Inappropriate Left Ventricular Mass Changes During Treatment Adversely Affects Cardiovascular Prognosis in Hypertensive Patients
Inappropriate left ventricular mass (LVM; ie, the value of LVM exceeding individual needs to compensate hemodynamic load) predicts the risk of cardiovascular (CV) events, independent of risk factors, either in the presence or in the absence of traditionally defined LV hypertrophy. The relation between changes in appropriateness of LVM during antihypertensive treatment and subsequent prognosis was evaluated in 436 prospectively identified uncomplicated hypertensive subjects, with a baseline and follow-up standard clinical evaluation, laboratory examinations, and echocardiogram (last examination: 6±3 years apart), followed for additional 4.5±2.5 years. The appropriateness of LVM to cardiac workload was calculated by the ratio of observed LVM to the value predicted for individual sex, height, and stroke work at rest. At baseline, low or appropriate LVM (≤128% of predicted) was found in 178 patients, and 258 had inappropriate LVM. A first CV event occurred in 82 patients. Event rate (×100 patient-years) was 3.18 among patients with inappropriate LVM persistence (n=152), 0.97 among patients with inappropriate LVM regression (n=104), 1.87 among patients with inappropriate LVM development (n=75), and 0.81 among patients with persistence of appropriate LVM from baseline to the follow-up (n=105; log-rank test: P=0.0001). Cox’s proportional hazard model, considering all of the known CV risk factors, indicated that age, male sex, persistence, or development of inappropriate LVM, in addition to persistence and development of LVH, were independently associated with the occurrence of CV events (P<0.001). The presence of inappropriate LVM during antihypertensive treatment may adversely influence subsequent CV prognosis.
In hypertension, left ventricular (LV) hypertrophy (LVH) is initially a useful compensatory process to abnormal loading conditions, but it is also the first step toward the development of overt clinical disease.1–4 In the attempt to discriminate between normal (compensatory) and abnormal (noncompensatory) increase in LV mass (LVM), it has been proposed recently to evaluate LVM increase in hypertensive patients taking into account gender and cardiac loading conditions rather than some measure of body size.5 Patients with inappropriate LVM, exceeding the amount needed to adapt to stroke work for a given gender and body size, tend to cluster with metabolic risk factors.6 More interestingly, prevalence of low systolic myocardial function, as well as of abnormal relaxation, is greater in hypertensive patients with inappropriate LVM, suggesting that this condition may represent an accelerated phase of transition from compensatory LVH toward heart failure.7,8 Two studies have demonstrated that the presence of inappropriate LVM implies a greater risk of cardiovascular (CV) events, either in the presence or in the absence of traditionally defined LVH.9,10
The reduction of echocardiographically determined LVM, along with the normalization of LV geometry, during antihypertensive treatment, has been associated with a reduction in risk for subsequent CV disease.11,12 No data are presently available on changes in appropriateness of LVM during antihypertensive treatment in hypertensive patients. Therefore, we aimed to investigate the relation between changes in appropriateness of LVM during antihypertensive treatment and subsequent prognosis for CV events in a large group of prospectively identified essential hypertensive patients with and without LVH undergoing usual medical treatment.
We identified uncomplicated hypertensive subjects, selected from an ongoing prospective registry of morbidity and mortality, including patients referred for high blood pressure (BP) diagnostic workup. At entry, all of the patients were never treated (n=247) or had withdrawn for ≥4 weeks before antihypertensive treatment (n=189); they had clinic systolic BP ≥140 and/or diastolic BP ≥90 mm Hg on ≥3 visits at 1-week intervals and no clinic or laboratory evidence of heart failure, renal failure, coronary heart disease, valvular heart disease, and secondary hypertension or severe medical illness, including diabetes mellitus.
At baseline and follow-up visits, patients underwent a through clinical examination. CV risk factors were carefully assessed, and a documented clinical history was collected. BP was measured using a mercury sphygmomanometer (Korotkoff phase V for diastolic BP), with a regular adult cuff, after the echocardiographic examination in a semirecumbent position and thereafter in a sitting position (3 measurements in each position).
