Author + information
- Received May 27, 2005
- Revision received September 30, 2005
- Accepted October 17, 2005
- Published online March 7, 2006.
- Iacopo Olivotto, MD⁎,⁎ (, )
- Franco Cecchi, MD⁎,
- Roberto Gistri, MD⁎,
- Roberto Lorenzoni, MD‡,
- Giampaolo Chiriatti, MD§,
- Francesca Girolami, BSc†,
- Francesca Torricelli, BSc† and
- Paolo G. Camici, MD, FACC, FRCP∥,¶
- ↵⁎Reprint requests and correspondence:
Dr. Iacopo Olivotto, Centro di Riferimento per le Cardiomiopatie, Cardiologia San Luca, Azienda Ospedaliera Universitaria Careggi, Viale Pieraccini 19, 50134 Firenze, Italy
Objectives This study sought to evaluate whether the entity of microvascular dysfunction, assessed by positron emission tomography (PET), predicts the long-term development of left ventricular (LV) remodeling and systolic dysfunction in hypertrophic cardiomyopathy (HCM).
Background A subgroup of patients with HCM developed LV dilation and systolic impairment. A causal role of coronary microvascular dysfunction has been suggested as the underlying pathophysiological mechanism.
Methods Fifty-one patients (New York Heart Association functional class I to II) were followed up for 8.1 ± 2.1 years after measurement of resting and dipyridamole (Dip) myocardial blood flow (MBF). Left ventricular systolic dysfunction was defined as an ejection fraction (LVEF) <50%.
Results The Dip-MBF was blunted in HCM patients compared with a group of healthy control patients (1.50 ± 0.69 ml/min/g vs. 2.71 ± 0.94 ml/min/g; p < 0.001). At final evaluation, 11 patients (22%) had an LVEF <50%; in most (n = 7), systolic dysfunction was associated with a significant increase in LV cavity dimensions (>5 mm) during follow-up. These 11 patients showed lower Dip-MBF than the 40 with preserved LV function (1.04 ± 0.38 ml/min/g vs. 1.63 ± 0.71 ml/min/g, respectively; p = 0.001); Dip-MBF was particularly blunted in five patients with clinical progression to severe heart failure symptoms or death (Dip-MBF 0.89 ± 0.15 ml/min/g). At multivariate analysis, the two independent predictors of systolic dysfunction were Dip-MBF in the lowest tertile (<1.1 ml/min/g; relative hazard, 7.5; p = 0.038) and an end-diastolic LV dimension in the highest tertile (>45 mm; relative hazard, 12.3; p = 0.031).
Conclusions Severe microvascular dysfunction is a potent long-term predictor of adverse LV remodeling and systolic dysfunction in HCM. Our findings indicate microvascular dysfunction as a potential target for prevention of disease progression and heart failure in HCM.
Since its early descriptions, hypertrophic cardiomyopathy (HCM) has been characterized by a small, markedly hypertrophied, hypercontractile left ventricle (LV) (1). In a significant minority of patients, however, progressive LV dilation and impairment of systolic function may become prominent clinical features, and lead to the so-called end-stage phase (1–7). Although other causes may concur, this process is thought to represent the consequence of recurrent, diffuse myocardial ischemia attributable to coronary microvascular dysfunction (8–11).
Using positron emission tomography (PET), we have recently shown that severe blunting of the coronary vasodilator response to dipyridamole (Dip), reflecting microvascular dysfunction, is a potent predictor of outcome in a cohort of HCM patients, mostly related to heart failure and its complications (12). That work, however, did not address the relationship between microvascular dysfunction and long-term changes in LV morphology and function, so that such an association remains unproven. Therefore, in the same HCM patient cohort, we aimed to assess whether the severity of coronary microvascular dysfunction was predictive of adverse long-term LV remodeling, systolic dysfunction, and progression to the end-stage phase.
