Left Ventricular Geometric Patterns After 1 Year of Antihypertensive Treatment

Abstract

Left ventricular hypertrophy increases the risk for cardiovascular target organ damage, myocardial infarction, and stroke. The authors assessed the patterns of ventricular adaptation in 107 essential hypertensives whose treatment had been withdrawn and its modification after 1 year of hypertension treatment. Blood pressure decreased from 158+17/96+12 mm Hg to 137+15/83+10 mm Hg (mean + SD; p<0.001); 45% of the patients (49 of 107) had their blood pressure controlled below 140 mm Hg and 90 mm Hg. Although a significant decrease of left ventricular mass index was found in the study, the percentage of patients with normal left ventricular geometry at the completion of the study increased by only 9% (27% to 36%, p>0.05). Left ventricular mass geometry improved in 31% of the patients, remained unaffected in 51%, and worsened in 18%. The data suggest that even while suboptimal antihypertensive treatment reduces left ventricular mass index, either left ventricular hypertrophy or concentric remodeling remains present in a significant number of patients at the end of a 1‐year treatment period. The authors conclude that these patients should be considered as a subgroup at high risk and should be treated more aggressively.

The pattern of left ventricular (LV) adaptation to increased afterload is an important determinant of the risk of future cardiovascular (CV) complications in essential hypertensive subjects. ¹ , ² In a long‐term prospective study, Koren et al. ³ found that the risk of CV events was increased not only in patients with both concentric and eccentric LV hypertrophy (LVH), but also in those with LV concentric remodeling, a situation defined by increases in relative wall thickness (RWT) with normal LV mass (LVM). In 1992, Ganau et al. ⁴ reported the patterns of ventricular geometric adaptation in 165 hypertensive subjects. They found that 13% had concentric remodeling, 27% eccentric LVH (increase LVM with normal RWT), and only 8% had concentric LVH (increased LVM and RWT). Similar results were obtained in two other studies conducted in Japan ⁵ and Spain. ⁶ In the Spanish study, ⁶ only 35% of 218 essential hypertensives showed a normal LVM, while hypertension was associated with concentric hypertrophy (23%), eccentric hypertrophy (23%), or concentric remodeling (19%).

The increased risk for clinical events in hypertensive subjects is markedly reduced and even normalized following regression of LVH. ⁷ , ⁸ , ⁹ For this reason, regression of LVH has emerged as an important goal in the treatment of hypertensive patients. Data from comparative studies indicate a reduction of LVM with almost all antihypertensive drugs. ¹⁰ , ¹¹ However, the results of these studies usually are presented only in terms of reduction in both wall dimensions and LVM index (LVMI). Although these parameters are useful, no studies to date have characterized the effect of antihypertensive treatment on the different patterns of LV adaptation. Therefore, we assessed the prevalence of ventricular geometric adaptation patterns in untreated essential hypertensives and its modification after 1 year of treatment.

METHODS

Patients

One hundred seven essential hypertensive patients (29% women) consecutively studied in the hypertension unit of the Hospital Clinico San Carlos with an echocardiogram (Echo) of good quality were accepted to participate in the study. Essential hypertension was diagnosed after an extensive investigation of secondary causes of hypertension, including isotopic renogram. A written informed consent was obtained from all of the subjects and the protocol was approved by the Institutional Review Board. Antihypertensive drugs were stopped at least 6 weeks before obtaining a baseline Echo. Patients with diabetes mellitus, body mass index >30 kg/m ² , renal insufficiency (serum creatinine >1.5 mg/dL), or antecedents of angina, myocardial infarction, or stroke were excluded from the study.

Office blood pressure (BP) was measured at two separate visits spaced 5‐7 days apart, and again twice before the Echo procedure in the supine position. The day of the Echo, 24‐hour urine was collected to measure sodium excretion. The patients were examined again 1 year later. Treatment was not standardized since the follow‐up therapy was managed by referring physicians. Treatment was initiated with combination therapy in 26% of the patients, whereas monotherapy with diuretics was used in 5%, P blockers in 20%, calcium antagonists in 6%, angiotensin‐converting enzyme inhibitors in 30%, and nonpharmacologic measures in 13%. The goal of treatment was to achieve a BP below 140 mm Hg and 90 mm Hg. During the follow‐up, a total of 44 patients (41%) required more than one drug to control their BP.

