Relation of Left Ventricular Hypertrophy to Regional Cerebral Blood Flow: Single Photon Emission Computed Tomography Abnormalities in Essential Hypertension

Abstract

Several reports have shown that left ventricular hypertrophy (LVH) is an independent predictor of acute cerebrovascular events. The aim of the present study was to investigate the relationship between LVH and cerebral blood flow in middle‐aged patients with essential hypertension. Forty never‐treated hypertensive patients (24 men, 16 women, aged 50–60 years) without clinical evidence of target organ damage were studied. Regional cerebral blood flow was measured by means of single photon emission computed tomography of the brain. Twenty‐nine patients showed echocardiographic criteria of LVH; 11 patients did not show this feature. No differences were found in regional cerebral blood flow ratio of all brain areas studied between hypertensives with or without LVH except for the striatum area. The regional cerebral blood flow ratio was significantly reduced in the striatum region of hypertensive patients with LVH, compared with patients without LVH (91.5±7.4 vs 98.1±8.3; P=.023). This relationship remained significant after adjusting for blood pressure. The authors conclude that the presence of LVH in middle‐aged patients with essential hypertension is associated with a reduction of regional cerebral blood flow in the striatum area.

A number of studies have reported that echocardiographically determined left ventricular (LV) hypertrophy (LVH) is an independent risk factor for cardiovascular morbidity and mortality in both the general population ¹ and essential hypertensive patients. ² In addition, Bikkina et al ³ demonstrated that LV mass is associated with an increased risk of cerebrovascular events, such as stroke and transient ischemic attacks, in an elderly cohort from the Framingham Heart Study. Verdecchia et al ⁴ also showed that LVH is an independent predictor of stroke in essential hypertension. There are some reports that have also shown a relationship between LVH and asymptomatic cerebrovascular damage, such as lacunar infarcts and white matter lesions, in patients with essential hypertension. ⁵ , ⁶ , ⁷

It is well known that hypertension is a major risk factor for stroke. High blood pressure (BP) markedly influences cerebral blood flow (CBF), causing adaptive vascular changes and an increased incidence of stroke. ⁸ Hypertensive vascular adaptation of CBF autoregulation includes vessel wall connective tissue proliferation, muscular hyperplasia, and degenerative changes. It has been shown that a reduced CBF may be found in hypertensive patients with neurologic deficits such as transient ischemic attacks ⁹ and in people without clinically evident cerebral involvement, especially if untreated. ¹⁰

The fact that LVH is related to stroke could be due to some regional CBF (rCBF) abnormalities that are probably present at early stages of hypertensive cerebrovascular damage. The presence of LVH and its possible relationship with CBF has not been fully explored. Nobili et al ¹¹ did not find significant correlation between CBF, measured by the xenon Xe 133 inhalation method, and LVH, determined by an electrocardiogram (ECG), in a group of 101 hypertensive patients, most of them on antihypertensive therapy.

It has been demonstrated that cerebral single photon emission computed tomography (SPECT) with technetium Tc 99m‐hexamethyl‐propylene amine oxime (HMPAO) is an accurate measure of rCBF. ¹² , ¹³

The aim of the present study was to investigate the relationship between LVH, determined by echocardiography, and rCBF‐SPECT in asymptomatic middle‐aged patients with essential hypertension.

METHODS

Patient Selection

Forty never‐treated essential hypertensive patients of both genders, aged 50–60 years, were consecutively selected from the Hypertension Unit where they had been referred for diagnosis and management of hypertension. They were recruited on their first visit if they satisfied the inclusion criteria. The diagnosis of essential hypertension was considered if no known cause of high BP could be detected after complete clinical, biochemical, and radiologic examination. All patients had systolic BP ≥140 mm Hg and/or diastolic BP ≥90 mm Hg measured over at least 3 separate 1‐week intervals. Exclusion criteria included type 2 diabetes mellitus (fasting plasma glucose >6.6 mmol/L), carotid stenosis >50% measured by ultrasonography, alcohol intake >30 g/d of pure ethanol, clinical evidence of cerebrovascular or coronary heart diseases (positive clinical history, ECG abnormalities, or positive exercise test), cardiac failure, atrial fibrillation, papilloedema, and renal impairment (serum creatinine >115 μmol/L). Patients were not receiving pharmacologic treatment. All patients gave informed consent. The study was approved by the Ethics Committee of the Hospital Clinic of Barcelona and by the Spanish Health Authority.

