Cerebral Microangiopathy in Treatment-Resistant Hypertension
Prof Dr Roland E. Schmieder, Nephrology and Hypertensiology, University Hospital Erlangen, Krankenhausstrasse 12, 91054 Erlangen, Germany
J Clin Hypertens (Greenwich). 2011;13:582–587. ©2011 Wiley Periodicals, Inc.
Cerebral microangiopathy is a cause of cognitive impairment and indicates high risk for clinically overt cerebrovascular disease. It develops in patients with or without hypertension, and different pathologies may play a supporting role. In this pilot study, the authors aimed to elucidate risk factors contributing to the deleterious action of hypertension on cerebral small vessels. A cross-sectional study in 42 patients with treatment-resistant hypertension was performed. Microangiopathy was investigated by cerebral magnetic resonance imaging (MRI). Determinants were identified by clinical investigation, computed tomography, intima-media thickness and pulse wave velocity measurement, and urinary albumin excretion. Nineteen of 42 patients had cerebral microangiopathy (23 controls). Patients were different with respect to heart rate (60.5±10.2 vs 69.7±15.1 beats per minute; P=.029) and systolic blood pressure during nighttime (138±13 mm Hg vs 126±18 mm Hg; P=.019). In addition, there were significant differences in pulse wave velocity (10.7±2.0 m/s vs 9.4±1.4 m/s; P=.034), peripheral pulse pressure (70.8±16.3 mm Hg vs 59.2±13.6 mm Hg; P=.016), central pulse pressure (62.9±15.8 mm Hg vs 50.3±14.2 mm Hg; P=.012), and aortic augmentation pressure (15.9±6.0 vs 11.8±6.6; P=.040). Systolic blood pressure and signs of hypertensive vasculopathy such as peripheral and central pulse pressure and pulse wave velocity were associated with cerebral microangiopathy in patients with long-standing treatment-resistant hypertension.
Cerebral microangiopathy is commonly detected on magnetic resonance imaging (MRI) in elderly hypertensive patients and includes the very early stage of microvascular disease (white matter hyperintensities [WMHs] and lacunar infarctions). Smooth muscle hypertrophy, replacement by extracellular, matrix and enhanced small-vessel permeability are generally considered key pathogenetic features, with mural deposition of serum protein detectable specifically in areas of blood–brain barrier breakdown (hyaline arteriolosclerosis, lipohyalinosis). Cerebral microangiopathy identifies a group of individuals at high risk for clinically overt cerebrovascular disease.1 It also has been discussed to be a frequent cause of cognitive impairment and dementia in the elderly; however, the association between the extent of the disease as assessed by cross-sectional MRI studies and cognitive impairment has repeatedly been shown to be weak or even absent.2–4
Established risk factors are advancing age,5–7 cardiac disease,7 diabetes mellitus,8 and hypercholesterolemia.9 Hypertension is a further risk factor,5–7,10 which has been shown to be of particular importance because not only the initiation but also the progression of cerebral microangiopathy appears to depend on actual blood pressure (BP) readings. This was confirmed in the Austrian Stroke Prevention Study, in which the progression of WMHs in 273 clinically healthy patients was studied. A total of 45.1% patients in the group had hypertension with WMHs and 28.1% patients had hypertension without WMHs (P=.006). At 3-year follow-up WMHs were increased in 18% of patients, and the baseline degree of WMH and diastolic BP were the strongest determinants of progression.11 On the other hand, the data illustrate that hypertension is not a prerequisite for the development or progression of cerebral microangiopathy.11 This was also shown in a prior autopsy study in which cerebral small-vessel disease (usually manifested as concentric hyaline wall thickening but not lipohyalinosis and fibrinoid necrosis) was also present in patients who were nonelderly, nondiabetic, and nonhypertensive.12
Taken together, there appears to be a complex interrelationship between BP and a number of other risk factors contributing to cerebral microangiopathy. There is further possibility that different pathologies in patients with or without hypertension may play a supporting role. This is further complicated by the unknown threshold below which an impact of BP can certainly be ruled out and above which there is a definite impact. Because of these considerations, we focused on patients with long-standing uncontrolled hypertension (treatment-resistant hypertension) in this pilot study to gain an in-depth understanding of risk factors that contribute to the deleterious action of long-standing uncontrolled hypertension on cerebral small vessels.