Echocardiography

Two‐dimensional guided M‐mode Echos (Hewlett‐Packard 5.500, Andover, MA) were obtained with the patients lying in the left lateral position with the head elevated 30°. The transducer was placed in the third or fourth intercostal space and the measurement was done distal to the tip of the mitral leaflets for five consecutive cycles. The interventricular septum thickness, posterior wall thickness, and LV internal dimensions in diastole were measured at the peak of the QRS complex on a simultaneously recorded electrocardiogram according to the Penn convention. ¹² LVM was calculated using the Devereux formula divided by body surface area in square meters to obtain LVMI. ¹² LVH was defined as an LVMI >125 g/m ² because of the well documented prognostic value of such a limit in hypertensive patients. ⁹ RWT was calculated by two posterior wall thicknesses divided by LV internal dimensions in diastole. Patients were classified into four groups according to the baseline values of RWT and LVMI following Ganau's criteria. ⁴ A partition value >0.45 for RWT was used to identify the presence of concentric hypertrophy or, if LVMI was normal, to define concentric LV remodeling. All Echos were recorded and readings were done by two trained sonographers who had no knowledge of the subjects' characteristics. The intra‐observer variability of LVMI, assessed as the difference between two measurements of the same registration, was <1%. The inter‐observer variability was 6%–8%.

Statistical Analysis

Values are expressed as mean ±SD. Data were analyzed by two‐way analysis of variance. Differences in mean values were evaluated by the Student t test. Comparisons of percentages were made using the chi‐square test. A probability value of p<0.05 was considered statistically significant

RESULTS

The mean age of the patients was 42±15 years; 71% were men. The duration of hypertension was 7±2 years. BP averaged 147±12/92±10 mm Hg on therapy before the washout period and 158±17/96±12 mm Hg at the completion of the 6‐week washout period (p<0.01). Heart rate was 77±13 beats per minute (Table I). Neither body mass index nor 24‐hour urinary sodium excretion was above normal for the Spanish population. Twelve months of therapy was associated with significant reductions in BP and heart rate, but no significant changes in body mass index or urinary sodium excretion (Table I).

Table I.

Patient Characteristics at Baseline and After 1 Year of Follow‐Up

Variable	Baseline	i Year*
Systolic blood pressure (mm Hg)	158±17	137±15**
Diastolic blood pressure (mm Hg)	96±12	83±10**
Heart rate (bpm)	77±13	70±12**
Body mass index (kg/m2)	26±3	26±3
Urinary Na+ excretion (mmol/24 h)	159±75	159±70
Values are means ± SD of data collected before and after a 1‐year follow‐up period in 107 hypertensive subjects; *interim blood pressures are not shown; **p<0.001 compared to baseline

The Echo results in the whole group at baseline and after 1 year of treatment are documented in Table II. Significant decreases of LV wall thickness and LVMI were found after 1 year of antihyperten‐sive treatment. RWT also decreased significantly. However, only a small percentage of patients attained normal LV geometry (from 27% to 36%, p>0.05) after 1 year of treatment.

Table II.

Left Ventricular Dimensions and Left Ventricular Geometry by Two‐Dimensional Guided M‐Mode Echocardiography at Baseline and After 1‐Year Follow‐Up Period

Variable	Baseline	One Year
Left ventricular function indices
Interventricular septum (mm)	12.2±2.4	11.3±1.4*
Posterior wall thickness (mm)	10.9±1.6	10.2±1.8*
Left ventricular internal diastolic diameter (mm)	47±5	47±4
Relative wall thickness (mm)	0.47±0.09	0.45±0.07 *
Left ventricular mass index (g/m2)	129±36	116±27*
Left ventricular geometric patterns
Normal (n [%])	29 (27)	39 (36)
Concentric remodeling (n [%])	18 (17)	21 (20)
Eccentric left ventricular hypertrophy (n [%])	26 (24)	18 (17)
Concentric left ventricular hypertrophy (n [%])	34 (32)	29 (27)
Data are mean ± 1 SD of echocardiographic data and number of patients and percentage of the studied population obtained in 107 hypertensive patients; *p<0.001 compared with baseline

The evolution of LV geometry in the 107 essential hypertensives is shown in Table III. Twenty‐five of 29 patients with normal LVM at the beginning of the study had normal LVM after 1 year of treatment. Three patients developed concentric remodeling, while concentric hypertrophy was detected at the end of the study period in one subject. Three of 18 patients with Echo evidence of concentric remodeling at baseline normalized their LVM at the end of the 12‐month follow‐up period, whereas eight continued to show Echo evidence of concentric remodeling. In addition, one subject developed eccentric hypertrophy and six had concentric hypertrophy. LV was normalized in only eight of 26 patients with eccentric hypertrophy at baseline, whereas at the end of the study, two patients showed Echo signs of concentric remodeling, eight others demonstrated concentric hypertrophy, and eight remained with eccentric hypertrophy. Finally, LVM was normalized in three out of 34 patients with concentric LVH at baseline, whereas eight had concentric remodeling, nine showed eccentric LVH, and in 14 patients, concentric LVH was still found after 1 year of treatment.

