Left Ventricular Hypertrophy

HomeCirculationVol. 102, No. 4Left Ventricular Hypertrophy Free AccessOtherPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyRedditDiggEmail Jump toFree AccessOtherPDF/EPUBLeft Ventricular Hypertrophy Pathogenesis, Detection, and Prognosis Beverly H. Lorell and Blase A. Carabello Beverly H. LorellBeverly H. Lorell From the Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Mass (B.H.L.), and the Department of Medicine, Baylor College of Medicine, and Veterans Affairs Medical Center, Houston, Tex Search for more papers by this author and Blase A. CarabelloBlase A. Carabello From the Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Mass (B.H.L.), and the Department of Medicine, Baylor College of Medicine, and Veterans Affairs Medical Center, Houston, Tex Search for more papers by this author Originally published25 Jul 2000https://doi.org/10.1161/01.CIR.102.4.470Circulation. 2000;102:470–479When the heart faces a hemodynamic burden, it can do the following to compensate: (1) use the Frank-Starling mechanism to increase crossbridge formation; (2) augment muscle mass to bear the extra load; and (3) recruit neurohormonal mechanisms to increase contractility. The first mechanism is limited in its scope, and the third is deleterious as a chronic adjustment. Thus, increasing mass assumes a key role in the compensation for hemodynamic overload. This increase in mass is due to the hypertrophy of existing myocytes rather than hyperplasia, because cardiomyocytes become terminally differentiated soon after birth. In response to pressure overload in conditions such as aortic stenosis or hypertension, the parallel addition of sarcomeres causes an increase in myocyte width, which in turn increases wall thickness. This remodeling results in concentric hypertrophy (increase in ratio of wall thickness/chamber dimension).According to LaPlace’s Law, the load on any region of the myocardium is given as follows: (pressure×radius)/(2×wall thickness); thus, an increase in pressure can be offset by an increase in wall thickness. Because systolic stress (afterload) is a major determinant of ejection performance, the normalization of systolic stress helps maintain a normal ejection fraction even when needing to generate high levels of systolic pressure.1 Volume overload in conditions such as chronic aortic regurgitation, mitral regurgitation, or anemia engenders myocyte lengthening by sarcomere replication in series and an increase in ventricular volume. This pattern of eccentric hypertrophy (cavity dilatation with a decrease in ratio of wall thickness/chamber dimension) is also initially compensatory, such that the heart can meet the demand to sustain a high stroke volume. However, chronic hypertrophy may be deleterious because it increases the risk for the development of heart failure and premature death.This review will focus on the pathogenesis of pressure- versus volume-overload types of left ventricular hypertrophy (LVH), detection, clinical manifestations, and prognosis. Where possible, observations from human studies will be presented that have led to major insights regarding the pathogenesis of hypertrophy and the potential for its reversal.Pathogenesis of Load-Induced HypertrophyPressure Versus Volume OverloadIt is generally believed that a mechanical signal initiates a cascade of biological events leading to coordinated cardiac growth. If this is true, the signals for volume versus pressure overload are either quite different or result in remarkably different patterns and mechanisms of growth. Within hours after a pressure overload occurs in the heart in vivo, myosin heavy chain synthesis increases by ≈35%; this increase is initially mediated by an increase in translational efficiency2 (Figure 1). In contradistinction, in severe, pure volume overload, as is seen in mitral regurgitation, much of the increase in left ventricular (LV) mass may accrue from a decrease in the myosin heavy chain degradation rate.3 If hypertrophy were perfectly regulated by the mechanical signal using a typical feedback loop, changes in radius, thickness, and pressure would be orchestrated such that wall stress would be constantly normalized. However, this often does not occur. For example, a large myocardial infarction imposes a volume overload on the remaining myocardium, and cardiac dilatation with an increase in LV mass rapidly occurs.4 Although the initial dilatation may be compensatory to maintain stroke volume, adverse remodeling often develops whereby the ventricle becomes progressively more spherical and wall stress increases, perpetuating the dilatation. In individuals with aortic stenosis and hypertension, especially older women, exuberant hypertrophy develops, wall stress is subnormal, and ejection performance is normal or supernormal.56Both forms of hypertrophy are usually accompanied by complex changes in gene reprogramming.7 These changes include the