Patients and methods
The study cohort comprises 51 patients from a large regional HCM population, closely followed up at two community-based institutions in Tuscany. The baseline features and clinical outcome of the study group have been previously reported (12). The diagnosis of HCM was based on the echocardiographic evidence of myocardial hypertrophy (wall thickness >15 mm), in the absence of any other cardiac or systemic cause of LV hypertrophy (1). Exclusion criteria were a history of hypertension or coronary artery disease and severe congestive heart failure (New York Heart Association [NYHA] functional class III to IV). All patients had Maron type II or III morphology (1). The mean age was 44 ± 13 years, 36 (71%) patients were male, and 21 (41%) patients were in NYHA functional class II. Fourteen patients (27%) complained of typical angina and were enrolled after documentation of angiographically normal coronary arteries.
The control group included 12 patients with an atypical chest pain syndrome (4 men) ages 51 ± 8 years (p = 0.1 vs. patients). All had a normal physical examination, electrocardiogram, echocardiogram, treadmill exercise test result, and coronary and LV angiograms.
Measurement of myocardial blood flow
All PET scans were performed at the Institute of Clinical Physiology, in Pisa, between June 1990 and May 1993 (9,12). Regional myocardial blood flow (MBF) was measured using PET with nitrogen-13-labeled ammonia under basal conditions and during near maximal hyperemia induced by Dip (0.56 mg/kg, administered intravenously over 4 min) as previously reported (9). For each MBF measurement, a dynamic acquisition was started simultaneously with an intravenous bolus of nitrogen-13–labeled ammonia (0.25 mCi/kg of body weight). The Dip-MBF was measured 50 min after the basal scan and 4 min after the end of Dip infusion, following the same protocol. Absolute regional MBF (in ml/min/g) was calculated as previously reported (9). The study protocol was approved by the research ethics committees of each institution, and written consent was obtained from each patient.
Clinical and echocardiographic follow-up
Patients were followed up for an average of 8.1 ± 2.1 years after the PET scan with clinical and echocardiographic examinations. For each patient, we compared echocardiographic measurements at the time of PET with the last available echocardiogram at the end of follow-up.
Standard measurements of left atrial, LV end-diastolic, and end-systolic diameter were obtained in the parasternal long-axis view (13). In all patients, maximum LV wall thickness values were in the basal or mid anterior septum and were measured in the same region of the LV wall both at the time of PET and at final evaluation. Obstruction of the LV outflow was considered present when a peak outflow gradient of ≥30 mm Hg was present under basal conditions (1). The LV ejection fraction (LVEF) was measured in the standard four-chamber view by the area-length method (13). The LV systolic dysfunction was defined as an LVEF <50%.
For the purpose of this study, patients were grouped into tertiles of Dip-MBF, and the relationship of this ordinal variable with the development of systolic dysfunction was assessed. Data were expressed as mean values ± SD. For the comparison of normally distributed data, the Student ttest was used for single comparisons; one-way analysis of variance was used for comparisons among tertiles of Dip-MBF, followed by the Bonferroni post-hoc test. Chi-square or Fisher exact tests, as appropriate, were used to compare non-continuous variables expressed as proportions. Relative hazard and 95% confidence intervals (CIs) were calculated using a stepwise forward Cox proportional hazard regression model, with an entry probability for each variable set at 0.05. All p values are two-sided and considered significant when <0.05.
MBF and baseline echocardiographic features
Resting MBF was not significantly different in patients and control patients (0.84 ± 0.31 ml/min/g vs. 1.00 ± 0.23 ml/min/g, respectively; p = 0.10). By contrast, Dip-MBF was severely blunted in HCM patients (1.50 ± 0.69 ml/min/g vs. 2.71 ± 0.94 ml/min/g in the control patients; p < 0.001). Table 1shows patients’ features with respect to tertiles of Dip-MBF. At the time of PET, all patients were in NYHA functional class I or II. On average, patients in the lower tertile of Dip-MBF (0.59 to 1.11 ml/min/g) had larger LV cavity dimensions and a lower EF compared with patients in the other two tertiles (Table 1).