Table III.

Evolution of Left Ventricular Geometric Patterns After 1 Year of Treatment

			i Year
	Baseline (n [%])	Normal LV (n)	Concentric Remodeling (n)	Eccentric LVH (n)	Concentric LVH (n)
Normal LV	29 (27)	25	3	0	1
Concentric R	18 (17)	3	8	1	6
Eccentric LVH	26 (24)	8	2	8	8
Concentric LVH	34 (32)	3	8	9	14
BP <140/90 mm Hg (n [%])		22 (56)	11 (52)	4(22)	12 (41)
LV=left ventricle; LVH=left ventricular hypertrophy; R=remodeling; BP=blood pressure

Antihypertensive therapy, which decreased BP significantly in 107 patients during a 12‐month period, was associated with an improvement in the LVM geometry in 31% of the treated subjects, whereas it had no effect on 51%. LV geometry worsened in 18% of the subjects included in the study. Table III also shows the percentage of the patients with optimal control of BP (systolic and diastolic pressure <140 mm Hg and <90 mm Hg, respectively) at the end of the follow‐up, according to their final LV geometry. BP control was not as good in the groups with either concentric or eccentric LVH as in the group with normal LVM (p<0.05). At the end of the follow‐up period, among the 49 well‐controlled subjects, only 16 patients (33%) showed LVH by Echo measures. Among the 58 uncontrolled patients, the prevalence of LVH after 1 year of treatment was 40% (23 patients).

DISCUSSION

Clinical trials conducted in thousands of hypertensive patients have demonstrated an incremental benefit of drug therapy on the reversal of surrogate end points such as proteinuria and cardiac remodeling. Several epidemiologic and observational studies have reported that reversal of cardiac hypertrophy is a predictor of reduced morbidity and stroke mortality. Although a majority of published studies have reported the effects of antihypertensive drugs on LVM expressed as the reduction of LVMI from baseline, few studies have reported the percentage of patients who normalize their LVM with antihypertensive treatment. The present study confirms that antihypertensive therapy is effective enough to reduce BP in a majority of hypertensive subjects during a 1‐year treatment period and is associated with significant reductions in LVMI. However, a detailed analysis of the changes in LV geometry showed that the percentage of patients with normal LV geometry was substantially unchanged 12 months after initiation of antihypertensive therapy. Our data showed that LVM was normalized in only 14 patients with pathologic LV geometry at baseline. And, because four patients with normal LVM at baseline developed either concentric remodeling or concentric LVH during the follow‐up period, the percentage of patients with normal LV geometry 1 year after initiation of therapy did not change significantly. The failure of antihypertensive therapy to control BP may not entirely explain these findings since, in our study, 46% of the patients enrolled in the protocol had their BP controlled below 140/90 mm Hg. A limitation of this observation is that interim BPs were not known.

LVH in essential hypertensives is associated with an increased risk of CV events, and its reduction is associated with a decrease of major CV complications. ⁷ , ⁸ , ⁹ LV remodeling is associated with an increase in CV risk and hypertension‐related clinical events. ⁷ In a 4‐year study, Yurenev et al. ⁷ reported that the frequency of CV complications in patients with no changes in LVM was significantly higher than in the 230 patients in whom LVM decreased by an average of 30 g. In a study by Muiesan et al., ⁸ 151 hypertensive patients were studied before and after an average of 10 years of follow‐up. The rate of CV events was higher in subjects showing no evidence of LVH regression when compared with those maintaining normal LVM. In their study, 66 out of the 151 patients (44%) had LVH at baseline, and after the 10‐year follow‐up, LVH was still present in 27% of the subjects.

In a more definitive study, Verdecchia et al. ⁹ found a prevalence of LVH (LVMI >125 g/m ² ) of 26% among 430 essential hypertensives. After a mean follow‐up of 3.2 years, 54% out of the 112 patients with LVH still had an enlarged LV. The rate of CV events was significantly lower in the group of patients with regression of LVH (LVMI <125 g/m ² ) than in those whose LVMI remained above 125 g/m ² . In contrast, Cipriano et al. ¹³ found no differences between the incidence of CV complications and LVH regression during a 7.4‐year follow‐up period, in part because of the low number of CV events recorded in the study. Although not statistically significant, Cipriano et al. ¹³ reported a tendency for increased CV events in patients with LVMI >125 g/m ² .