re-expression of immature fetal cardiac genes, including the following: (1) genes that modify motor unit composition and regulation, (2) genes that modify energy metabolism, and (3) genes that encode components of hormonal pathways (eg, atrial natriuretic peptide, angiotensin converting enzyme). In addition, variable or later blunted expression occurs in other genes that modify intracellular ion homeostasis (eg, downregulation of sarcoplasmic reticulum calcium ATPase [SERCA-2], with variable upregulation of the Na+/Ca2+ exchanger), and key parasympathetic and sympathetic receptors are downregulated (eg, downregulation of β1-adrenergic receptors and M2 muscarinic receptors and increase in ratio of angiotensin II AT2 to AT1 receptor subtypes). Some of these switches, such as the increased expression of the slow myosin ATPase isoform β-myosin heavy chain relative to the fast myosin ATPase isoform α-myosin heavy chain, are adaptive and promote a more favorable myoenergetic economy. However, the long-term functional implications of many of the changes in gene expression are still unclear in the context of integrated cardiovascular function in vivo.How Is Mechanical Information Transduced to Cardiac Growth?The essence of hypertrophy is an increase in the number of force-generating units (sarcomeres) in the myocyte. How does an increase in force on the myocyte trigger an increase in force-generating units (sarcomeres) in the myocyte? The implication is that mechanical input is transduced into a biochemical event that modifies gene transcription in the nucleus. An excellent candidate for such a transducer is the focal adhesion complex, whereby integrins connect the internal cytoskeleton of the cell (which is connected to the nucleus) to the extracellular matrix (ECM).8 Multiple tyrosine-phosphorylated kinases and serine-threonine kinases that are implicated in the signaling of hypertrophy can be found in the ECM.9 Although critical proximal steps in mechanosignal transduction are not yet well understood, there is now evidence that the disruption of cell-cell and cell-ECM contact is sufficient in itself to modulate both cell growth and apoptosis (anoikis).10 In chronic hypertrophy, there are changes in integrin expression11 and possible integrin shedding into adjacent ECM,12 which raises the potential for disordered biomechanical signal transduction for growth and suboptimal myocyte-ECM coupling for force generation.Acute biomechanical signal transduction in experimental models is often accompanied by recruitment of the G-protein–coupled neurohormones (such as angiotensin II and endothelin-1), whose activation likely serves to amplify the growth signaling triggered by the mechanical event itself. The current review by Sugden13 discusses current models of growth-signaling pathways activated by G-protein–coupled neurohormones. In clinical medicine, a critical question is whether a proximal neurohormonal signaling molecule serves as a master switch for load-induced hypertrophy.Some have postulated that angiotensin II, via the AT1 receptor, plays a mandatory role in the induction of hypertrophy because this hormone can directly induce the molecular events of early cardiac growth,141516 its synthetic machinery is upregulated in hypertrophied rat17 and human18 myocardium, and it seems to be required for the growth of stretched neonatal myocytes in vitro.19 However, recent studies in experimental hypertrophy20 and the observations that pressure overload produces robust hypertrophy in transgenic mice with AT1a receptor knockout2122 show that angiotensin II is not mandatory for load-induced hypertrophy.The avid search for a signaling molecule that serves as a master switch for clinical hypertrophy recently shifted to calcineurin, a calcium calmodulin-dependent phosphatase. Transgenic mice that overexpress components of the calcineurin signaling pathway develop a hypertrophic phenotype that can be suppressed by pharmacological inhibitors of calcineurin.23 However, calcineurin inhibitors fail to suppress experimental hypertrophy in several animal models2425 and in humans with hypertension after cardiac transplantation.26 Taken together, these experimental animal and human observations suggest that redundant signaling pathways are likely to modulate load-induced hypertrophy, with the potential for recruitment of alternate signaling cascades when a single pathway is suppressed.27Hypertrophy and Connective TissueFor myocyte growth to support an increased biomechanical load, it must be accompanied by coordinated increases in the surrounding architecture of connective tissue and ground substance, as well as the capillary and nerve networks. The connective tissue itself is primarily composed of collagen with smaller amounts of elastin, laminin, and fibronectin. Although collagen types I, III, and V are found in the myocardium, type I comprises ≈85% of the total collagen in the area. The complex