In all of the patients, a technically optimal M-mode echocardiogram (under 2D control) was obtained at baseline, and the patients were included into the study. M-mode echocardiographic tracings were recorded on a videotape and printed on a strip-chart recorder. Two different readers measured all of the echocardiograms, before and after treatment, in a blind manner; the average of the 2 calculations was considered for the study. The LV internal dimensions, interventricular septum, and posterior wall thicknesses were measured according to the recommendations of the American Society of Echocardiography,13 and derived anatomic variables were calculated14; LVM index (LVMI) was obtained by normalization of LVM for height to the 2.7 power,15 and LVH was prospectively defined as a value of LVMI ≥50 g/m2.7 in men and 47 g/m2.7 in women.15,16 End-diastolic relative wall thickness (RWT) was calculated as the ratio of posterior wall thickness to one-half LV internal dimension,17 and concentric geometry was defined as RWT ≥0.44. Ejection fraction and endocardial and midwall fractional shortening were calculated by standard methods.7,12 LV end-diastolic and end-systolic volumes were calculated with the Teicholz’s correction of the cube formula.18,19 Stroke work was estimated as systolic BP (measured in a semirecumbent position after the echocardiographic study) times stroke volume and converted in gram meters by multiplying by 0.0144. The theoretical value of LVM was estimated using an equation previously developed5: predicted LVM (pLVM)=55.37+ 6.64×height (m2.7)+0.64×stroke work−18.07×sex (where sex was coded as male=1 and female=2).
LVM measured by M-mode tracing14 (observed LVM [oLVM]) was divided by the predicted LVM and was expressed as a percentage (oLVM/pLVM). oLVM/pLVM was categorized using the 95th percentiles of the distribution in the normotensive, normal weight reference adult population,5 and inappropriate LVM was defined as an excess of >28% from the predicted value (ie, oLVM/pLVM >128%). In 200 participants in the Reliability of M-Mode Echocardiographic Studies (RES) Study,20 it has been demonstrated that a difference of −22 or +20% in the ratio of oLVM/pLVM, measured on 2 different days, has a 90% probability of a true biological change.
Family doctors, who prescribed standard lifestyle and pharmacological medication, followed up all of the patients. Results of baseline tests were made available to family doctors, although it cannot be established to what extent they might have influenced the patients’ management.
A follow-up examination was performed, including clinical evaluation, standard laboratory tests, and an echocardiogram of optimal or suboptimal quality, after a median period of 60 months. Thereafter, direct contact or telephone interviews were regularly undertaken with patients, relatives, and physicians to ascertain the incidence of major medical events or complications of hypertension for an additional median period of 48 months. The comparison between the baseline and follow-up echocardiogram was done before the occurrence of the first nonfatal or fatal event, if any, and none of the patients had developed a CV morbid event at the time of the follow-up visit.
The institution committee on human research approved the study protocol, and all of the subjects gave informed consent to the study. The procedures followed were in accordance with institutional guidelines.
CV events were considered sudden death, fatal and nonfatal stroke, transient cerebral ischemic attack, fatal and nonfatal myocardial infarction, new-onset angina requiring hospitalization, progressive heart failure requiring hospitalization, coronary artery by pass or angioplasty, carotid endarterectomy, and renal failure requiring dialysis. Two independent medical members of the local ethical committee, who were unaware of the echocardiographic data recorded, decided the attribution of CV events to patients, on the basis of available hospital records and all of the other documents collected.
Data were stored and analyzed with the SPSS 13.0 statistical software package. All of the data are expressed as mean±SD. Differences among prospectively defined subgroups were analyzed by ANOVA (Scheffè posthoc) and χ2 tests when appropriate. P<0.05 was considered statistically significant.
Event-free survival analysis was done with the Kaplan-Meier method, and the groups were compared using the Mantel test. The relative importance of each prognostic factor, adjusting for the others, was assessed using Cox’s proportional hazard model, with stepwise procedure. Age, sex, smoking habits (never and former versus current smokers) glucose and cholesterol levels, body mass index, clinic systolic and diastolic BP at baseline, their change from baseline to follow-up (considered as a categorical variable, ie, BP <140/90 mm Hg during treatment), ejection fraction, changes from baseline of RWT, and changes of appropriateness of LVM from baseline to follow-up (categorical variable) were considered as covariates. When changes in traditionally defined LVH from baseline to follow-up (categorical variable) were used in the Cox model, Cox-Snell residuals (individual cumulative hazard function estimate) were generated and compared among patients with inappropriate LVM regression or with appropriate LVM persistence and with inappropriate LVM development or persistence.
A total of 522 patients were included into the study; among these patients, 20 patients died (12 for a CV event), 7 had a morbid CV event before the follow-up evaluation, 49 refused to undergo the follow-up visit, and 10 had a follow-up echocardiogram of suboptimal quality and were excluded from the analysis. Therefore, 436 patients (249 males and 187 females; age range: 18 to 71 years; mean age: 52±9) completed the follow-up. Baseline demographic characteristic of patients lost and of those re-evaluated at follow-up did not differ (data not shown).