Forty-two patients were recruited for this study by the clinical research competence center in Erlangen-Nürnberg (http://www.crc-erlangen.de). All patients had treatment-resistant hypertension as defined by a casual BP of at least 140/90 mm Hg despite treatment with ≥3 antihypertensive agents including 1 diuretic agent. Secondary hypertension was excluded by patient history, laboratory analysis, and Doppler sonography for renal artery stenosis. If there were any uncertainties as to whether secondary hypertension could be present, the patient was not included in the study. All patients underwent an intensive screening for cardiovascular target organ damage. None of the patients had diabetes mellitus or recent cardiovascular events such as myocardial infarction, congestive heart failure, stroke or similar event.
The study was performed in adherence with the principles of the Declaration of Helsinki and according to good clinical practice standards. Before enrollment in the study, written informed consent was obtained from each participant. The study protocol was approved by the clinical investigations ethics committee of the University of Erlangen-Nürnberg, Germany.
For casual and standing BP, a standardized sphygmomanometer was used, with the cuff size adjusted to the volunteer’s arm circumference. BP readings were computed as the mean of 3 consecutive measurements at each study visit, with the patient seated for 5 minutes. Ambulatory 24-hour BP measurements were performed by an automatic portable device (Spacelab No. 90207; Redmont, CA). Measurement intervals were 15 minutes during the day (7 am–10 pm) and 30 minutes during the night.
Treatment-resistant hypertension was defined by elevated BP readings despite the use of 3 antihypertensive drugs, including a diuretic, in combination with lifestyle measures.13
Cerebral Magnetic Resonance Imaging
Magnetic resonance images were acquired on a 3.0 Tesla scanner (Magnetom Tim Trio; Siemens Healthcare AG, Erlangen, Germany) using a standard protocol including axial fast fluid-attenuated inversion-recovery (FLAIR-), DWI- and GRE-, and T2-weighted sequences. All scans were reviewed by an experienced neuroradiologist (AD) unaware of the clinical or other information. Cerebral microangiopathy on conventional MRI was rated using a modification of the Fazekas scale,14 which was recently developed as part of a multinational European study on age-related white matter changes15 as follows: no WMHs/cerebral microangiopathy (grade 0), punctate (grade 1), early confluent (grade 2), and confluent (grade 3) WMH.
Multiple breath-hold electrocardiographic-gated cine MRI in sequential 8-mm short-axis slices (no gap) were obtained with patients positioned supine in a 1.5-T MRI scanner (Gyroscan ACS/NT; Philips Medical Systems, Best, The Netherlands) using a commercial cardiac coil as described in detail.16,17 For image analysis, a commercially available work station (View Form; Philips Medical Systems) was used. Left ventricular (LV) mass was calculated by manual planimetry of the endocardial and epicardial borders of the LV myocardium on successive short-axis cine images at end-diastole as previously described by Olivotto and colleagues.16 Papillary muscles were excluded from the LV mass calculation. All measurements were performed by one experienced observer who was blinded to all clinical data, including age, weight, and sex of the patient.
CT Imaging for Coronary Calcification
For quantification of coronary artery calcification, multidetector computed tomography (CT) was performed with a 16-section scanner (Sensation 16; Siemens Medical Solutions, Forchheim, Germany). The scan protocol included the acquisition of a low-energy tomogram and a nonenhanced coronary calcification scan according to standard protocols as described in detail previously.18 Evaluation of the acquired transverse images was performed on an offline workstation (Leonardo; Siemens Medical Solutions) by an experienced observer (SA). The Agatston score was used to quantify the amount of coronary calcifications.19
Measurement of the Intima-Media Thickness
Intima-media thickness (IMT) measurements of the right and left common carotid artery were determined in the far walls according to the Mannheim Carotid Intima Media Thickness Consensus using a Siemens G60 S ultrasound machine (Siemens Healthcare AG) with 10 MHz-linear ultrasound transducer. For analysis, IMT values from the left and right side were averaged.20
Measurement of Pulse Wave Analysis
To derive the central arterial waveform, a validated system (Sphygmocor; AtCor Medical, Sydney, NSW, Australia) was used that employs high-fidelity applanation tonometry for noninvasive registration of peripheral arterial pressure waves and appropriate computer software for pressure wave analysis (Sphygmocor). Pressure calibration was accomplished through automatically, noninvasively obtained supine BP of the brachial artery of the dominant arm after a 30-minute rest (Dinamap Compact T; Johnson & Johnson Medical Ltd, Newport, UK). BP was measured 5 times during 10 minutes, and the mean of the last 3 measurements were taken for calibration.