Finally, Koren et al. ¹⁴ found a prevalence of LVH of 22% among 172 essential hypertensives. After a mean follow‐up of 5.5 years, 16 of 38 patients with LVH had a reduction of LVM to normal levels, but 20 of 134 patients with an initially normal LVM developed LVH. However, greater BP reduction between baseline and follow‐up was observed in patients whose LVM either did not change or was decreased when compared with those with an increased LVM (−7.7 mm Hg vs. −0.9 mm Hg; −5.3 mm Hg vs. −1.5 mm Hg). The rate of CV events was significantly lower in the patients with no evidence of LVH on follow‐up Echo (9.2% vs. 28.6%; p<0.004). While these studies collectively suggest a benefit of antihypertensive therapy on reversing LV remodeling, the benefit is not universal since, in most of the studies, LVH was not changed. However, as in our study, interim BP control was not always established. It is conceivable that some patients whose BP was <140/90 mm Hg at the end of a study had remained hypertensive for a large part of the follow‐up period. This interpretation agrees with the conclusions obtained by Verdecchia et al., ⁹ who showed no changes in LVH in 14% of the subjects studied; Cipriano et al. ¹³ (38%); Muiesan et al. ⁸ (27%); and those investigated by Koren et al. (58%). ¹⁴

In the evaluation of whether any one particular class of antihypertensive medication is better than any other in preventing CV events, the Losartan Intervention for End‐Point Reduction (LIFE) ¹⁵ study suggested that the choice of BP‐lowering drugs matters since patients randomized to a losartan‐based regimen had fewer strokes than those assigned to an atenolol‐based regimen. In this large study, the effects of the therapies on LVH (assessed by electro‐cardiography) showed that although both therapies were equivalent in their degree of BP control, regression of LVH was significantly better in subjects randomized to losartan. In the LIFE study, eccentric and concentric hypertrophy were the most common abnormal geometric patterns, while fewer than 20% had normal LV geometry. ¹⁶ , ¹⁷ Twelve months into the study, the decrease of LVMI in the LIFE study was 14 g/m ² (compared with 13 g/m ² in our study). A reduction in both septal and posterior wall thickness associated with small changes in LV internal dimension accounted for the decrease in LVMI in both the data reported in the LIFE study ¹⁶ , ¹⁷ and in the present study. While the data obtained in the LIFE study do not contradict our findings, it is important to note that the Echo criteria used in the LIFE study to define both LVH and RWT were different than those employed in this study (>104 g/m ² in women and >116 g/m ² in men for LVH and >0.430 for RWT).

In our study, 60 out of 107 hypertensives (56%) had LVH at baseline and 18 patients had concentric remodeling. At the end of the 1‐year follow‐up, 47 patients (44%) still had LVH and 21 exhibited concentric remodeling, a situation clearly associated with an increase of CV risk. ³ , ¹⁸ Thus, despite a significant reduction of LVMI after 1 year of treatment, there were a number of patients with an increased risk due to the LV geometric pattern. We found that 51% of the patients enrolled in our study did not modify their LVM pattern and 31% showed an improvement in LVM profile, whereas 18% had an unfavorable evolution of LVM.

We do not know why antihypertensive treatment did not produce regression of LVH in all hypertensive patients. Although LVM is associated with a load‐dependent effect, the prevalence of LVH in patients achieving BP control (<140 and <90 mm Hg) was lower (33%), although not statistically significant in difference from those in patients with unsatisfactory BP control. These data suggest that better BP control does reduce LVH. Similar results were reported by Cuspidi et al. ¹⁹ in 700 patients treated and followed for at least 6 months. In this study, the prevalence of LVH in well‐controlled patients was 15% compared with 33% in patients with BP levels > 140/90 mm Hg. In keeping with other findings, our study showed that BP control was worse in patients with LVH compared with those presenting with a normal LVM. These data indicate that LV enlargement is a manifestation of poor BP control. ²⁰

CONCLUSIONS

Our results confirm that antihypertensive treatment reduces LVMI, although LVH or concentric remodeling was still present in a significant number of our patients at the end of the follow‐up period. Because the rate of major CV complications of hypertension is higher in patients who do not achieve regression of LVH, these patients should be considered as a subgroup at high risk and should be treated more aggressively to obtain a better control of BP. Further studies with careful interim BP determinations aimed to clarify why antihypertensive treatments do not achieve regression of LVM in a significant number of hypertensive patients are needed.