β-Blockers, diuretics, calcium-channel blockers, and angiotensin-converting enzyme inhibitors were used alone in 14%, 4%, 15.5%, and 12.5% of patients, respectively, and in various combinations in 42% of patients (we considered the drug class taken for the longest period of treatment during follow-up). All of the patients were treated at the follow-up visit, although in 12% of patients, antihypertensive treatment was temporarily withdrawn and changed several times.
CV Morbidity and Mortality
Eighty-two patients had a documented CV event occurring after the follow-up echocardiographic examination (Table 1). In patients who experienced CV events, age, systolic BP, RWT, LVMI, and the prevalence of patients with inappropriate LVM were greater, whereas midwall fractional shortening was lower at baseline and at follow-up than in patients without events (Table 2).
LVM Appropriateness and CV Risk
At baseline, 170 patients had appropriate LVM, and 258 had inappropriate LVM. Eight patients with low LVM (<73% of predicted) were included in the group of patients with appropriate LVM, according to the results of the MAssa Ventricolare sinistra nell Ipertensione arteriosa (MAVI) Study.10
LVH was present at baseline in 262 patients (61 with appropriate LVM and 201 with inappropriate LVM); of 174 patients without LVH, 117 patients had appropriate LVM, and 57 had inappropriate LVM (Tables S1 and S2, available online at http://hyper.ahajournals.org). CV events occurred in fewer patients with low and appropriate LVM than in those with inappropriate LVM (12% versus 24%; χ2 P=0.02), both in patients with (13.1% versus 25.4%; χ2 P=0.05) or without traditionally defined LVH (11% versus 17.5%; χ2 P=0.24).
Changes in Appropriate LVM and CV Events
According to the presence of inappropriate LVM at baseline and at the end of follow-up, patients were divided in 4 groups: with persistence of appropriate LVM, with regression of inappropriate LVM, with development of inappropriate LVM, and with persistence of inappropriate LVM. Patients with inappropriate LVM regression were younger, whereas patients with persistence of inappropriate LVMI had a higher body mass index and were predominantly males (Table 3). Systolic and diastolic BP were higher in patients with persistence of appropriate LVM. BP values decreased from baseline to follow-up in all of the groups and remained slightly but significantly higher in patients with persistence of appropriate LVM and with regression of inappropriate LVM compared with patients with persistence of inappropriate LVM.
A greater LVMI and RWT were observed in patients with persistence of inappropriate LVM both at baseline and at follow-up (Table 4). A significant decrease of LVMI and of RWT was observed in patients with inappropriate LVM regression but also in those with inappropriate LVM persistence.
Stroke volume, endocardial and midwall fractional shortening, and ejection fraction were increased from baseline to follow-up in patients with regression of inappropriate LVM, whereas they did not change in patients with persistence of appropriate LVM and were decreased in patients with development of inappropriate LVM.
The rate of CV events in patients with persistence of inappropriate LVM was higher in respect to that observed in patients with a persistence of appropriate LVM (Table 1 and Figure 1) and did not differ in respect to those with regression of inappropriate LVM; no difference was observed between patients with persistence and those with development of inappropriate LVM. Life-table analyses showed a significant difference in event-free survival among subgroups (Figure 2).
Outcome was not related to the class of drugs used during follow-up; in the group of patients with persistence or development of inappropriate LVM, a greater percentage of patients receiving association treatment was observed (48 versus 36%; χ2 P=0.015), whereas no differences among monotherapy regimens were noted. At follow-up, appropriate LVM was found in 209 patients and inappropriate LVM in 227 patients. LVMI was normal in 233 patients (160 with appropriate LVM, whereas 73 had inappropriate LVM), and LVH in the presence of appropriate and inappropriate LVM was observed in 49 and 154 patients, respectively.
In patients with development or persistence of LVH at follow-up, a higher number of events was observed in subjects with inappropriate LVM as compared with those with appropriate LVM (33% versus 14%; P=0.01). A similar trend was observed in patients without LVH at follow-up, although it did not reach statistical significance (15% versus 8%; P=0.11 in patients with inappropriate and appropriate LVM, respectively), possibly because the small number of events in this subgroup (n=23).
In a multivariate Cox proportional hazard analysis, the risk of CV events was 2 times higher (95% confidence interval: 1.15 to 3.57) in patients with development and persistence of inappropriate LVM (clustered versus patients with persistence of appropriate LVM or regression of inappropriate LVM), independent of age (adjusted hazard ratio: 1.07; 95% confidence interval: 1.04 to 1.11), male sex (adjusted hazard ratio: 1.79; 95% confidence interval: 1.08 to 2.97), and persistence or development of traditionally defined LVH (clustered versus persistence of normal LVM or regression of LVH; adjusted hazard ratio: 1.8; 95% confidence interval: 1.04 to 3.008). Other covariates did not enter the model.