Pressure wave recording was then performed at the radial artery of the same arm with the wrist gently hyperextended. The pressure wave was averaged from single pressure waves recorded consecutively for 8 seconds. Averaged pressure waves were accepted only if variation of peak and bottom pressures of single pressure waves was <5%. The central pressure wave was then automatically synthesized from the radial pressures by a built-in generalized transfer function. Prior to analysis, a visual check for correct detection of inflection points was performed in each radial and central pressure wave by an independent blinded investigator. From the derived central waveforms, data are given on central systolic BP (SBP) and diastolic BP (DBP), time to the first shoulder determined by the outgoing pressure wave (cP1), time to the second shoulder determined by the reflected pressure wave (cP2), either absolutely or as a percentage of ejection duration, as well as augmentation pressure as the pressure height difference between cP2−cP1, and cAI defined as: (pressure difference between cP2−cP1)/pulse pressure (PP).
Pulse Wave Velocity
Aortic pulse wave velocity (PWV) was determined using the foot-to-foot method as described previously.21 Pulse waveforms of the common carotid artery and the femoral artery were sequentially obtained and PWV was calculated as the distance between the suprasternal notch and the femoral artery recording site, divided by the time interval between the feet of the flow waves.
Urinary Albumin Excretion
Urinary albumin excretion was determined by 24-hour urine collection applying the standard laboratory method of nephelometry. For analysis, low-grade albuminuria was used with a threshold of 10 mg/d in men and 15 mg/d in women as previously described.22
All statistical analyses were carried out using SPSS software (release 15.0; SPSS Inc, Chicago, IL). Results are given as mean±standard deviation for parametric data in the text and tables. Comparisons between groups were performed using Student t test for parametric and Mann–Whitney U test for nonparametric data (two-sided).
Baseline characteristics of the 42 patients are displayed in Table I. Nineteen patients had signs of cerebral microangiopathy and 23 patients served as controls. Patient groups were only nominally different with respect to age (mean 61.2±6.8 years vs 56.4±8.7 years; P=.062), but body mass index (28.4±2.8 vs 30.5±3.8; P=.048) and casual heart rate were lower in patients with cerebral microangiopathy (60.5±10.2 beats per minute vs 69.7±15.1 beat per minute; P=.029). Pharmacotherapy was largely comparable between patient groups. All patients received a blocker of the renin-angiotensin system and a diuretic (Table II). The third combination partner was either a calcium channel blocker or a β-blocker in both patient groups. Differences were nonsignificant.
Table I. Patient Characteristics
|Body mass index, kg/m2||28.4±2.8||30.5±3.8||.048|
|Waist circumference, cm||102.4±8.8||106.5±13.6||.274|
|Hypertension duration, mo||218±164||156±98||.138|
| Casual SBP, mm Hg||164.0±17.9||154.7±19.4||.115|
| Casual DBP, mm Hg||94.7±11.5||94.2±11.6||.887|
| Casual heart rate, bpm||60.5±10.2||69.7±15.1||.029|
| Standing SBP, mm Hg||162.3±15.7||154.2±17.5||.123|
| Standing DBP, mm Hg||95.8±8.4||95.4±11.9||.890|
| Standing heart rate, bpm||64.8±12.6||71.4±14.0||.122|
|Heart rate (ECG) under resting and supine conditions||57.6±10.7||60.7±10.1||.331|
|Sokolow Lyon index, mV||2.43±0.76||2.36±0.92||.790|
Table II. Antihypertensive Pharmacotherapy
|RAS blockers, No. (%)||19 (100)||23 (100)||N/A|
|Diuretics, No. (%)||19 (100)||23 (100)||N/A|
|CCBs, No. (%)||12 (63)||15 (65)||1.000|
|β-Blockers, No. (%)||12 (63)||16 (70)||.748|
|α-Antagonists, No. (%)||4 (21)||3 (13)||.682|
|Central sympatholytics, No. (%)||5 (26)||7 (30)||1.000|
24-Hour Ambulatory BP Measurement
Parameters obtained during 24-hour ambulatory BP measurement were comparable in both groups except for SBP during nighttime (Table III). This was higher in patients with microangiopathy (138±13 mm Hg) compared with patients without microangiopathy (126±18 mm Hg; P=.019).