BP Measurement

Twenty‐four‐hour ambulatory BP monitoring was performed with a portable noninvasive oscillometric device (SpaceLabs 90207, Spacelabs Inc, Issaquah, WA). BP readings were obtained automatically every 15 minutes throughout the 24 hours. The following ambulatory BP monitoring parameters were evaluated: averages of 24‐hour, daytime (from 8 am to 11 pm), and nighttime (from 11 pm to 8 am) systolic BP, mean BP, diastolic BP, pulse pressure (difference between systolic and diastolic BP), and heart rate.

Echocardiographic Studies

A 2‐dimensional M‐mode echocardiogram was obtained with the patient lying partially on the left side, after a 10‐minute rest. The recordings were assessed by two previously trained examiners. In accordance with the criteria of the American Society of Echocardiography, ¹⁴ the following parameters relative to the left ventricle were obtained, each as an average of at least 3 measurements: LV end‐diastolic diameter, LV diastolic posterior wall thickness, and interventricular septum thickness. LV mass was determined by the Penn convention criteria ¹⁵ and divided by the body surface area to calculate LV mass index in g/m². LVH was diagnosed if LVMI exceeded 110 g/m² in women and 130 g/m² in men.¹⁶

Brain SPECT

rCBF images were obtained by SPECT using ^99mTc‐HMPAO (Ceretec, Nycomed Amersham Imaging, Princeton, NJ). The acquisition and reconstruction protocols used in our Department of Nuclear Medicine have been described elsewhere. ¹⁷ , ¹⁸ Briefly, a brain SPECT was performed using a rotating dual‐head gamma camera (Helix, GE Healthcare, Chalfont St Giles, UK), fitted with a high‐resolution fanbeam collimator. Data acquisition started 20 minutes after IV injection of 740 MBq of ^99mTc‐HMPAO in a quiet room. Sixty 30‐second frames were collected in a 360° circular orbit, step and shoot mode, using a 128 ′ 128 matrix. Image data were processed on an Elscint SP1 computer (Apex SP‐X, software version 3.12, Elscint Ltd, Tel Aviv, Israel). Reconstruction was performed by filtered back projection and using a Metz filter (full width at half maximum = 10; power factor = 3) without attenuation correction. The final pixel size was 3.9 mm and spatial resolution was 10 mm (full width at half maximum) in the transaxial plane.

Semiquantitative rCBF analysis was performed by drawing irregular regions of interest (ROIs) in 8 standardized 9‐mm thick oblique slices taken from the frontocerebellar direction, as described previously by Goldenberg et al. ¹⁹ The ROIs were drawn by the same observer, blind to clinical data, on the left hemisphere, and a mirrored ROI was placed on the right hemisphere. Uptake indices were obtained in each hemisphere for the following regions: anterior frontal, posterior frontal, anterior temporal, posterior temporal, parietal, occipital, cerebellum, superior cingulate, inferior cingulate, thalamus, and striatum. The cerebellum was chosen as a reference area since it is not generally affected by either anatomic or functional abnormalities. ²⁰ Regional indices of CBF in the ROIs were obtained, expressed as the ratio between average pixel counts in each ROI and average pixel counts of the cerebellum. This ratio allows the uptake of each region from each hemisphere to be obtained relative to that in the reference area.

Brain SPECT and BP monitoring were performed the same day.

Statistical Analyses

Values are expressed as mean (SD) for normally distributed variables and as median (interquartile range) for variables that deviated from normal distribution. Differences between hypertensive patients with and without LVH were analyzed by the 2‐tailed unpaired Student t test or the nonparametric Mann‐Whitney test for numeric variables. Chi‐square or Fisher exact tests, as appropriate, were applied for categoric variables. Analysis of covariance was used to assess the association of rCBF ratios and LVH, adjusting for BP values. A value of P<.05 was considered the level of statistical significance.

RESULTS

Twenty‐nine of the 40 hypertensive patients had echocardiographic criteria of LVH. Essential hypertensive patients were classified into 2 groups on the basis of presence or absence of LVH. As shown in Table I, main baseline characteristics, including age, gender distribution, body mass index, serum fasting glucose, lipid profile, renal function, duration of hypertension, and smoking status did not differ between groups. Echocardiographic data defining LV mass characteristics are shown in Table II.

Table I.