collagen weave provides a mechanism for translating individual myocyte force generation into ventricular contraction, it restrains the development of interstitial edema, and it is responsible for much of the ventricle’s passive diastolic stiffness.28In pressure-overload hypertrophy, the increase in collagen production that occurs as an adaptation to overload must be distinguished from pathological collagen deposition, which is characterized by both perivascular and interstitial fibrosis.28293031 It is not clear whether the initiation of reactive fibrosis in some models is, in part, triggered by defective cell-ECM contact, myocardial ischemia, or the local activation of trophic peptides such as angiotensin II, aldosterone, and/or catecholamines, which results in the sequential expression of transforming growth factor-β1, fibronectin, and relative increase in collagen I.293031 Autopsy and biopsy studies of patients with severe chronic hypertension or aortic stenosis frequently show changes in collagen architecture, as well as severe increases in the percentage of fibrosis occupying the myocardium; this fibrosis reaches a maximum at ≈30%.3132Role of Metalloproteineases in Volume-Overload HypertrophyECM remodeling seems to be very different in volume-overload hypertrophy, in which cavity dilatation occurs in part due to both myocyte elongation and changes in collagen cross-linking and the collagen weave.33343536 Dissolution of the collagen weave leads to increased elasticity, muscle fiber slippage, and an increase in chamber size.37 Such dissolution is predominantly related to the activation of matrix metalloproteineases (MMPs), a family of zinc-containing proteins that includes stromalysins, collagenases, gelatinases, and membrane-type MMPs.3839 Observations in animal models394041 and humans42 with end-stage dilated cardiomyopathies show that an increase in MMP activation and a downregulation of localized tissue inhibitors is critical for the changes in the collagen matrix that permit chamber expansion. The role of MMPs is less well understood in concentric hypertrophy; however, preliminary observations from our laboratory show that MMPs are also activated in experimental pressure overload hypertrophy (E. Goldsmith, PhD, et al, written communication, February 2000).Diagnosis of LVHEchocardiographic DetectionPathological hypertrophy may be associated with an absence of symptoms for many years before the development of congestive heart failure or unexpected sudden death. Thus, in contemporary clinical practice and population studies, the diagnosis of LVH depends predominantly on echocardiographic measurements or novel noninvasive imaging techniques. Methods for 2D targeted M-mode echocardiographic measurements of LV dimensions and the calculation of LV mass are standardized and have been reported in detail elsewhere.43 The detection of pathological LVH requires adjustments for sex, height, and body mass. Although multiple studies have offered echocardiographic criteria for LVH, analyses of the large original cohort and offspring subjects (n=6148) of the Framingham Heart Study have provided criteria that are based on a healthy population distribution of LV mass (Table).44 Using these mass/height criteria, the prevalence of LVH in the entire Framingham Study population is 16% in women and 19% in men.44 Echocardiographic LVH is more prevalent than LVH detected by electrocardiography, with overall rates of 17.4% versus 2.4%, respectively.Normal ranges of LV and right ventricular mass have been described in healthy male and female subjects using cine-MRI as well as ultrafast CT. In a recent study of 75 healthy subjects, the upper limit (95% confidence limit) of LV mass normalized to body surface area was 113 g/m2 in men and 95 g/m2 in women.45 The recent review by Lorenz et al45 summarizes normative sex-based values of LV mass reported by additional contemporary MRI and CT studies. In comparison with the Framingham Heart Study, which echocardiographically detected LVH, current novel imaging studies are limited by the much smaller size of the study populations and less robust longitudinal outcome data.Prognostic Implications of LVHAnalyses from the Framingham Heart Study have unequivocally demonstrated the prognostic value of echocardiographically detected LVH. First, echocardiographic LVH identifies a population at high risk for cardiovascular disease. Subjects with LVH are older, more obese, have higher blood pressures, and are more likely to have preexisting coronary disease and depressed LV systolic function (ejection fraction).46 Second, echocardiographic LVH predicts an increased risk of cardiovascular morbidity and death, even after adjustment for other major risk factors (age, blood pressure, pulse pressure, treatment for hypertension, cigarette use, diabetes, obesity, cholesterol profile, and electrocardiographic evidence of LVH).47 In otherwise healthy subjects followed