT, Odudu A, McIntyre CW: Randomized clinical trial of dialysate cooling and effects on brain white matter. J Am Soc Nephrol 26: 957–965, 201525234925
14. Chou MC, Hsieh TJ, Lin YL, Hsieh YT, Li WZ, Chang JM, et al.: Widespread white matter alterations in patients with end-stage renal disease: A voxelwise diffusion tensor imaging study. AJNR Am J Neuroradiol 34: 1945–1951, 201323788598
15. Zhang R, Liu K, Yang L, Zhou T, Qian S, Li B, et al.: Reduced white matter integrity and cognitive deficits in maintenance hemodialysis ESRD patients: A diffusion-tensor study. Eur Radiol 25: 661–668, 201525326436
16. Lau WL, Nunes ACF, Vasilevko V, Floriolli D, Lertpanit L, Savoj J, et al.: Chronic kidney disease increases cerebral microbleeds in mouse and man. Transl Stroke Res 11: 122–134, 202031055735
17. Lau WL, Huisa BN, Fisher M: The cerebrovascular-chronic kidney disease connection: Perspectives and mechanisms. Transl Stroke Res 8: 67–76, 201727628245
18. 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, 19947905681
19. Polinder-Bos HA, García DV, Kuipers J, Elting JWJ, Aries MJH, Krijnen WP, et al.: Hemodialysis induces an acute decline in cerebral blood flow in elderly patients. J Am Soc Nephrol 29: 1317–1325, 201829496888
20. Findlay MD, Dawson J, Dickie DA, Forbes KP, McGlynn D, Quinn T, et al.: Investigating the relationship between cerebral blood flow and cognitive function in hemodialysis patients. J Am Soc Nephrol 30: 147–158, 201930530658
21. Sasaki O, Hattori N, Nakahama H, Inoue N, Nakamura S, Inenaga T, et al.: Positive correlations between cerebral choline and renal dysfunction in chronic renal failure. Neuroradiology 48: 300–306, 200616586119
22. Su L, Blamire AM, Watson R, He J, Hayes L, O’Brien JT: Whole-brain patterns of (1)H-magnetic resonance spectroscopy imaging in Alzheimer’s disease and dementia with Lewy bodies. Transl Psychiatry 6: e877, 201627576166
23. Sotak CH: The role of diffusion tensor imaging in the evaluation of ischemic brain injury - a review. NMR Biomed 15: 561–569, 200212489102
24. Mayo CD, Garcia-Barrera MA, Mazerolle EL, Ritchie LJ, Fisk JD, Gawryluk JR; Alzheimer’s Disease Neuroimaging Initiative: Relationship between DTI metrics and cognitive function in Alzheimer’s disease. Front Aging Neurosci 10: 436, 201930687081
25. Ferguson KJ, MacLullich AM, Marshall I, Deary IJ, Starr JM, Seckl JR, et al.: Magnetic resonance spectroscopy and cognitive function in healthy elderly men. Brain 125: 2743–2749, 200212429601
26. Murray AM, Tupper DE, Knopman DS, Gilbertson DT, Pederson SL, Li S, et al.: Cognitive impairment in hemodialysis patients is common. Neurology 67: 216–223, 200616864811
27. Friedman SD, Brooks WM, Jung RE, Hart BL, Yeo RA: Proton MR spectroscopic findings correspond to neuropsychological function in traumatic brain injury. AJNR Am J Neuroradiol 19: 1879–1885, 19989874540
28. Kraus MF, Susmaras T, Caughlin BP, Walker CJ, Sweeney JA, Little DM: White matter integrity and cognition in chronic traumatic brain injury: A diffusion tensor imaging study. Brain 130: 2508–2519, 200717872928
29. Mrak RE, Griffin ST, Graham DI: Aging-associated changes in human brain. J Neuropathol Exp Neurol 56: 1269–1275, 19979413275
30. De Deyn PP, Goeman J, Engelborghs S, Hauben U, D’Hooge R, Baro F, et al.: From neuronal and vascular impairment to dementia. Pharmacopsychiatry 32[Suppl 1]: 17–24, 199910338104
31. Gupta A, Lepping RJ, Yu AS, Perea RD, Honea RA, Johnson DK, et al.: Cognitive function and white matter changes associated with renal transplantation. Am J Nephrol 43: 50–57, 201626894920
32. Amen DG, Egan S, Meysami S, Raji CA, George N: Patterns of regional cerebral blood flow as a function of age throughout the lifespan. J Alzheimers Dis 65: 1087–1092, 201830103336
33. Ku E, Yang W, McCulloch CE, Feldman HI, Go AS, Lash J, et al.; CRIC Study Investigators: Race and mortality in CKD and dialysis: Findings from the chronic renal insufficiency cohort (CRIC) study. Am J Kidney Dis 75: 394–403, 202031732235
34. Nitsch D, Grams M, Sang Y, Black C, Cirillo M, Djurdjev O, et al.; Chronic Kidney Disease Prognosis Consortium: Associations of estimated glomerular filtration rate and albuminuria with mortality and renal failure by sex: A meta-analysis. BMJ 346: f324, 201323360717
35. Mortel KF, Meyer JS, Herod B, Thornby J: Education and occupation as risk factors for dementias of the Alzheimer and ischemic vascular types. Dementia 6: 55–62, 19957728220
36. Ostojic J, Kozic D, Ostojic SM: N-acetylaspartate-to-creatine ratio in twelve brain locations among healthy men and women with different levels of education. Neurosci Lett 692: 23–26, 201930367956
