Decreased white matter lesion volume and improved cognitive function after liver transplantation


  • Alex Rovira,

    Corresponding author
    1. Unitat de Ressonància Magnètica (I.D.I), Departament de Radiologia, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    • Unitat de Ressonància Magnètica, Servei de Radiologia, Hospital General Universitari Vall d'Hebron, Pg. Vall d'Hebron 119-129, Barcelona 08035, Spain===

    Search for more papers by this author
    • fax: )34( 934286059

  • Beatriz Mínguez,

    1. Servei de Medicina Interna-Hepatologia, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    2. Departament de Medicina. Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain
    3. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), Spain
    Search for more papers by this author
  • F. Xavier Aymerich,

    1. Unitat de Ressonància Magnètica (I.D.I), Departament de Radiologia, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    2. Departament d'Enginyeria de Sistemes, Automàtica i Informàtica Industrial (ESAII), Universitat Politècnica de Catalunya, Barcelona, Spain
    Search for more papers by this author
  • Carlos Jacas,

    1. Departament de Medicina. Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain
    2. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), Spain
    3. Unitat de Neuropsicología, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    Search for more papers by this author
  • Elena Huerga,

    1. Unitat de Ressonància Magnètica (I.D.I), Departament de Radiologia, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    Search for more papers by this author
  • Juan Córdoba,

    1. Servei de Medicina Interna-Hepatologia, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    2. Departament de Medicina. Universitat Autònoma de Barcelona, Cerdanyola del Vallès, Spain
    3. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), Spain
    Search for more papers by this author
  • Juli Alonso

    1. Unitat de Ressonància Magnètica (I.D.I), Departament de Radiologia, Hospital Universitari Vall d'Hebron, Barcelona, Spain
    2. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas (CIBEREHD), Spain
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.


Focal T2-weighted white matter lesions (WML) on brain magnetic resonance imaging (MRI), mimicking those seen in cerebrovascular small-vessel disease described in patients with persistent hepatic encephalopathy, decreased in volume with the improvement of hepatic encephalopathy. This outcome has been interpreted as a decrease in the edema that it is proposed to be involved in the pathogenesis of hepatic encephalopathy. We designed a study to further investigate potential changes in focal WML in the brains of patients with cirrhosis following liver transplantation and to study the relationship between these changes and overall cognitive function. We used MRI to measure the volume of supratentorial focal WML and a neuropsychological examination to assess cognitive function before and after liver transplantation in 27 patients with cirrhosis without signs of overt hepatic encephalopathy. Baseline MRI identified focal T2-weighted lesions in 19 patients (70.3%). The presence of WML was associated with older age but not with vascular risk factors, severity of liver function, or psychometric tests. A significant reduction in lesion volume was observed after liver transplantation (from a median of 1.306 cm3 to 0.671 cm3, P = 0.001). This decrease correlated with an improvement in an index of global cognitive function (r = −0.663; P < 0.001). This evolution indicates that lesion volume is partially related to a reversible type of tissue damage, which is compatible with brain edema. Conclusion: Focal WML probably induced by age-related microvascular injury can decrease their volume with liver transplantation. The associated improvement of cognitive function supports a relationship between brain edema and minimal hepatic encephalopathy. (HEPATOLOGY 2007.)

Focal white matter lesions (WML) secondary to cerebrovascular small-vessel disease are commonly found in the general population over 60 years of age.1 In a previous report, we described focal brain WML on T2-weighted images in patients with hepatic encephalopathy (HE).2 This abnormality, which is radiologically indistinguishable from the features of small-vessel disease or normal aging, was partially reversible with improvements in HE. This evolution pattern contrasts with what occurs in focal WML attributable to small-vessel disease or brain aging, which remain stable or progress in number and size over time, but never decrease.3–5 The most plausible explanation for decreasing the volume of WML with improvement of HE is a decrease in the edema component of WML.

There is a large body of evidence indicating an increase in the amount of water in the brain of patients with liver cirrhosis that is related to the development of HE.6 The most accepted hypothesis for this low-grade brain edema is that astrocytic accumulation of glutamine induces edema as a result of ammonia detoxification. This hypothesis is supported by experimental data showing astrocyte swelling and changes in brain organic osmolytes in models of HE,7 and by brain proton magnetic resonance spectroscopy studies in patients with liver cirrhosis, which consistently show increases in the glutamine/glutamate signal accompanied by myoinositol depletion,8 the latter being part of an osmoregulatory mechanism that prevents massive brain edema in patients with chronic liver failure.