for 4 years in whom LVH was defined as an LV mass adjusted for height of >143 g/m in men and >102 g/m in women, the relative risk of developing cardiovascular disease was 1.49 in men and 1.57 in women for each increment of 50 g/m in LV mass. This increment of LV mass was also associated with a relative risk of cardiovascular death of 1.73 in men and 2.12 in women and a relative risk of all-cause death of 1.49 in men and 2.01 in women.Increased LV mass is also associated with an increased risk for sudden cardiac death,48 which is more pronounced in men than in women. Knowledge of geometric remodeling patterns provides little additional prognostic information beyond LV mass and traditional cardiovascular risk factors.49 A limitation of these data is that the Framingham Heart Study is composed predominantly of white adults. The prevalence of echocardiographic LVH is reported to be higher in black adults, and LVH is associated with a doubling of mortality in both white and black cohorts.50 Ongoing studies, such as the Jackson Heart Studies, may address unanswered issues regarding the variance and development of hypertrophy in black subjects.Detection of Physiological HypertrophyMeasurements of LV mass must be interpreted in the clinical context. In both men and women, age, height, systolic (but not diastolic) blood pressure, and body mass index (a measure of obesity) are highly significant and independent predictors of LV mass.51 Other demographic studies, corrected for sex and height, have shown that body weight is the most powerful independent predictor of LV mass52 ; in individual patients, increases and reductions in body weight are accompanied by changes in LV mass.53 It is not yet known if increases in LV mass associated with obesity confer an increased risk for cardiovascular morbidity or mortality.Cardiac hypertrophy occurs during changes in load and is not deleterious in the following 3 settings: maturation in infancy and childhood, pregnancy, and exercise. It is plausible that a key difference in the biomechanical signals in these states is the intermittence or transient duration of excess load, compared with the sustained load of hypertension or aortic stenosis. In humans, the heart grows in proportion to body growth in a roughly linear relationship, such that the LV weight in grams is 3 to 4 times the body weight in kilograms. Obviously, the 10-fold increase in LV mass that occurs from childhood to adulthood is necessary and not deleterious. During pregnancy, the requirement for an increased stroke volume and cardiac output is accompanied by a substantial increase in LV dimension and mass, which regresses over months in the postpartum period.54 Finally, both the concentric hypertrophy that occurs in the trained athlete who specializes in sports requiring isometric skeletal muscle contraction (ie, weight lifting, wrestling) or the eccentric hypertrophy that occurs in sports requiring isotonic exercise (ie, long-distance running, cycling) are consistent with normal LV systolic and diastolic function.5556 In the Framingham Study, leisure-time physical activity was associated with LV mass in men, but not in women.51 Thus, an increase in LV mass by itself does not result in muscle dysfunction and must be carefully interpreted in the clinical context.Clinical Manifestations of HypertrophyPatients with LVH due to continuous pressure overload (aortic stenosis) or volume overload (mitral regurgitation) may remain in a compensatory phase with no symptoms and normal or near-normal exercise reserve for years. Others have a transition to heart failure that may be due to diastolic dysfunction, systolic dysfunction, or both.Diastolic Dysfunction in Hypertrophy: Mechanisms and DetectionIn patients with LVH, abnormalities in both myocardial relaxation and passive filling have been detected. Myocardial relaxation, which reflects the time course and extent of crossbridge dissociation after systolic contraction, is modified by the load imposed on the muscle,57 the rapid reduction of cytosolic calcium to basal levels, and those factors such as intracellular pH that modify myofilament sensitivity to calcium. The initial rapid fall of cytosolic calcium is achieved by the ATP-dependent sarcoplasmic reticulum pumps (SERCA-2), which move intracellular calcium “uphill” against a concentration gradient into the sarcoplasmic reticulum. The kinetics of these pumps and optimal calcium loading in the sarcoplasmic reticulum are modified both by the ATP-dependent energy charge and the phosphorylation state of the inhibitory regulatory protein phospholamban.5859 A slower phase of extrusion of calcium, that entered during depolarization, depends on the low affinity, high-capacity sarcolemmal Na+/Ca2+ exchanger.60The downregulation of SERCA-2 is nearly ubiquitous in animal models of advanced pressure overload hypertrophy, and compelling evidence shows that changes in SERCA-2 levels have the