Baseline Characteristics of Essential Hypertensives Patients With and Without Left Ventricular Hypertrophy (LVH)

Parameter	With LVH (n=29)	Without LVH (n=11)	p Value
Age, y	54.5 (4.7)	54.5 (2.9)	.979
Sex, No., men/women	18/11	6/5	.728
Body mass index, kg/m2	29.3 (3.6)	29.5 (1.9)	.774
Smokers, %	20.6	27.2	.686
Duration of hypertension, mo*	27 (2–48)	22 (5–74)	.562
Serum creatinine, (μmol/L	81.0(13.5)	74.7 (18.6)	.243
Serum cholesterol, mmol/L	5.2 (0.7)	5.4 (1.3)	.388
HDL cholesterol, mmol/L	1.26 (0.31)	1.37 (0.37)	.375
LDL cholesterol, mmol/L	3.31 (0.63)	3.55 (1.17)	.412
Serum glucose, mmol/L	5.3 (0.6)	5.2 (0.5)	.284
Values are mean (SD) except as otherwise indicated. HDL indicates high‐density lipoprotein; LDL, low‐density lipoprotein. *Median (interquartile range).

Table II.

Differences in Echocardiographic Parameters Between Hypertensive Patients With and Without Left Ventricular (LV) Hypertrophy (LVH)

Parameter Mean (SD)	With LVH	Without LVH	p Value
Posterior wall thickness, mm	11.34(1.04)	9.80 (1.39)	.001
Interventricular septum thickness, mm	12.10(1.39)	10.70 (1.88)	.017
LV end‐diastolic diameter, mm	52.6 (4.2)	47.2 (4.2)	.001
Relative wall thickness ratio, mm	0.43 (0.06)	0.42 (0.09)	.593
LV mass, g	295.0 (50.6)	200.6 (45.2)	.000
LV mass index, g/m2	152.3 (19.8)	106.1 (15.7)	.000
Fractional shortening, %	38.9 (4.7)	37.6 (5.9)	.476

Essential hypertensive patients with LVH showed significantly higher values of 24‐hour systolic BP (145.2±14.8 mm Hg vs 136.2±11.1 mm Hg; P=.012) and pulse pressure (54.2±9.3 mm Hg vs 47.6±9.5 mm Hg; P=.013), compared with those without LVH (Table III). Mean 24‐hour diastolic BP tended to be higher in patients with LVH compared with those without LVH, although this difference did not reach statistical significance (P=.218).

Table III.

Mean 24‐Hour Blood Pressure (BP) and Heart Rate in Hypertensive Patients With and Without Left Ventricular Hypertrophy (LVH)

Parameter, Mean (SD)	With LVH	Without LVH	p Value
Systolic BP, mm Hg	145.2 (14.8)	136.2 (11.1)	.012
Diastolic BP, mm Hg	91.9(10.1)	87.6 (8.7)	.218
Pulse pressure, mm Hg	53.2 (9.3)	48.6 (9.5)	.013
Heart rate, bpm	69.2 (8.7)	74.4 (6.5)	.081

With respect to brain SPECT, the average uptake ratios for individual brain ROIs/cerebellum for hypertensive patients with and without LVH are shown in Table IV. No significant differences were found in the rCBF ratios of all brain areas studied between hypertensives with or without LVH except for the striatum area. We particularly found that patients with LVH showed a significantly reduced perfusion in the striatum region compared with those without LVH (91.5±7.4 vs 98.1±8.3; P=.023). This association remained significant (P=.021) after controlling for 24‐hour BP values. All rCBF ratios were normally distributed.

Table IV.

Regional Cerebral Blood Flow: SPECT Values* in Hypertensive Patients With and Without Left Ventricular Hypertrophy (LVH)

Cerebral Region	With LVH	Without LVH	p Value
Anterior frontal	97.6 (7.3)	97.6	(6.7) .991
Posterior frontal	91.3 (8.3)	95.3	(8.1) .205
Anterior temporal	96.3 (4.8)	98.3	(6.2) .320
Posterior temporal	101.1 (7.2)	104.3	(6.7) .228
Parietal	99.9 (8.5)	105.0	(11.7) .145
Occipital	104.9 (6.9)	109.1	(10.7) .158
Superior cingulate	94.1 (10.5)	97.2	(9.3) .415
Inferior cingulate	101.1 (8.6)	103.1	(8.2) .545
Thalamus	98.1 (10.0)	98.1	(8.0) .988
Striatum	91.5 (7.4)	98.1	(8.3) .023
Values are mean (SD). SPECT indicates single photon emission‐computed tomography. *Expressed as percentage of activity compared with that of the cerebellum (technetium Tc 99m‐hexamethyl‐propylene amine oxime uptake ratio).

DISCUSSION

The present study shows a relationship between the presence of LVH and a reduction of an area of rCBF‐SPECT in asymptomatic middle‐aged untreated essential hypertensive patients. We found a significant relationship between the presence of LVH and a reduced perfusion in the striatum area, independent of BP values.