LV remodeling and prevalence of systolic dysfunction
Average follow-up was 8.1 ± 2.1 years. Compared with baseline, final evaluation showed an overall increase in LV end-diastolic diameter (+8 ± 10%) and a decrease in LVEF (−15 ± 13%) (Table 1, Fig. 1).At initial evaluation, three patients (6%) had an LVEF <50%; eight additional patients developed systolic dysfunction during follow-up. Thus, the total number of patients with LVEF <50% at final evaluation was 11 (22%); in this group, the average decrease in LVEF from baseline was −27 ± 11% (Table 2).In most (n = 7), development of systolic dysfunction was associated with a >5 mm increase in LV end-diastolic diameter (average, 9 mm); however, four patients showed little variation (<5 mm; average, 3 mm) (Table 2).
At final evaluation, 5 of the 11 patients with LVEF <50% progressed to NYHA functional class III to IV, including two patients who eventually died of heart failure and two who have been evaluated for cardiac transplantation (Fig. 2,Table 2). Among these five patients, the final LV end-diastolic dimension ranged from 44 to 68 mm, and the LVEF from 25% to 46%. Of note, only two patients already had an LVEF <50% at the time of PET (Table 2).
Relevance of MBF impairment to LV remodeling
Average Dip-MBF was significantly lower in the 11 patients in whom LV systolic dysfunction developed, compared with the 40 with preserved LV function (1.04 ± 0.38 ml/min/g vs. 1.63 ± 0.71 ml/min/g, respectively; p = 0.001). The MBF impairment was particularly severe among the five patients whose condition progressed to NYHA functional class III to IV or who died (Dip-MBF, 0.89 ± 0.15 ml/min/g). Prospectively, patients in the lowest tertile of Dip-MBF (≤1.11 ml/min/g) showed a greater increase in LV dimensions and decrease in LVEF during follow-up as compared with each other tertile, even though there was considerable individual variability in each group (Table 1, Fig. 3).Specifically, 9 of the 11 patients with systolic dysfunction belonged to the lowest Dip-MBF tertile, compared with only 2 of the 40 patients with preserved LV function (Table 2). After exclusion of the three patients with LVEF <50% at initial evaluation, the positive and negative predictive values of Dip-MBF ≤1.11 ml/min/g to foretell the development of systolic dysfunction were 40% and 94%, respectively.
In a multivariate Cox regression model including age, gender, NYHA functional class, initial LV dimensions, maximum LV thickness, and magnitude of LV outflow gradient, the only two independent long-term predictors of LV systolic dysfunction were a Dip-MBF in the lowest tertile (relative hazard, 7.5; 95% CI: 1.1 to 50.4; p = 0.038) and an end-diastolic LV dimension in the highest tertile (>45 mm; relative hazard, 12.3; 95% CI: 1.2 to 121.5; p = 0.031).
MBF impairment and LV systolic dysfunction in HCM
The main finding of the present study was that severe impairment of the coronary vasodilator response to Dip is strongly associated with long-term adverse LV remodeling and systolic dysfunction in HCM patients. Over an average follow-up >8 years, patients with a Dip-MBF in the lowest tertile showed a 7.5-fold higher risk of developing systolic dysfunction (defined as an LVEF <50%) compared with the other two tertiles. The only other independent predictor of systolic dysfunction was a baseline end-diastolic LV dimension in the highest tertile (>45 mm), possibly suggesting an early propensity toward cavity dilation.
We previously described the association of an impaired Dip-MBF with increased LV dimensions and reduced fractional shortening in HCM patients studied cross-sectionally (14). Indeed, the baseline data of the present study are consistent with a certain degree of LV remodeling before study enrolment among patients in the lowest tertile of Dip-MBF. However, we now show that even in the presence of a normal LV cavity size and function, the degree of microvascular dysfunction is a potent predictor of remodeling and systolic impairment. Indeed, with a negative predictive value as high as 94%, a Dip-MBF in the lowest tertile seemed to be a necessary prerequisite for long-term impairment of systolic dysfunction in HCM patients, whereas a relatively preserved vasodilator capacity showed a powerful protective effect. Thus, PET studies of microvascular function potentially represent a valuable tool for the identification of HCM patients at risk of severe disease progression, even in individuals with no or mild symptoms and normal LV function (12). This is particularly relevant in the face of preliminary evidence that certain pharmacologic agents, such as verapamil and angiotensin-converting enzyme inhibitors, may positively influence the coronary microcirculation (15,16).