Cumulative hazard functions were individually computed (Cox-Snell residuals) from a Cox model including all of the previously listed covariates, except changes in LVM appropriateness. The cumulative hazard function was significantly higher in patients with development and persistence of inappropriate LVM than in patients with inappropriate LVM regression and with persistence of appropriate LVM (0.23±0.20 versus 0.14±0.15; P=0.001).
The main result of this study is that changes in the appropriateness of LVM from baseline to follow-up during treatment have a prognostic value, because they may predict the risk for a subsequent CV event in uncomplicated essential hypertensive patients. The predictive value of changes in inappropriateness of LVM in this study seems to have some advantage to that obtained by the traditional approach based on LVM normalized for body height and on geometry of the left ventricle,9–12,21 particularly in patients with increased LVMI; this finding could have been influenced by the lower event rate observed in patients with persistence of normal LVM or with regression of LVH. The changes of the deviation of LVM from the values predicted maintained prognostic value independent of age, sex, and of changes in LVH (according to standard partition values of LVMI), indicating that this approach may help in discriminating between the amount of LVM adequate to compensate the hemodynamic load from the amount in excess to loading conditions (and, therefore, inappropriate or not compensatory).5
The mechanisms by which persistence or development of inappropriate LVM may predispose to the occurrence of events and, conversely, by which regression of inappropriate LVM is associated with an improvement in CV prognosis, cannot be assumed from our study. It has been shown previously that the excess of LVM that is inappropriate to cardiac workload is associated with a greater prevalence of concentric geometry, impaired systolic performance measured at the midwall, and more marked alterations of diastolic relaxation and filling.7,8,22–24 In patients with inappropriate LVM, stroke volume is lower, suggesting that inappropriate LVM may better represent a phenotype of the transition phase from compensatory increase of LVM toward heart failure, in the absence of LV remodeling secondary to ischemic heart disease.8 In patients with inappropriate LVM, lower clinic BP values were observed, possibly reflecting an early decline in cardiac pump function, as found previously in other studies.7,22
One speculation is that the excessive growth of LVM is associated with changes in myocardial structure, with a disproportionate increase in extracellular matrix and myocardial fibrosis.25 Myocardial fibrosis could determine abnormalities in coronary microcirculation blood supply, contributing to myocardial ischemia,26 diastolic filling, relaxation impairment, and favoring heart failure and/or supraventricular arrhythmias. In our study we could observe that, in patients with regression of inappropriate LVM, a significant increase in stroke volume was observed, whereas the opposite occurred in patients with inappropriate LVM development, possibly influencing the incidence of heart failure; 13 heart failure events occurred in patients with persistence or development of inappropriate LVM, and only 1 occurred in patients with persistently appropriate LVM.
First, in this study, the prevalence of inappropriate LVM is quite higher than reported in other studies, and similar to the prevalence (46%) observed in the Losartan Intervention For Endpoint reduction in hypertension (LIFE) Study.7 In the Hypertension Genetic Epidemiology Network population22 the prevalence was 15%, although a more strict partition value was used to identify inappropriately high LVM (>133% instead of 128%). All of the patients were not receiving any treatment at baseline, whereas in other studies,10,22 the influence of antihypertensive treatment cannot be excluded. In our study, patients were seen for the first time in the outpatient clinic, and selection of more severe patients might have occurred, as shown by the high prevalence of traditionally defined LVH; in addition, coronary artery disease was excluded by a negative exercise test, and transmitral flow Doppler parameters for diastolic dysfunction diagnosis were obtained only in a portion of subjects. The use of M-mode echocardiography might have also influenced patients’ selection.27 However, in this sample of hypertensive patients of young age and with a low prevalence of obesity, only 5% of patients at baseline and 2.2% at follow-up were excluded because of an echocardiogram of suboptimal quality.
Second, the calculation of appropriate LVM from echocardiography has been demonstrated to correlate with angiocardiographically determined LVM,25 and the calculation of stroke volume using Teicholz’ correction of the cube formula has been compared with Doppler-derived stroke volume with a mean difference of 2 mL per beat in a large population sample.28 In this study, modifications of the ob/p LVM were >20% in all of the patients with regression or development of inappropriate LVM, suggesting true biological relevance.20
Third, BP values at baseline and follow-up or their changes during the follow-up were not associated with CV risk in these patients. Patients with inappropriate LVM seem to have higher mean systolic and diastolic BP, measured by 24-hour ambulatory monitoring in comparison with patients with appropriate LVM,6 although this finding needs to be confirmed in a larger number of patients. Higher 24-hour BP could have explained, at least in part, the higher incidence of cerebrovascular events.
