J Neurol (2005) 252 : 163 167 DOI 10.1007/s00415-005-0624-3 ORIGINAL COMMUNICATION Sumeet Singhal Hugh S. Markus Cerebrovascular reactivity and dynamic autoregulation in nondemented patients with CADASIL (cerebral autosomal infarcts and leukoencephalopathy) Received: 25 February 2004 Received in revised form: 17 July 2004 Accepted: 30 July 2004 S. Singhal, BSc H. S. Markus, MD Department of Clinical Neuroscience St George s Hospital Medical School London, UK S. Singhal ( ) 3 Finlarigg Drive Edgbaston, Birmingham B15 3RH Tel.: +44-7870/133050 E-Mail: sumeet@zinc.freeserve.co.uk The study was approved by the Local Research Ethics Committee. Abstract Reduced cerebrovascular reactivity has been reported in patients with cerebral autosomal infarcts and leukoencephalopathy (CADASIL), and the measurement has been suggested as a useful surrogate marker of disease progression. Previous studies have not determined whether cerebral autoregulation is also impaired. We measured dynamic cerebral autoregulation and carbon dioxide reactivity in 24 nondemented CADASIL patients and 20 controls, using transcranial Doppler ultrasound (TCD). No impairment in either measure was found in the CADASIL group. We conclude that either cerebrovascular reactivity and autoregulation are not impaired in early disease, or that TCD may not be a sufficiently sensitive tool to detect haemodynamic changes in early disease. TCD is unlikely to be useful for disease monitoring in patients without advanced disease. Key words cerebrovascular reactivity cerebral autoregulation CADASIL Introduction Cerebral autosomal infarcts and leukoencephalopathy (CADASIL) is an autosomal dominant disease characterised by migraine and recurrent premature subcortical strokes in early adulthood, leading to a subcortical dementia with advancing disease. The typical neuroimaging findings of lacunar infarcts and leukoaraiosis occur early in the disease process. CADASIL is caused by mutations in the NOTCH3 gene [6]. Neuropathologically, a diffuse cerebral arteriopathy is found, most severe in small arteries supplying the deep white matter, with loss of medial smooth muscle cells [10]. The arteriopathy appears to lead to cerebral hypoperfusion, which has been detected at the level of the major arteries,and at the capillary level within areas of leukoaraiosis [3, 13]. Hypoperfusion alone may be insufficient to explain the acute and chronic ischaemia seen in this condition, as the cerebral circulation has autoregulatory mechanisms in place to compensate for reductions in perfusion pressure. Thus, it has been further suggested that the arteriopathy results in an earthen pipe state, in which both constriction and dilatation are impaired,with consequent failure of cerebral autoregulation [10]. Cerebral autoregulation refers to the brain s ability to maintain adequate perfusion despite changes in blood pressure, through compensatory changes in cerebral blood flow (CBF). Dynamic autoregulation can now be measured non-invasively using transcranial Doppler ultrasound (TCD). Following a sudden drop in blood pressure, the rate of rise of middle cerebral artery blood flow velocity (MCA-V) is compared with that of continuously monitored blood pressure [1]. However, the mechanism of autoregulation is complex and incompletely understood. It is at least partly dependent on changes in vessel radius, which is inversely related to resistance. Changes in cerebrovascular resistance lead to changes in CBF. Smaller vessels experience a greater proportionate change in radius than larger vessels, leading to the hypothesis that the vessels most involved in determining JON 1624
164 autoregulation are the distal small arteries and arterioles. If this is the case, then a small vessel arteriopathy such as CADASIL might be particularly expected to show impaired autoregulatory responses. Cerebrovascular reactivity is a more straightforward concept than cerebral autoregulation. It is directly dependent on the ability of vessels to dilate or constrict in the face of appropriate stimuli. There is evidence that reactivity to carbon dioxide (CO 2 ) as measured by TCD is impaired in patients with non-inherited small vessel disease, for example in patients with lacunar infarction [7] or hypertension [8]. Reactivity studies in CADASIL, using both MRI and TCD approaches, have shown variable results. Reactivity to the vasodilator acetazolamide was reduced in an initial MRI perfusion study [3], but a recent study measuring large artery flow in patients with early disease reported intact vasodilatory responses [13]. The TCD approach uses change in middle cerebral artery blood flow as an indirect marker of change in blood flow in smaller arteries downstream. Impaired CO 2 reactivity as measured by TCD has been reported in a cohort of 29 CADASIL patients [11], with the greatest reduction occurring in disabled patients. However, this study used a 5 % CO 2 concentration, which causes variable degrees of blood CO 2 rise. This can be controlled for by measuring endtidal CO 2 concentrations, but this was not performed in this study. Although vasoactive dysfunction may partly underlie both cerebrovascular reactivity and autoregulation, it has been suggested that each reflects a different mechanism controlling cerebral blood flow [5]. Consistent with this is evidence that the pathophysiological bases of the hypercapnic and autoregulatory responses differ [14, 16]. Therefore, in order to know if autoregulation in CADASIL is impaired, it becomes necessary to specifically assess the autoregulatory response, but to our knowledge this has not yet been attempted. As well