The presence of mild diffuse edema within the brain tissue in patients with cirrhosis is also supported by the findings of several magnetic resonance imaging (MRI) techniques such as magnetization transfer ratio measurements, which show low values in otherwise normal-appearing white matter, and fast- fluid attenuated inversion recovery (FLAIR) T2-weighted images, which show diffuse high-signal intensity along the brain hemispheric white matter in or around the corticospinal tract. Both abnormalities return to nearly normal with restoration of liver function, supporting the hypothesis that they reflect mild diffuse brain edema.9–11

Disturbances in cell volume homeostasis may participate in the pathogenesis of minimal HE, which is a cognitive disturbance that is common among patients with cirrhosis that undergo liver transplantation.10 Because focal WML are not uncommon in patients over 60 years of age,1 we expect to find them in a significant proportion of patients undergoing liver transplantation. We hypothesized that, if present, changes in their volume after liver transplantation could indicate the presence of brain edema before liver transplantation. All this led us to perform a systematic assessment of the presence of focal T2 WML in a group of patients with cirrhosis without overt HE who underwent liver transplantation, with the aim of analyzing their evolution after normalization of liver function and studying their association with cognitive function.


FLAIR, fluid attenuated inversion recovery; HE, hepatic encephalopathy; MRI, magnetic resonance imaging; SD, standard deviation; T2LL, T2 lesion load; WML: white matter lesion.

Patients and Methods


The study includes a consecutive series of patients with cirrhosis with no clinical evidence of overt HE, who were evaluated for liver transplantation at our institution following standard procedures. At that time, all patients were informed that the study was designed to evaluate manifestations of minimal HE and that they would complete psychometric tests and undergo brain MRI before, and 6 to 14 months after, liver transplantation. Patients with a history of drug abuse, those affected by neurologic or psychiatric diseases, and those receiving medication known to have significant effects on the central nervous system were excluded. The diagnosis of cirrhosis was based on a consistent clinical history, radiological studies, and liver biopsy when available.

Thirty patients were initially included in the study, although 3 of them could not be reassessed after liver transplantation because of death or severe medical complications. Therefore, the final cohort was composed of 27 patients, 23 men and 4 women, with a mean age of 55 years (range, 30-68 years). All of these patients underwent neurologic assessment and brain MRI on the same day. The clinical characteristics of the patients included in the study are described in Table 1. The study was approved by the Institutional Review Board of Hospital Universitari Vall d'Hebron, and all subjects gave written consent for participation.

Table 1. Clinical Characteristics of Patients Classified According to the Presence of Focal White Matter Lesions Before Liver Transplantation
CharacteristicFocal white matter lesions
  • *

    P ≤ 0.01. Abbreviation: WML, white matter lesions.

Number of patients819
Age (years)46 ± 1160 ± 9*
Vascular risk factors  
 Arterial hypertension02
 Diabetes mellitus24
 Child Pugh Score10.0 ± 1.68.3 ± 1.4*
 Viral hepatitis012*
 Viral hepatitis + alcohol33
Prior decompensation  
 Hepatic encephalopathy28
 Variceal bleeding47
 Month since first decompensation36 ± 2949 ± 51
Drug therapy  
Laboratory tests  
 ALT (IU/L)46 ± 4576 ± 49
 Bilirubin (mg/dL)4.5 ± 2.82.5 ± 1.7
 Albumin (mg/dL)2.5 ± 2.62.7 ± 0.7
 Prothrombin activity (%)44 ± 1464 ± 15*
 Creatinine (mg/dL)0.8 ± 0.11.1 ± 0.6

Neurologic Assessment.

