Article Text

Short report
Executive dysfunction in adults with moyamoya disease is associated with increased diffusion in frontal white matter
  1. Lionel Calviere1,
  2. Guillaume Ssi Yan Kai2,
  3. Isabelle Catalaa2,
  4. Fabienne Marlats1,
  5. Fabrice Bonneville2,
  6. V Larrue1
  1. 1Department of Vascular Neurology, University of Toulouse, France
  2. 2Department of Neuroradiology, University of Toulouse, France
  1. Correspondence to Dr Lionel Calviere, Department of Vascular Neurology, University of Toulouse, Toulouse 31059, France; liocalviere{at}


Background and purpose Alteration of the cerebrovascular reserve (CVR) in the frontal lobes has been associated with cognitive dysfunction in adults with moyamoya disease (MMD). Elevation of the apparent diffusion coefficient (ADC) in normal-appearing white matter on conventional MRI may occur as a consequence of chronic haemodynamic failure. In the present study, the authors examined the relation of ADC with CVR and cognitive dysfunction in adults with MMD.

Methods The authors measured ADC and CVR in the normal-appearing frontal white matter. CVR was calculated using dynamic susceptibility contrast-enhanced MRI and the acetazolamide challenge. A standardised and validated neuropsychological assessment test battery focusing on executive function was used.

Results 14 patients, 9 women and 5 men (mean age 36.6±12.9 years), were included. The authors found executive dysfunction in 7 of 13 tested patients. ADC and CVR were negatively correlated (Spearman coefficient: −0.46; p=0.015). Elevation of ADC predicted executive dysfunction (area under receiver operating characteristic curve (95% CI): 0.85 (0.59 to 1.16); p=0.032).

Conclusion Elevation of ADC in the normal-appearing frontal white matter of adults with MMD was associated with reduced CVR and executive dysfunction. This preliminary study suggests that measurement of ADC might be used to detect patients at risk for cerebral ischaemia and cognitive impairment.

  • Diffusion
  • dysexecutive syndrome
  • moyamoya disease
  • cerebrovascular disease
  • subarachnoid haemorrhage
  • stroke
  • ultrasound
  • cerebral blood flow
  • neuroradiology

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Moyamoya disease (MMD) is characterised by the progressive stenosis and occlusion of the intracranial carotid artery and its main branches associated with the development of a basal collateral network.1 The impairment in cerebral perfusion resulting from occlusive arterial lesions can lead to ischaemic stroke and cognitive dysfunction. Cognitive dysfunction is common in adult patients with MMD.2–4 It affects predominantly the executive function and has been associated with reduced cerebrovascular reserve (CVR) in frontal areas as measured with perfusion MRI and the acetazolamide challenge.4

Measurement of CVR is, however, a time-consuming technique and has not been standardised across centres. The apparent diffusion coefficient (ADC) calculated using diffusion-weighted images (DWIs) can be used to detect subtle structural changes in the normal-appearing cerebral white matter on T2 and fluid attenuated inversion recuperation (FLAIR) sequences.5 6 In a recent study, regions of impaired CVR were spatially correlated with elevated ADC in otherwise normal-appearing white matter of patients with MMD.7

In the present study, we measured ADC in the normal-appearing frontal white matter of adult patients with MMD and examined the relation of this parameter with CVR and cognitive dysfunction.

Patients and methods

We identified adult patients with MMD followed at our institution and invited them to participate in this prospective study. The study was approved by our medical institutional review board and local ethics committee. All patients gave written informed consent.

MMD was diagnosed using established criteria.8 Diagnosis was based on digital subtraction angiography (12 patients) or magnetic resonance angiography (two patients).

Exclusion criteria were age <18 years and any associated disease potentially responsible for the arterial lesions.9 Patients were included at least 3 months after a stroke and before any surgical procedure.

Neuropsychological assessment

We used a standardised and validated neuropsychological assessment test battery focusing on executive function.4 The individual raw scores were converted into Z-scores adjusted for age, gender and education level according to published normative data. The results were expressed in SD. Impairment was defined as a Z score ≥1.7 SDs below the normative mean.4 Executive dysfunction was defined by impairment in three or more tests of executive function or attention.4

Measurement of ADC

MRI studies were performed on a 1.5-T scanner (Gyroscan NT Intera; Philips, Best, The Netherlands).

DWIs were acquired using a spin-echo single-shot echo-planar sequence with the following parameters: temps d'echo (TE) ‘30 ms’, field of view (FOV) ‘69’, matrix, ‘128×128’, flip angle ‘40°’, 7 mm thick contiguous sections, number of excitations (NEX) ‘2’, b=0–1000 mm2/s. DWIs were transferred to a separate workstation for data analysis. In each hemisphere, ADC was calculated from an average of two regions of interest (ROIs) located in the normal-appearing frontal white matter (centrum semiovale or corona radiata). Each ROI consisted of a 1-cm diameter circle. We carefully avoided areas with infarction or leukoaraiosis identified on T2- and FLAIR-weighted images (figure 1 in online supplementary material). All ROIs were placed by the same neuroradiologist (GSYK) blinded to clinical data.