as possibly illuminating the mechanism behind acute and chronic ischaemia in CADASIL, TCD may have an additional role. With the advent of clinical trials of putative therapeutic agents, for example drugs increasing cerebral perfusion, markers of response to therapy will be required. The unpredictable and saltatory nature of clinical events in CADASIL means that identification of surrogate markers of clinical disease progression to assess the response to treatment will become necessary. TCD measures of reactivity and autoregulation may offer a simple, non-invasive approach to disease monitoring. We measured both CO 2 reactivity and dynamic autoregulation in a group of nondemented CADASIL subjects without advanced disease, and in normal controls. We studied subjects with early disease, because we wished to determine if alterations in haemodynamic parameters could be detected prior to severe tissue damage, as this would be when preventative therapies were most likely to be beneficial. Subjects and methods Twenty-four CADASIL subjects were recruited; all had known NOTCH3 mutations. Cases were excluded if they had had a stroke in the past 3 months, as this might transiently alter autoregulatory responses. Twenty control subjects were recruited from patients spouses and hospital staff. Carotid stenosis, which can affect CO 2 reactivity and autoregulation [14], was excluded by ultrasound. Subjects discontinued any antihypertensive drugs for 12 hours, and statins for 72 hours, prior to testing, and were asked not to smoke on their study day. The study was approved by the Local Research Ethics Committee and all subjects gave written informed consent prior to their inclusion in the study. TCD recordings Bilateral simultaneous MCA-Vs were recorded via the transtemporal window using a Multidop X4 TCD system (DWL, Sipplingen, Germany). Continuous arterial blood pressure recordings were made by finger plethysmography (Finapres 2300, Ohmeda, Ky, USA). Subjects inspired air, followed by 6 % and then 8 % CO 2,and end-tidal pco 2 was monitored. Inspiration continued until MCA-V had plateaued. Reactivities were calculated as the percentage change in MCA-V from baseline (air) to 6 % or 8 % CO 2.As 6 % is a submaximal stimulus, the MCA-V at this concentration was divided by the change in steadystate end-tidal pco 2.For dynamic autoregulation testing, the thigh cuff technique was used to induce a stepped drop of at least 10 mm Hg in blood pressure [1]. The autoregulatory index (ARI) was derived using semi-automated software supplied with the TCD system. It is measured on a scale of 0 9, the higher the ARI, the better the autoregulatory response. In one CADASIL subject, aged 66 years, an acoustic window was absent. In all remaining 23 cases and all 20 controls successful 8 % reactivity measurements were performed; two CADASIL subjects declined full reactivity testing and underwent only 8 % testing. Technically successful dynamic autoregulation studies were achieved in 14 CADASIL subjects and 16 controls. In the remaining cases, ARI could not be calculated. This was either because of inadequate blood pressure signal quality, or because of insufficient size or quality of blood pressure drops (either < 10 mmhg or not a stepped drop). Data analysis and statistics Data were analysed off-line by a single experienced rater (HM) blinded to the identity and case/control status of the subjects. For both ARI and reactivity, a mean value was calculated from right and left values. Independent samples group comparisons were carried out using t-tests for normally distributed variables, or the Kruskal-Wallis test for gender differences, using the statistical software package SPSS 11.0 (SPSS, Chicago, USA). Results Three CADASIL patients were asymptomatic and another ten suffered with migraine alone. Nine (39 %) had had a previous cerebrovascular event, but only two (3 %) had had recurrent strokes. All scored 27/30 or more on the Folstein Mini-Mental State Examination (MMSE). There were no differences between cases and controls in
165 mean (SD) age (44.7(9.6) vs 45.6(11.8) years, p = 0.795), gender (male 10/23 vs 8/20, p = 0.818), serum cholesterol (5.85 vs 5.22, p = 0.128), or systolic/diastolic blood pressures (128/80 vs 120/75, p = 0.181/p = 0.135). Mean (SD) resting MCA-V was lower in CADASIL subjects than in controls [55.0 (16.6) vs 71.2 (14.9) cm/s, p=0.002]. Mean (SD) pulsatility index was higher in CADASIL subjects than in controls [0.931 (0.141) vs 0.830 (0.169), p = 0.039]. There were no significant differences in mean reactivities for either 6 % CO 2 (CADASIL 26.1 vs controls 28.1 %/kpa; p = 0.569); or 8 % CO 2 (CADASIL 60.3 vs controls 53.4 %; p = 0.373) Figs. 1 and 2. For autoregulation testing, the mean blood pressure drop was lower in the CADASIL group (15 vs 21 mmhg; p=0.013). There was no difference in mean ARI (CADASIL 5.8 vs controls 5.6; p = 0.562) Fig. 3. Discussion This study did not find an impairment of either carbon dioxide reactivity or dynamic cerebral autoregulation in CADASIL. The finding of reduced resting MCA blood flow velocity is consistent with cerebral hypoperfusion. The finding of increased MCA pulsatility index is consistent with increased distal cerebrovascular resistance, which suggests reduced luminal diameter. Previous imaging studies in patients with established disease have also noted cerebral hypoperfusion in CADASIL. Lower cerebral blood flow has been found in subcortical and cortical white matter compared with Fig. 2 Scatterplot of values of mean right and left middle cerebral artery 8 % carbon dioxide reactivities for individual CADASIL and