None of the patients showed signs of overt HE; they were perfectly alert, without flapping tremor, and were oriented in space, person, and time. Results of the neurologic examination were considered normal in all cases. The patients completed a battery of neuropsychological tests designed to detect the abnormalities that are most frequent in minimal HE and to give an overall appraisal of neuropsychological function12: Stroop test, Trail Making test (part A), Symbol Digits (oral version), Grooved Pegboard test (dominant and nondominant hand), Auditory Verbal Learning, Judgment of Line Orientation, Hooper Test of Visual Organization, and Controlled Oral Word Association test. The scores of each test were transformed into T values with the aid of metanorms adjusted by age, sex, and years of education. T values were calculated according to the formula T = 50 +10 ([x–xn]/SDn), where x represents the raw result of the test, xn represents the mean value, and SDn the standard deviation value for the test in the normal population.13 Scores on the different tests were grouped into indexes of memory (Auditory Verbal Learning test), attention (Trail A, Symbol Digit), executive function (Controlled Oral Word Association test), psychomotor function (Grooved Pegboard), and visual perceptive function (Judgment of Line Orientation, Hooper Test). An overall score was calculated as the average of all the tests. Impairment was defined as a T-value of 30–40 (mild), 20–29 (moderate), and <20 (severe), according to standard procedures.13


Brain MRI was obtained prior to liver transplantation [mean, 50 days; standard deviation (SD), 36 days; range, 3–131 days), and repeated after liver transplantation in all patients (mean, 283 days; SD, 89 days; range, 177–404 days) with a 1.5-T Magneton Vision-plus superconductive magnet (Siemens, Erlangen, Germany) using a quadrature transmit/receive head coil. The MRI protocol included the following pulse sequences: transverse T2-weighted fast spin-echo (3000/85/2=TR/TE/acquisitions) and fast-FLAIR (9900/110/2500/1=TR/TE/inversion time/acquisitions) and T1-weighted spin-echo (600/15/2). The sequences were registered using the following parameters: section thickness 5 mm, interleaved imaging mode, intersection gap 1.5 mm, pixel size approximately 1 × 1 mm, and acquisition matrix 256 × 256 mm.

All fast-FLAIR images obtained in the baseline and follow-up scans were transferred to a Sun Ultra 10 Workstation (Sun Microsystems Inc., Palo Alto, CA) to calculate T2 lesion load (T2LL). Baseline lesions were initially identified on all available sequences, and then marked on the fast-FLAIR hard copies by the same neuroradiologist. Only focal WML located in the brain hemispheres and at least 3 mm in size were considered for the volume measurement. A single rater previously trained to ensure a high level of reproducibility performed the T2LL analysis as follows: (1) Hard copies with all lesions outlined were placed alongside the identical computer-generated image; (2) All lesions marked on the hard copies were outlined on the computer image using a semiautomatic local thresholding contour technique (DISPImage program, Dave Plummer, London, UK). If the lesion could not be outlined satisfactorily with this approach, outlining was done manually; and (3) A computer program added up the individual lesion volumes, and the final T2LL was obtained. Follow-up scans were assessed by the same rater, taking into account that only lesions identified on the baseline MRI scan could be considered in the analysis. The rater was not blinded to the date of the examination, but she was not aware of the aim of the study or the clinical and neuropsychological condition of the patients.

Statistical Analysis.

The Sigma Stat statistical package (v 3.0) was used for statistical calculations. Results are expressed as mean ± standard deviation, if not expressed otherwise. Depending on the behavior of the variables, parametric or nonparametric tests were applied to study the differences between groups of patients (Student t test or the Mann-Whitney rank sum test) and to study the intrasubject differences before and after liver transplantation (paired Student t test or Wilcoxon signed-rank test). A chi-square test or Fisher's exact test was used to study significant differences between nominal variables. Finally, Pearson's correlation coefficient was applied to study correlations among variables.


Baseline MR imaging identified focal T2-weighted WML, of the type commonly seen in patients with small-vessel disease, in 19 patients (70.3%) (Fig. 1). These patients were significantly older than the group without lesions but exhibited better hepatic function according to Child-Pugh score and prothrombin activity (Table 1). There was no association between vascular risk factors and the presence of focal WML. In patients with focal WML, cirrhosis was most commonly attributable to viral hepatitis (hepatitis C, 10; hepatitis B, 1; hepatitis B and C, 1).

Figure 1.

(A) Baseline MRI study (transverse fast-FLAIR T2-weighted imaging) in a 56-year-old patient with hepatitis C cirrhosis before liver transplantation. Focal lesions are scattered in the subcortical white matter (arrows). (B) The same MRI study 6 months after successful liver transplantation. There is a decrease in the size of focal white matter lesions.