37. Cnaan A, Laird NM, Slasor P: Using the general linear mixed model to analyse unbalanced repeated measures and longitudinal data. Stat Med 16: 2349–2380, 19979351170
38. Gueorguieva R, Krystal JH: Move over ANOVA: Progress in analyzing repeated-measures data and its reflection in papers published in the archives of general psychiatry. Arch Gen Psychiatry 61: 310–317, 200414993119
39. Gupta A, Mahnken JD, Johnson DK, Thomas TS, Subramaniam D, Polshak T, et al.: Prevalence and correlates of cognitive impairment in kidney transplant recipients. BMC Nephrol 18: 158, 201728499360
40. Gupta A, Montgomery RN, Bedros V, Lesko J, Mahnken JD, Chakraborty S, et al.: Subclinical cognitive impairment and listing for kidney transplantation. Clin J Am Soc Nephrol 14: 567–575, 201930890576
41. Liu H-S, Hartung EA, Jawad AF, Ware JB, Laney N, Port AM, et al.: Regional cerebral blood flow in children and young adults with chronic kidney disease. Radiology 288: 849–858, 201829893643
42. Kamano C, Komaba Y, Sakayori O, Iino Y, Katayama Y: Decreased cerebral blood flow in renal transplant recipients. Intern Med 41: 677–683, 200212322791
43. Sandwijk MSv, Berge IJMt, Caan MWA, Düring M, Gool WAv, Majoie CBLM, et al.: Cognitive improvement after kidney transplantation is associated with structural and functional changes on MRI. Transplant Direct 6: e531, 2020
44. Sedaghat S, Vernooij MW, Loehrer E, Mattace-Raso FU, Hofman A, van der Lugt A, et al.: Kidney function and cerebral blood flow: The Rotterdam study. J Am Soc Nephrol 27: 715–721, 201626251352
45. Fazekas G, Fazekas F, Schmidt R, Flooh E, Valetitsch H, Kapeller P, et al.: Pattern of cerebral blood flow and cognition in patients undergoing chronic haemodialysis treatment. Nucl Med Commun 17: 603–608, 19968843120
46. Wintermark M, Sesay M, Barbier E, Borbély K, Dillon WP, Eastwood JD, et al.: Comparative overview of brain perfusion imaging techniques. J Neuroradiol 32: 294–314, 200516424829
47. Davenport A, Buscombe JR: Watershed cerebral infarction in a hemodialysis patient. Kidney Int 77: 1140, 201020508665
48. Gur RC, Gur RE, Obrist WD, Skolnick BE, Reivich M: Age and regional cerebral blood flow at rest and during cognitive activity. Arch Gen Psychiatry 44: 617–621, 19873606327
49. Stubbs JR, House JA, Ocque AJ, Zhang S, Johnson C, Kimber C, et al.: Serum trimethylamine-N-oxide is elevated in CKD and correlates with coronary atherosclerosis burden. J Am Soc Nephrol 27: 305–313, 201626229137
50. Heianza Y, Ma W, Manson JE, Rexrode KM, Qi L: Gut microbiota metabolites and risk of major adverse cardiovascular disease events and death: A systematic review and meta-analysis of prospective studies. J Am Heart Assoc 6: e004947, 201728663251
51. Clements RS Jr., DeJesus PV Jr., Winegrad AI: Raised plasma-myoinositol levels in uraemia and experimental neuropathy. Lancet 1: 1137–1141, 19734123536
52. Acara M, Roch-Ramel F, Rennick B: Bidirectional renal tubular transport of free choline: A micropuncture study. Am J Physiol 236: F112–F118, 1979420292
53. Silbernagl S, Völker K, Dantzler WH: Tubular reabsorption of myo-inositol vs. that of D-glucose in rat kidney in vivo et situ. Am J Physiol Renal Physiol 284: F1181–F1189, 200312736166
54. Melmed S, Lewin LM, Bank H: Myo-inositol clearance in renal failure and in patients with normal kidney function. Am J Med Sci 274: 55–59, 1977900156
55. Spector R: Myo-inositol transport through the blood-brain barrier. Neurochem Res 13: 785–787, 19883173626
56. Allen DD, Lockman PR: The blood-brain barrier choline transporter as a brain drug delivery vector. Life Sci 73: 1609–1615, 200312875893
57. Takado Y, Sato N, Kanbe Y, Tomiyasu M, Xin L, Near J, et al.: Association between brain and plasma glutamine levels in healthy young subjects investigated by MRS and LC/MS. Nutrients 11: 1649, 201931330962
58. Benarroch EE: N-acetylaspartate and N-acetylaspartylglutamate: Neurobiology and clinical significance. Neurology 70: 1353–1357, 200818413589
59. Drew DA, Koo B-B, Bhadelia R, Weiner DE, Duncan S, la Garza MM-D, et al.: White matter damage in maintenance hemodialysis patients: A diffusion tensor imaging study. BMC Nephrol 18: 213, 201728676035
60. Reetz K, Abbas Z, Costa AS, Gras V, Tiffin-Richards F, Mirzazade S, et al.: Increased cerebral water content in hemodialysis patients. PLoS One 10: e0122188, 201525826269
61. Lu H, Clingman C, Golay X, van Zijl PC: Determining the longitudinal relaxation time (T1) of blood at 3.0 Tesla. Magn Reson Med 52: 679–682, 200415334591

ESKD; kidney transplantation; cerebral blood flow; neurochemicals; white matter integrity; ASL; MRSI; DTI

Copyright © 2021 by the American Society of Nephrology