Table III. 24-Hour Ambulatory BP
| Systolic, mm Hg||145±11||138±15||.082|
| Diastolic, mm Hg||86±8||85±8||.664|
| Heart rate, bpm||64±10||69±10||.117|
| Systolic, mm Hg||149±12||143±15||.177|
| Diastolic, mm Hg||89±9.5||88±8.3||.778|
| Heart rate, bpm||66±11||72±12||.102|
| Systolic, mm Hg||138±13||126±18||.019|
| Diastolic, mm Hg||78±7.8||75±10||.246|
| Heart rate, bpm||60±9.5||62±9.4||.450|
Cardiovascular Risk Factors and End Organ Damage
Additionally, a number of cardiovascular risk factors were determined (Table IV). The prevalence of these risk factors was not significantly different between groups; however, the number of pack-years of smoking was nominally higher in patients with cerebral microangiopathy (16.5±21.1 pack-years vs 9.4±14.6 pack-years) as were high-density lipoprotein cholesterol levels (62.0±16.3 mg/dL vs 54±9.4 mg/dL).
Table IV. Cardiovascular Risk Factors
| Pack, y||16.5±21.1||9.4±14.6||.135a|
|Fasting glucose, mg/dL||92±16||94±12||.765|
|eGFR, mL/min/1.73 m2||88±17||95±21||.661|
| Serum aldosterone||225±89||204±97||.475|
| Renin concentration||11.8±16.7||16.4±28.3||.781a|
| Aldosterone/renin ratio||134±400||52±55||.830a|
| PRA, ng/mL/h||1.37±2.5||1.67±3.0||.243a|
| Angiotensin II||4.80±5.3||5.60±8.8||.927a|
|24-Hour urine measurements|
| Sodium excretion||191±80||213±59||.464|
| Albumin excretion, g/d (geometric mean)||7.65±21.37||10.21±16.1||.445a|
There were no differences in left or right ventricular mass, IMT, or the Agatston score (which quantifies coronary calcification) (Table V). On the other hand, there were significant differences in pulse wave velocity (10.7±1.96 m/s vs 9.39±1.39 m/s; P=.034), peripheral pulse pressure (70.8±16 mm Hg vs 59.2± 13.6 mm Hg; P=.016), central pulse pressure (63±16 mm Hg vs 51±14 mm Hg; P=.012), and aortic augmentation pressure (15.9±6.0 mm Hg vs 11.8±6.5 mm Hg; P=.038).
Table V. End-Organ Damage in Patients With and Without Cerebral Microangiopathy
| LV mass, g||154±39||144±24||.333|
| RV mass, g||42±10||40±7.7||.466|
|Intima-media thickness, mm||0.75±0.07||0.76±0.12||.823|
|Pulse wave velocity, m/s||10.7±1.96||9.39±1.39||.034|
|Peripheral pulse pressure, mm Hg||70.8±16.3||59.2±13.6||.016|
|Central pulse pressure, mm Hg||63±16||51±14||.012|
|Aortic AP (heart rate, 75)||15.9±6.0||11.8±6.5||.038|
About 44.1% of patients had cerebral microangiopathy in the present cohort of patients with treatment-resistant hypertension. The analysis of patient and clinical characteristics documents that among the variables tested, age, low casual heart rate, SBP (specifically at night), and signs of hypertensive vasculopathy such as peripheral and central pulse pressure and pulse wave velocity were determinants of cerebral microangiopathy. None of the other risk factors tested proved to be predictive.
Role of SBP
Hypertension has been made responsible for structural alterations in small-caliber vessels and, in particular, the perforating arteries that irrigate the cerebral white matter.23 Accordingly, SBP derived from casual and standing BP measurements and from 24-hour monitoring (during the day and night) was higher in patients than in controls in our analysis. However, only the elevation of SBP readings during nighttime was significant (138±13 mm Hg vs 127±13 mm Hg; P=.039). On the other hand, there was no such trend for DBP values.