In patients with WML, median lesion volume was 1.306 cm3 (range, 0.198-10.991 cm3). There was no association between the volume of focal WMLs and age, cause of cirrhosis, Child-Pugh score, or laboratory findings. Following liver transplantation, focal WMLs decreased to a median volume of 0.671 cm3 (range, 0.155–9.769 cm3; P > 0.0001), which represents a 21.7% decrease on average. Figure 2 shows T2LL before and after liver transplantation for each patient. It should be noted that a slight increment in T2LL occurred in 2 patients, from 0.210 to 0.378 cm3 in 1 and from 0.939 to 1.070 cm3 in the other. At the time of the second MRI scan, liver function was considered good, as corroborated by laboratory findings [alanine aminotransferase (ALT), 74 ± 63 IU/L; bilirubin, 0.9 ± 0.5 mg/dL; albumin, 4 ± 0.4 mg/dL; and prothrombin, 95% ± 14%]. All patients were receiving immunosuppression with tacrolimus, combined with steroids in 4 cases and mycophenolate in 3 cases. Ten patients developed diabetes mellitus and 8 patients arterial hypertension, which were likely caused by the immunosuppressor medication. Similarly, there was mild deterioration of renal function (creatinine, 1.3 ± 0.2), which can be attributed to tacrolimus.

Figure 2.

Changes in volume of focal white matter lesions after liver transplantation. Each line represents 1 patient before (pre-liver Tx) and after liver transplantation (post-liver Tx). There is a significant decrease in median volume from 1.306 cm3 to 0.671 cm3 (−21.7%) (P < 0.001).

Neuropsychological tests before liver transplantation showed impairment in several domains. The most highly affected were attention and motor function. Attention was impaired in 13 patients (mild, 8; moderate, 4; severe, 1) and motor function in 14 patients (mild, 7; moderate, 6; severe, 1). There was a significant improvement in neuropsychological function following liver transplantation (Table 2). Before transplantation, the overall score was abnormal in 9 patients (mild, 7; moderate, 2) and following liver transplantation only 1 patient exhibited mild impairment. Neuropsychological indexes before liver transplantation did not differ between the group of patients with focal WML (overall score, 45.1 ± 7.4) and the group without WMLs (overall score, 43.5 ± 8.5). However, following liver transplantation, the percent decrease in focal WML volume showed a positive correlation with the percent improvement in overall cognitive function (R = −0.663, P < 0.001, Fig. 3).

Table 2. Neuropsychological Tests in Patients with Cirrhosis Before and After Liver Transplantation
TestLiver transplantation
  • *

    P < 0.01, results expressed in t-values.

Attention38.4 ± 9.544.8 ± 7.0*
Motor38.9 ± 12.044.5 ± 11.1*
Visual perceptive49.7 ± 8.753.4 ± 6.6*
Executive function49.7 ± 11.853.9 ± 7.9
Memory46.4 ± 9.351.3 ± 9.4*
Global44.6 ± 7.749.9 ± 5.7*
Figure 3.

Correlation between changes in the global neuropsychological score and volume of focal white matter lesions (r = −0.663; P < 0.001). Both indexes are expressed as percentage of change compared with baseline [(posttransplantation − pretransplantation)/pretransplantation].


The current study shows that focal brain T2-weighted WML present in patients with liver cirrhosis can decrease in volume within a few months after liver transplantation. These lesions resemble those commonly seen in patients with several types of small-vessel cerebrovascular disease (atherosclerosis, cerebral amyloid angiopathy, or cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy) and in the normal elderly population.1, 14, 15 These focal WML are strongly associated with advanced age, high blood pressure, diabetes, stroke, and myocardial infarction.5, 14–18 The neuropathologic substrate corresponds to areas of demyelination and gliosis as well as silent infarcts.14, 19 The lesions are considered ischemic in nature and caused by hypoperfusion in the distal deep arterial or arteriolar territories. Multiple clinical studies have found that WML are irreversible and relate to cognitive impairment, dementia, depression, and gait disturbances.17, 20–22

The aim of this study was not to investigate the prevalence of incidental focal WML in patients with cirrhosis, but the 70% prevalence in a group of relatively young subjects (55 ± 11 years), with no history of stroke and a low proportion of arterial hypertension (7%), is higher than would be expected.16, 17 Other factors in addition to small-vessel disease may be responsible for their presence. Liver failure in itself or its duration do not seem to be responsible for these WML. In our study, the Child-Pugh Score was higher in the group with no WML, and there was no relationship with the time since the first decompensation. The association with viral hepatitis is intriguing, but may be explained by the lower age of patients with cirrhosis of other origins (mainly alcoholic).