potential to modify the time course of the calcium transient, myocardial relaxation, and the force-frequency response.61626364 However, the marked reductions in SERCA-2 observed in both models of compensated hypertrophy and end-stage human dilated cardiomyopathy are associated with very divergent degrees of impaired relaxation (as well as systolic performance), which indicate that the reduced expression of this pump cannot be the sole mechanism of impaired function. In humans with dilated cardiomyopathy, reductions in protein levels of SERCA-2 may be partially compensated for by increases in levels of the Na+/Ca2+ exchanger.65 In humans with load-induced hypertrophy, potential coordinated changes in SERCA-2 and the Na+/Ca2+ exchanger have not yet been well characterized, distinct from end-stage heart failure. In addition to these changes, other molecular adaptations have the potential to modify crossbridge attachment and perturb myocardial relaxation; these adaptations include increased β-myosin ATPase activity, changes in troponin subunit isoform expression and phosphorylation, and blunted cAMP-mediated phosphorylation of regulatory proteins such as phospholamban.6667In humans, myocardial relaxation properties are often estimated by surrogate measurements of the first derivative of LV pressure decay (−dp/dt) and by modeling the time course of LV pressure decay between aortic valve closure and mitral valve opening to an exponential function to obtain a time constant (τ) of LV pressure decline. These techniques provide clues to the presence of impaired relaxation but have serious limitations in diseased hearts because isovolumic pressure decay is often not exponential, and the active process of relaxation may be prolonged and dyssynchronous, extending well beyond mitral valve opening.Diastolic FillingThe dynamics of passive LV filling and the relationship between diastolic volume and pressure are influenced by the time course of active relaxation and the passive deformation properties of the myocardium, including the thickness of the wall and its composition, particularly collagen deposition and its architecture. The hypertrophied myocyte in isolation has only a limited role in increasing chamber stiffness.68 Doppler echocardiographic techniques are widely used to assess transmitral valve flow velocity curves, which characterize left atrial emptying and ventricular diastolic filling.6970 Alternatively, radionuclide ventriculography is sometimes used in clinical research to estimate the rates of early and peak ventricular filling.69 The rate and magnitude of diastolic ventricular filling just after mitral valve opening is directly related to the pressure gradient across the mitral valve, which is determined by both left atrial pressure and the active fall of LV pressure to its nadir during this relaxation filling period.Three patterns of LV filling as assessed by Doppler flow velocity curves are helpful in identifying progressively worse diastolic function.6970 These patterns are (1) “slowed relaxation,” which is characterized by reduced early diastolic inflow velocity with a compensatory increase in filling due to left atrial contraction (decreased E/A ratio); (2) “pseudonormalization,” which has a preserved ratio of the contributions of early diastolic filling and atrial contraction (normal E/A ratio) but a rapid deceleration of early mitral inflow; and (3) a “restrictive pattern,” in which almost all filling occurs explosively in early diastole in association with a very short deceleration time, which is suggestive of a high left atrial pressure driving filling into a “stiff” LV. This latter pattern of severe diastolic dysfunction is characterized by an S3 gallop as the auscultatory marker of abrupt cessation of ventricular filling in early diastole and, frequently, increased left atrial regurgitant flow into the pulmonary veins. These patterns must always be interpreted in the context of other clinical features, and they can change abruptly in response to volume overload and exercise, and long-term during normal pregnancy54 and childhood development.71 Controversy and conflicting observations still exist regarding the depression of isovolumic relaxation and filling indices during aging in normal humans in the absence of hypertension and hypertrophy.72737475In athletes with moderate hypertrophy, there is usually no evidence of altered systolic contractile indices or indices of diastolic filling.5556 In contrast, in patients with LVH due to sustained pressure overload (hypertension and aortic stenosis), a hemodynamic hallmark is the elevation of LV end-diastolic pressure relative to a normal or small LV diastolic cavity volume. This decrease in diastolic chamber distensibility is predominantly related to altered passive properties causing an increase in myocardial stiffness, which is described formally by the mechanical stress/strain relationship. Dynamic abnormalities of