control cases. Horizontal bars show mean values for each group Fig. 3 Scatterplot of values of mean right and left middle cerebral artery autoregulatory indices (ARI) for individual CADASIL and control cases. Horizontal bars show mean values for each group Fig. 1 Scatterplot of values of mean right and left middle cerebral artery 6 % carbon dioxide reactivities for individual CADASIL and control cases. Horizontal bars show mean values for each group controls [9], and in white matter hyperintensities compared with normal appearing white matter in patients [2, 3]. Our results suggest that cerebral hypoperfusion occurs early in the disease process. It could be speculated that cerebral hypoperfusion is the initial pathogenic insult that leads to the clinical and radiological changes seen in CADASIL, such as lacunar strokes and leukoaraiosis. This may occur through narrowing of the
166 vessel lumen, which could be reflected in the increased pulsatility index found in our study. Indeed, pathological studies have consistently found vessel wall thickening and hyaline degeneration, although occlusive disease appears rare [10, 12]. This pathogenic mechanism requires cerebral hypoperfusion to be a primary event, but it could be argued that it occurs as a secondary response to ischaemic damage. Against this is the finding from a dynamic contrast MRI perfusion study that the degree of hypoperfusion does not significantly differ between patients with mild or moderate-severe white matter changes [13]. Ultimately, longitudinal studies would be required to determine whether hypoperfusion truly precedes the development of symptoms and radiological changes. The failure to find impaired cerebrovascular reactivity contrasts with two previous studies [3, 11] and may have several explanations. It is possible reactivity was truly preserved in our CADASIL cohort. Our patients had relatively early disease, as evidenced by only two with recurrent strokes, none disabled and none with dementia. Hence, haemodynamic function at this stage may be mild and clinically undetectable. There is some supporting evidence for this from the recent finding of normal acetazolamide reactivity in patients with relatively early disease [13]. In contrast, the previous TCD and MRI perfusion studies, which found impairment of cerebrovascular reactivity, tended to feature patients with more advanced disease [3, 11]. It is also possible methodological differences are responsible for the contrasting results in cerebrovascular reactivity. The MRI perfusion study measuring reactivity at the capillary level found impaired responses [3], whereas the MRI study using measurements of total cerebral blood flow in the carotid and basilar arteries found intact responses [13]. In CADASIL, the small cortical arteries and arterioles are relatively preserved, so cortical responses would be expected to be normal. If changes in large artery flow were dominated by cortical rather than subcortical vessel responses, abnormal perforating vessel responses might not be detectable by MRI or TCD measures of carotid, basilar or middle cerebral artery flow, particularly in less advanced disease. This cannot explain the disparity between our results and those of the only previous TCD study to measure reactivity [11]. Aside from differences in disease severity, there was also a methodological difference between the two studies,which may explain the conflicting results.as this previous study used a 5 % submaximal concentration of carbon dioxide, there may have been a variable rise of pco 2.This could have accounted for the observed difference in reactivity between CADASIL and control subjects, if there had been study group differences in pco 2 rises. The failure to find impairment of the autoregulatory response may likewise have been due to a lack of sensitivity of TCD to subcortical vascular responses, coupled with early stage disease. In contrast to the situation with cerebrovascular reactivity, MRI perfusion techniques do not have the necessary temporal resolution to detect cerebral blood flow changes over the short time period of the dynamic autoregulatory response, and so would not be a useful alternative to TCD. In addition, it is possible that the autoregulatory response is selectively preserved in CADASIL. Neuropathological studies demonstrate severe cerebral perforating artery medial smooth muscle cell destruction, but less severe endothelial abnormalities [10]. Dynamic cerebral autoregulation may be endothelium dependent [16], thus relative preservation of the endothelium might contribute to a preserved autoregulatory response. The TCD approach has technical limitations. About 10 % of subjects have no acoustic window, and this occurred in one of our subjects. It is not always possible to achieve the necessary sudden stepwise drop in blood pressure required for autoregulatory testing. This has resulted in inadequate autoregulatory tests in approximately 20 % of subjects in previous studies [15]. Likewise in this study, we were unable to obtain measurements in over a quarter of our subjects. In conclusion, we were unable to show impairment of either carbon dioxide reactivity or dynamic autoregulation using a validated TCD approach in nondemented CADASIL subjects. This suggests that if abnormalities are present, they cannot be detected by TCD in early disease. The implication for this technique is that it is unlikely to be useful for monitoring disease and the response to therapy in this population. Acknowledgements S. S. is supported by the Atkinson Morley s Hospital Neuroscience Research Foundation Harrison Lectureship.