Perfusion MRI

We measured cerebral perfusion using dynamic susceptibility contrast-enhanced MRI. The perfusion MRI method with measurement of relative regional CVR using cerebellum as a reference region has been previously described in detail.4 Briefly, we analysed data without deconvolution by using summary parameters obtained directly from the concentration-time curve. The use of an arterial input function is not recommended in MMD owing to the proximal stenoses and the presence of collaterals, which can introduce delay and dispersion in the bolus of contrast.10 These potential errors can be minimised by using quantitative regional analysis based on summary parameters and a reference region in the cerebellum.4 10 We measured cerebral blood volume in each hemispheric ROI and two reference ROIs in the cerebellum. We calculated the relative regional CVR using the regional cerebral blood volume ratio before and after acetazolamide.

Statistical analysis

We assessed the correlation between ADC and CVR using the Spearman correlation test and linear regression analysis. A receiver operating characteristic (ROC) curve was constructed to determine how ADC averaged on both hemispheres can be predictive of executive dysfunction. We determined the best ADC value predictive of executive dysfunction using the Youden's index. Patients were divided into two groups according to this cut-off value. We tested the correlation between elevated ADC and executive dysfunction using the Fisher exact test.

Values are given as mean±SD. All tests were bilateral. p Values <0.05 were considered significant. Statistical analysis was performed with the SPSS V.14.0 software.


Fourteen patients, nine women and five men (mean age 36.6±12.9 years) were included (see online supplementary table 1). Ethnicities were Caucasian (12 patients), and African-West-Indian (two patients). Eight patients had a history of ischaemic stroke. The mean duration of disease from diagnosis to inclusion was 36.1±37.4 months.

All but one patient agreed to undergo neuropsychological testing. Seven of 13 patients had an executive dysfunction. Neuropsychological findings have been reported previously in detail.4 Z-scores on the trail-making test B and Wisconsin sorting test (categories) were impaired in all patients with executive dysfunction.

MRI data

Twenty-eight cerebral hemispheres were studied. There was no evidence of recent infarction on DWI in any patient. Five patients were free of old infarctions. Fifteen hemispheres showed at least one old cerebral infarct on T2-weighted images (figure 2 in online supplementary material). Cerebral infarcts were subcortical (four hemispheres), cortical (one hemisphere), or cortical and subcortical (10 hemispheres).

The respective mean values of ADC and CVR were 822.8±51.9×10−6 mm2/s and −1.3±25.9% on left hemispheres; 798.7±53.7×10−6 mm2/s and 9.4±28.5% on right hemispheres. We found a negative correlation between ADC and CVR (Spearman correlation coefficient: −0.46; p=0.01) (figure 1).

Figure 1

Scatter plot of cerebrovascular reserve (CVR) according to apparent diffusion coefficient (ADC) values (Spearman correlation coefficient: −0.46; p=0.01). The graph represents the linear regression analysis (R2=0.19; p=0.02).

Elevation of ADC significantly predicted executive dysfunction (ROC curve (95% CI): 0.85 (0.59 to 1.16); p=0.032). Using the threshold value determined with the Youden's index (802.6; sensitivity 85.7% and specificity 100%), we found a strong correlation between increased ADC and executive dysfunction (p=0.005) (table 1).

Table 1

Patients are classified according to the presence of a dysexecutive cognitive syndrome (DCS) and apparent diffusion coefficient (ADC) value in the frontal white matter


In this study of adult patients with MMD, elevated ADC in the frontal white matter, appearing normal on T2-weighted images, correlated with altered CVR and executive dysfunction.

The graded relationship between the elevation of ADC and reduction of CVR within the same ROIs suggest that elevated ADC reflected subtle ischaemic damage. Our finding is in good agreement with the previous study by Conklin et al.7 These authors used blood oxygen level-dependent MRI response to hypercapnia in patients with MMD. They demonstrated that regions of negative CVR, that is, with a steal phenomenon, were spatially correlated with elevated ADC in normal-appearing white matter.

There are only a few reports including a neuropsychological assessment of adult patients with MMD. These studies suggest that cognitive impairment is common, and executive function is predominantly affected.2–4 However, the mechanisms of cognitive impairment are uncertain. Our findings raise the hypothesis that subtle ischaemic injury of the frontal white matter, as documented by elevated ADC without abnormality on T2 images, may be responsible for executive dysfunction.11 However, it should be noted that most patients included in our study had associated cerebral infarction. Therefore, elevation of ADC could alternatively reflect secondary Wallerian degeneration of axons projecting into the normal-appearing frontal white matter from areas of infarction.5 Further studies are needed to determine whether our results can be confirmed in patients without overt cerebral infarction.

Using ADC may be of potential clinical interest for serial examinations of patients with MMD. ADC measurement is easier to perform and, possibly, more reproducible than CVR calculation. If complementary studies confirm our findings, ADC measurement might be used to detect patients at risk for cerebral ischaemia and cognitive impairment to allow therapeutic intervention at an early stage.


Supplementary materials


  • Competing interests None.

  • Ethics approval This study was approved by the medical institutional review board of “Centre Hospitalo-Universitaire” of Toulouse and the local ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement The data of the study are eventually accessible by contacting the principal author.