Lastly, in this study, we could not properly assess the specific influence of antihypertensive treatment, prescribed in a general practice setting, and it was not possible to relate outcome to the class of antihypertensive drugs used. Echocardiographic studies evaluating reversal of LVH with the use of different antihypertensive drugs have indicated that angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, and calcium antagonists may be more effective than β-blockers and diuretics in reducing LVM, for similar BP reduction,29 although BP control may require the use of antihypertensive agents in combination.
The results of this study have shown for the first time that regression of echocardiographically determined inappropriate LVM is associated with an improvement in prognosis. The evaluation of changes in LVM appropriateness may add prognostic information mainly in patients with persistence or development of traditionally defined LVH.
In hypertensive patients, the evaluation of LVM increase taking into account sex and cardiac loading conditions identifies patients at higher CV risk. The prognostic value of changes of echocardiographically determined appropriateness of LVM supports the role of this approach, in addition to traditionally defined LVH. Future research should more deeply investigate the meaning of inappropriate LVM in other groups of subjects.
- Received January 10, 2007.
- Revision received January 27, 2007.
- Accepted February 24, 2007.
Frohlich ED. Risk mechanisms in hypertensive heart disease. Hypertension. 1999; 34: 782–789.
Agabiti Rosei E, Muiesan ML. Hypertensive left ventricular hypertrophy: pathophysiological and clinical issues. Blood Pressure. 2001; 10: 1–11.
de Simone G, Devereux RB, Kimball TR, Mureddu F, Roman MJ, Contaldo F, Daniels S. Interaction between body size and cardiac workload. Influence on left ventricular mass during body growth and adulthood. Hypertension. 1998; 31: 1077–1082.
Palmieri V, de Simone G, Roman MJ, Schwartz JE, Pickering TG, Devereux RB. Ambulatory blood pressure and metabolic abnormalities in hypertensive subjects with inappropriately high left ventricular mass [published erratum appears in Hypertension. 2000;36:147] Hypertension. 1999; 34: 1032–1040.
Palmieri V, Watchell K, Gerdts E, Bella J, Papademetriou V, Tuxen C, Nieminen MS, Dahlof B, de Simone G, Devereux RB. Left ventricular function and hemodynamic features of inappropriate left ventricular hypertrophy in patients with systemic hypertension: the LIFE study. Am Heart J. 2001; 141: 784–791.
de Simone G, Verdecchia P, Pede S, Gorini M, Maggioni AP, on behalf of the MAVI investigators. Prognosis of inappropriate LV mass in hypertension The MAVI study. Hypertension. 2002; 40: 470–476.
Muiesan ML, Salvetti M, Monteduro C, Bonzi B, Paini A, Viola S, Poisa P, Rizzoni D, Castellano M, Agabiti-Rosei E. Left ventricular concentric geometry during treatment adversely affects cardiovascular prognosis in hypertensive patients. Hypertension. 2004; 43: 731–738.
Sahn DJ, DeMaria A, Kisslo J, Weyman A. The Committee on M-mode Standardization of the Amercican Society of Echocardiography: recomendations regarding quantitation in M-mode echocardiography. Results of a survey of echocardiographic measurements. Circulation. 1978; 58: 1072–1083.
Wallerson DC, Ganau A, Roman MJ, Devereux RB. Measurement of cardiac output by M-mode and two-dimensional echocardiography; application to patients with hypertension. Eur Heart J. 1990; 11 (suppl 1): 67–78.
Muiesan ML, de Simone G, Ganau A, Longhini C, Verdecchia C, Mancia G, Agabiti-Rosei E, on behalf of the Working Group on Heart and Hypertension of Italian Society of Hypertension. Inappropriate LV mass: reliability and limitations of echocardiographic measurement for risk stratification and follow-up in single patients: the RES study. J Hypertens. 2006; 24: 2293–2299.
Palmieri V, Watchell K, Bella JN, Gerdts E, Papademetriou V, Nieminen MS, Dahlof B, Roman MJ, Devereux RB. Usefulness of the assessment of the appropriateness of left ventricular mass to detect left ventricular systolic and diastolic abnormalities in absence of echocardiographic left ventricular hypertrophy; the LIFE study. J Hum Hypertens. 2004; 18: 423–430.