slowed isovolumic LV pressure decay, as well as slowed early diastolic mitral inflow velocity and ventricular filling with enhanced reliance on atrial transport (decreased E/A ratio), have been described in multiple studies using both invasive and noninvasive technologies; in some patients with advanced hypertrophy, this pattern may evolve to the more severe abnormality of a restrictive pattern of diastolic filling.7677787980In patients with aortic stenosis and regurgitation, combined hemodynamic and biopsy studies suggest that the prolongation of relaxation is closely related to the magnitude of hypertrophy, whereas abnormal increases in myocardial stiffness are more closely related to changes in collagen architecture.32 Abnormalities of relaxation and passive myocardial stiffness usually precede alterations in systolic ejection indices (end-systolic volume and ejection fraction) and are present in ≈50% of patients with pressure overload and normal systolic ejection indices,7677 although more subtle abnormalities, including depressed midwall shortening, may be present.8182Effects of Aging and Sex on Diastolic Function in LVHIn older patients with isolated systolic hypertension, concentric LVH is common.83 Diastolic dysfunction, including the presence of a Doppler filling pattern of impaired relaxation, has been observed in >80% of older hypertensives.84 In patients with aortic stenosis, senescence profoundly influences the pattern of hypertrophic growth and diastolic function. Using hemodynamic studies complemented by morphometric analyses of ventricular biopsies, Villari et al85 compared younger (<60 years) and elderly (>65 years) patients with comparable severities of aortic stenosis and showed that elderly patients with pressure overload were characterized by more severe hypertrophy and interstitial fibrosis, as well as more severe impairment of relaxation, myocardial stiffness, and filling indices. Ejection fraction and midwall shortening were similar in the 2 groups (younger and elderly patients).Gender also influences function in pressure-overload hypertrophy in humans.56 In men and women with aortic stenosis and similar aortic valve areas and gradients, men are more likely to have cavity enlargement, a lower ejection fraction, and increased diastolic myocardial stiffness associated with more severe changes in collagen architecture.86 Sex-based differences in diastolic function are recapitulated in rodent aortic stenosis models in which female animals demonstrate more favorable changes in cardiac geometry,8788 as well as better preservation of normal adult cardiac gene expression.89Prevalence and Prognosis of Heart Failure due to Diastolic DysfunctionIn severe cases of chronic LVH associated with diastolic dysfunction and preserved ejection fraction, both male and female patients may experience episodic severe congestive heart failure and hospitalization. Even in patients with milder degrees of hypertrophy and diastolic dysfunction, an inability to augment LV volume during exercise may severely limit exercise cardiac output and exercise reserve. In a case-control study from the Framingham Heart Study,90 51% of the subjects with congestive heart failure had a normal LV ejection fraction (≥50%). In these patients, diastolic dysfunction may be inferred, but it was not characterized by quantitative echocardiographic-Doppler indices. Women predominated in this group with normal ejection fractions (65%), whereas men predominated (75%) in the cohort with low ejection fractions.Although heart failure patients with normal LV ejection fractions had a lower mortality risk than those with reduced ejection fractions, heart failure patients with normal ejection fractions had an annual mortality of 18.9% versus 4.1% for matched control subjects during the 6-year study period, indicating a >4-fold mortality risk.90 Atrial fibrillation also modifies diastolic dysfunction in patients with LVH. The increased reliance on atrial contraction to fill the stiff left ventricle means that atrial fibrillation is usually very poorly tolerated. In addition, pressure overload from hypertension is responsible for more atrial fibrillation in the population (≈14% of cases) than any other risk factor, and atrial fibrillation confers a 4- to 5-fold increase in the risk of stroke.91Management of Heart Failure With LVH and Diastolic DysfunctionNo randomized clinical trials or evidence-based consensus guidelines exist regarding end points of survival, hospitalization, or quality-of-life to firmly guide the management of patients with LVH and heart failure due to diastolic dysfunction. On the basis of the above experimental insights, clinical observations, and consensus of expert opinion, current therapy is aimed at the following: (1) preserving sinus rhythm and suppressing tachycardia, (2) reducing elevated left atrial and diastolic ventricular pressures without excessively reducing preload and d