Because of their location and small size, WML in patients with cirrhosis often may be ignored or interpreted as normal involutive or chronic ischemic changes. In a previous report, we had the opportunity to assess 3 patients with vascular risk factors and persistent HE with MRI. These patients exhibited extensive focal WML that decreased in volume with improvement of HE.2 This improvement was achieved several weeks after the administration of medical treatment that included neomycin and branched chain amino acids. Similar MRI findings, although not so extensive, were present in our patients with liver cirrhosis but no overt HE. The volume of these WML significantly decreased following liver transplantation and closely correlated with the improvement in neuropsychological function. All these findings support the existence of a relationship between a reversible component of WML and HE.

The partial reversibility of focal WML found in this study contrasts with what occurs in patients with small-vessel cerebrovascular disease. WML may increase over time, particularly in patients with hypertension and diabetes, but they never decrease3–5 and therefore are permanently visible on MRI, indicating that they represent irreversible tissue damage. The decrease in the volume of WML observed in patients with cirrhosis can be explained by changes in the amount of brain edema. Reversible diffuse MRI changes, such as decreases in magnetization transfer ratio and high-signal white matter intensity on fast-FLAIR images have been described in patients with minimal HE.10, 11 The most accepted hypothesis for these MRI findings is the presence of mild diffuse brain edema associated with increased brain glutamine as a result of ammonia detoxification, which induces astrocytic swelling.23 Liver failure may be responsible for low-grade brain edema, which could be exacerbated in areas of small vessel disease and reverse after liver transplantation.

Magnetization transfer ratio and fast-FLAIR sequences are highly sensitive for detecting changes in brain tissue water.24, 25 However, these techniques cannot distinguish whether the increase is intracellular or extracellular. This issue can be resolved by diffusion-weighted MR imaging, which can potentially locate the compartment where the water increase is more prominent.26 A cross-sectional diffusion-weighted MRI study obtained in patients with cirrhosis showed a significant increase in brain water diffusivity, with highest diffusivity in patients with the most severe encephalopathy.27 According to the basic understanding of diffusivity,26 these findings likely reflect water accumulation in the extracellular compartment. A plausible explanation for the development of diffuse extracellular edema is increased blood–brain barrier permeability in particularly vulnerable areas, probably induced by age-related microvascular changes. In fact, hyperammonemia may induce an increase in blood–brain barrier permeability,28, 29 which would favor capillary water influx to the brain. Moreover, Alzheimer type II astrocytes, a common abnormality in HE that has been interpreted as a sign of astrocyte swelling, may indeed correspond to cellular damage and loss of cellular shape caused by oxidative stress, rather than intracellular edema.30

A recent hypothesis has proposed that in patients with small-vessel disease, leakage of toxic plasma components through an impaired blood–brain barrier at the level of disturbed microvasculature causes secondary interstitial edema and injury to axons, myelin, and glial cells. This injury can lead to neuropsychological decline and eventually result in dementia.31 Regardless of the potential causative mechanisms shared by focal WML observed in cerebrovascular small-vessel disease and in HE, both situations have been associated with a similar type of cognitive impairment. In fact, the typical pattern of cognitive impairment in minimal HE includes disturbances of attention and executive functioning with preserved memory,32 a pattern that is typical of vascular cognitive impairment.33

A possible source of uncertainty is the effect of immunosuppressor drugs. Several studies have shown that these drugs may affect brain function, but contrary to what we found their effect is to worsen neurologic function and induce white matter lesions.34 Neurotoxic effects of calcineurin inhibitors are understood as secondary to vasogenic brain edema caused by an increase in blood-brain barrier permeability.35 Prior magnetic resonance studies in recipients of kidney or heart transplants have described toxic lesions or infectious complications, but there is no documentation of changes in the size of small-vessel WML.36 Thus, there is no reason to suggest that the decrease in WML and the improvement of cognitive function could have been caused by immunosuppressor drugs.

In conclusion, focal WML present in the brain of patients with cirrhosis that resemble those seen in small-vessel cerebrovascular disease can decrease in volume following liver transplantation. This observation indicates that these lesions are partially related to a reversible component of tissue damage, most probably brain edema. Assessment of changes in the volume of focal WML may be useful for understanding the pathogenesis of HE, for adequate interpretation of cognitive impairment in patients with cirrhosis, and for assessing the effects of therapeutic measures focused on correcting this disorder.


We thank Celine L. Cavallo for English language support.