This corresponds with various prospective trials that have shown that systolic nighttime BP is superior to day or 24-hour BP in predicting cardiovascular complications.24,25 It is also consistent with data from the Honolulu-Asia Aging Study,26 which looked at a sample of 4000 Japanese-American men and evaluated the relationship between cognitive performance and the level of BP registered 20 to 25 years before. The analysis of the data demonstrated that SBP was inversely related to cognitive status, but DBP was not. In particular, for every 10-mm increase in the level of the SBP, there was a 9% increase in the risk of intellectual decline.
The lowering of BP from a previous hypertensive point into the normal range has been demonstrated for many years to result in stroke prevention.27 This is also accompanied by a reduction in the incidence of dementia during a 15-year follow-up.28 It is, however, most likely a result of the reduction of strokes and not because of a prevention of microangiopathy.23 The latter appears to be less responsive to antihypertensive treatment. The present dataset is limited in this respect by our focus on treatment resistant–hypertensive patients and the largely comparable antihypertensive pharmacotherapy, however.
Signs of hypertensive vasculopathy and arterial stiffness such as increases in peripheral and central pulse pressure as well as pulse wave velocity were determinants of cerebral microangiopathy in our cohort. These findings are supported by data of Henskens and colleagues29 who associated an increase of aortic pulse wave velocity with silent cerebral small-vessel disease in hypertensive patients. They examined 167 hypertensive patients (mean age, 51.8±13.1 years) with untreated hypertension and found that in multivariate analyses-adjusted for age, sex, brain volume, mean arterial pressure, and heart rate, a higher pulse wave velocity was significantly associated with a greater volume of WMH and the presence of lacunar infarcts. Further data by Kearney-Schwartz and colleagues30 also suggested that vascular abnormalities, independently of BP levels, may play a role for WMH in elderly hypertensive patients. They conducted a prospective, cross-sectional study in 198 elderly patients with a mean age of 69.3±6.2 years with subjective memory complaints and had WMH quantified by MRI. In their study, the severity of WMH was independently associated with increased carotid IMT and stiffness (as assessed by augmentation index). Further, after adjustment for a number of cardiovascular risk factors, increased arterial stiffness (as assessed by pulse wave velocity) was significantly and independently associated with memory impairment in men. This fits with data published by Waldstein and colleagues31 who prospectively examined the longitudinal relations of pulse pressure and pulse wave velocity to multiple domains of cognitive function (Baltimore Longitudinal Study of Aging) during a 14-year follow-up. Results of mixed-effects regression models revealed a prospective decline on tests of verbal learning, nonverbal memory, working memory, and a cognitive screening measure among patients with increasing levels of pulse pressure. Persons with higher baseline pulse wave velocity also exhibited prospective decline on tests of verbal learning and delayed recall, nonverbal memory, and a cognitive screening measure. Our data extend our understanding that at the stage of treatment-resistant hypertension, clinical markers of increased stiffness of large arteries are indicative of cerebral microvascular damage.
Aiming to identify determinants of cerebral microangiopathy in treatment-resistant hypertension, we conducted a small pilot study including a total of 42 patients. This may represent a limit to the present analysis in which a number of variables such as casual and standing BP values were nominally but not statistically different. A larger number of patients would have been helpful to further validate the results. On the other hand, typical determinants that have been reported in similar but larger analyses for patients with all stages of hypertension were also identified in the present analysis, making the results useful contributions to this research area. Further, due to the study design, namely inclusion of patients with treatment-resistant hypertension taking treatment with 3 antihypertensive drugs (of which one has to be a blocker of the renin-angiotensin system and one to be a diuretic), no differences in antihypertensive medication were detected and no relationship between antihypertensive drug use and cerebral microangiopathy could be established. This will be reserved for a future larger epidemiologic study.
SBP and signs of hypertensive vasculopathy such as peripheral and central pulse pressure and pulse wave velocity were associated with cerebral microangiopathy in patients with long-standing treatment-resistant hypertension, pointing at a linkage between systemic large and cerebral small-artery disease.