167 References 1. Aaslid R, Lindegaard KF, Sorteberg W, Nornes H (1989) Cerebral autoregulation dynamics in humans. Stroke 20: 45 52 2. Bruening R, Dichgans M, Berchtenbreiter C, Yousry T, Seelos KC, Wu RH (2001) Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy: decrease in regional cerebral blood volume in hyperintense subcortical lesions inversely correlates with disability and cognitive performance. Am J Neuroradiol 22:1268 1274 3. Chabriat H, Pappata S, Ostergaard L, Clark CA, Pachot-Clouard M, Vahedi K, Jobert A, Le Bihan D, Bousser MG (2000) Cerebral hemodynamics in CADASIL before and after acetazolamide challenge assessed with MRI bolus tracking. Stroke 31:1904 1912 4. Furst H, Hartl WH, Janssen I (1994) Patterns of cerebrovascular reactivity in patients with unilateral asymptomatic carotid artery stenosis. Stroke 25:1193 1200 5. Garnham J, Panerai RB, Naylor AR, Evans DH (1999) Cerebrovascular response to dynamic changes in pco 2. Cerebrovasc Dis 9:146 151 6. Joutel A, Corpechot C, Ducros A, Vahedi K, Chabriat H, Mouton P, Alamowitch S, Domenga V, Cecillion M, Marechal E, Maciazek J, Vayssiere C, Cruaud C, Cabanis EA, Ruchoux MM, Weissenbach J, Bach JF, Bousser MG, Tournier-Lasserve E (1996) Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature 383:707 710 7. Maeda H, Matsumoto M, Handa N, Hougaku H, Ogawa S, Itoh T, Tsukamoto Y, Kamada T (1993) Reactivity of cerebral blood flow to carbon dioxide in various types of ischemic cerebrovascular disease: evaluation by the transcranial Doppler method. Stroke 24:670 675 8. Maeda H, Matsumoto M, Handa N, Hougaku H, Ogawa S, Itoh T, Tsukamoto Y, Kamada T (1994) Reactivity of cerebral blood flow to carbon dioxide in hypertensive patients: evaluation by the transcranial Doppler method. J Hypertens 12:191 197 9. Mellies JK, Baumer T, Muller JA, Tournier-Lasserve E, Chabriat H, Knobloch O, Hackeloer HJ, Goebel HH, Wetzig L, Haller P (1998) SPECT study of a German CADASIL family: a phenotype with migraine and progressive dementia only. Neurology 15: 1715 1721 10. Okeda R, Arima K, Kawai M (2002) Arterial changes in cerebral autosomal infarcts and leukoencephalopathy (CADASIL) in relation to pathogenesis of diffuse myelin loss of cerebral white matter: examination of cerebral medullary arteries by reconstruction of serial sections of an autopsy case. Stroke 33:2565 2569 11. Pfefferkorn T, Stuckrad-Barre S, Herzog J, Gasser T, Hamann GF, Dichgans M (2001) Reduced cerebrovascular CO(2) reactivity in CADASIL: A transcranial Doppler sonography study. Stroke 32:17 21 12. Ruchoux MM, Maurage CA (1997) CADASIL: Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. J Neuropathol Exp Neurol 56:947 964 13. Van den Boom R, Lesnik Oberstein SA, Spilt A, Behloul F, Ferrari MD, Haan J, Westendorp RG,Van Buchem MA (2003) Cerebral hemodynamics and white matter hyperintensities in CADASIL. J Cereb Blood Flow Metab 23:599 604 14. White RP, Deane C, Vallance P, Markus HS (1998) Nitric oxide synthase inhibition in humans reduces cerebral blood flow but not the hyperemic response to hypercapnia. Stroke 29:467 472 15. White RP, Markus HS (1997) Impaired dynamic cerebral autoregulation in carotid artery stenosis. Stroke 28: 1340 1344 16. White RP, Vallance P, Markus HS (2000) Effect of inhibition of nitric oxide synthase on dynamic cerebral autoregulation in humans. Clin Sci (Colch) 99:555 560