Cerebral SPECT with ^99mTc‐HMPAO allows a semiquantitative evaluation of brain perfusion in humans. HMPAO is a lipophilic molecule that crosses the blood‐brain barrier and is converted within the brain cell into a hydrophilic form stable for many hours, permitting the display of areas of increased or decreased perfusion. The striatum and thalamus are the brain areas most susceptible to lacunar thrombotic infarction, particularly in hypertensive patients, because of vessel anatomy and vascular supply in these areas. ⁹ Cerebral blood flow in hypertensive patients is maintained at the same level as in normotensive individuals by means of cerebral autoregulation until significant arteriosclerosis and hypertensive vascular disease develop. ⁸ Studies of cerebral hemodynamics in essential hypertension have suggested that there are various degrees of cerebral arteriopathy between early and late stages of hypertension. ²¹ , ²² , ²³

A possible relationship between CBF and the existence of target organ damage in other organs in essential hypertensive patients is not well known. Nobili et al ¹¹ did not find a significant correlation between rCBF and the existence of LVH and retinopathy in 101 asymptomatic hypertensive patients (age range, 19–78 years). In that study, LVH was assessed by ECG, which is less sensitive than an echocardiogram, ¹⁵ and 61% of patients were receiving antihypertensive therapy.

To our knowledge, no studies to date have found a relationship between LVH and a decrease in rCBF, especially in the striatum area, in asymptomatic, middle‐aged, and never‐treated essential hypertensive patients. Limitations of the study are that present results are from a cross‐sectional study, and that the sample size was small to reliably detect differences. The presence of several comparisons could lead to a spurious finding, but there are some characteristics that may be emphasized. Because aging and associated factors can influence both cerebral hemodynamics ²³ , ²⁴ , ²⁵ and LVH, ²⁶ the present study included a homogeneous sample of middle‐aged patients with essential hypertension; risk factors for the development of cerebrovascular damage, such as diabetes or significant alcohol intake, were excluded. No patient had ever received anti‐hypertensive treatment that could have influenced the presence of LVH ²⁷ or cerebral hemodynamic features. ²⁵ , ²⁶ , ²⁷ , ²⁸ Patients with a history of cardiovascular disease were also excluded, since an association between these diseases and the presence of cerebral hemodynamic alterations has been reported. ²³ , ²⁹ , ³⁰

The present study shows a significant relationship between LVH and reduced perfusion of the striatum, independent of BP values. It is known that both brain and heart are targets of hypertension‐induced organ damage. ⁴ , ⁵ , ⁶ , ⁷ It has been hypothesized that the association of cardiac and cerebral injury could be due to a generalized impact of arterial hypertension that underlies both phenomena; however, these studies showed that the relationship between LVH and cerebrovascu‐lar disorders were independent of BP values. ⁴ , ⁵ , ⁶ , ⁷ In this sense, Verdecchia et al ³¹ have shown that interindividual variation in LVM is explained only slightly by BP, and that LVM might reflect long‐term exposure to several factors in addition to BP, such as genetic, hormonal, or metabolic factors. ⁴ , ³¹ , ³² On the other hand, Rogers et al ³³ showed in a longitudinal study (mean follow‐up, 50.12 months) performed in asymptomatic elderly hypertensive volunteers (mean age, 65.8 years) that patients with initial lowest CBF values showed a greater incidence of cerebrovascular disease than patients with higher values of CBF, despite the fact that there were no differences in BP.

The present study showed a relationship between the presence of LVH and a reduction of CBF in the striatum region. This region appears to be the most common area of the brain to suffer from hypertensive complications. A reduction of CBF could be a first step in the development of cerebrovascular complications in hypertensive patients with LVH; however, the mechanisms connecting LVH to cerebrovascular injury are not clear. It is difficult to differentiate the relative role of elevated BP from a direct contribution of LVH to the increased risk of developing cerebrovascular disorders; longitudinal studies are necessary.

Detection of cardiac hypertrophy may help to obtain a better assessment of global cardiovascular risk in middle‐aged hypertensive patients not only to discover cardiac damage but also to distinguish patients at risk of developing cerebrovascular injury.

Disclosure: This study was supported in part by grants from the Fondo de Investigaciones Sanitarias (FIS 02/0177), and Redes Temdticas de Investigation Cooperativa Sanitaria (RECAVA C03/01; Nodo IDIBAPS), Ministerio de Sanidad y Consumo, Spain.