Of note, some patients in the lowest tertile of Dip-MBF already had evidence of relative LV dilation and somewhat reduced systolic function at study entry, which may have emphasized the differences observed during follow-up compared with the other tertiles. For those patients, we could not quantify the morphologic change occurring before enrolment, nor its possible relevance to coronary microvascular function (e.g., because of elevated filling pressures) (8). Nevertheless, when initial LV end diastolic dimensions and LVEF were taken into account in a multivariate model, Dip-MBF remained a powerful independent predictor of systolic dysfunction in our cohort.
End-stage phase progression
The most dramatic clinical consequence of LV remodeling in HCM is represented by progression to the so-called “end-stage” phase, which may be difficult to distinguish from a primary dilated cardiomyopathy, is associated with severe clinical signs of heart failure and may require heart transplantation (1–7). In the present study, all patients with such progression had an extremely impaired vasodilator response to Dip, not exceeding and often below the resting MBF values measured in healthy control patients. Thus, although lesser degrees of LV remodeling also occurred in patients with relatively preserved Dip-MBF, severe degrees of systolic dysfunction, LV dilation, and clinical progression to heart failure were confined to patients with extreme impairment of flow.
We acknowledge that the prevalence of end-stage progression in our study cohort overestimates that in the general HCM population. Indeed, the prevalence of systolic dysfunction was found to be ≤5% in most reports (1). In the overall Florence HCM database, which includes more than 500 patients, 13% of the patients progressed to an end-stage phase over an average follow-up of about 10 years (F. Cecchi and I. Olivotto, unpublished observation, 2005) compared with 22% in the present study. Such a discrepancy suggests an undefined selection bias that led to a preferential inclusion of patients with progressive disease, despite the fact that individuals with overt heart failure were excluded by the study protocol.
Pathophysiology of LV remodeling in HCM
The most likely mechanism by which microvascular dysfunction is implicated in LV remodeling is represented by substantial lowering of the ischemic threshold, increased likelihood of recurrent ischemic damage to the myocardium, replacement fibrosis, chamber dilation, and loss of contractile function (2,3,5,9–12). A similar process is thought to occur in patients surviving acute myocardial infarction, in whom persistent microvascular dysfunction after successful reperfusion represents a powerful predictor of LV dilation and heart failure (17). Of note, although the absolute change in LVEF and cavity dimensions was greater in the lowest tertile of Dip-MBF, considerable variability existed within each tertile. This is consistent with the fact that the individual ischemic burden may vary considerably based on the relevance of triggers such as exercise, arrhythmias, and outflow obstruction (18–19). In addition, pathophysiological mechanisms such as primary fibrosis (20), apoptosis (21), and inappropriately increased adrenergic stimulation (22) may add to interindividual variability.
Finally, although the development of systolic dysfunction was generally associated with progressive LV dilation in our cohort, a minority of patients with progressive loss of systolic function had only a very limited increase in LV cavity size. A potential explanation for this is represented by extensive and diffuse fibrous replacement of the myocardium, leading to loss of contractile function without apparent remodeling of the LV (6). Indeed, systolic dysfunction is not uncommon in the context of a restrictive evolution of HCM, in which LV cavity size is normal or reduced.
Severe microvascular dysfunction is a potent long-term predictor of LV adverse remodeling and systolic dysfunction in HCM, which can be identified well before irreversible morphologic and functional changes occur. Thus, our findings indicate microvascular dysfunction as a potential target for the prevention of disease progression and heart failure in HCM.
Supported by grants from the Italian Ministry for Scientific and Technologic Research (MURST-COFIN 2004) and the Fondazione Ente Cassa di Risparmio di Firenze.
- Abbreviations and Acronyms
- hypertrophic cardiomyopathy
- left ventricle/ventricular
- myocardial blood flow
- New York Heart Association
- positron emission tomography
- Received May 27, 2005.
- Revision received September 30, 2005.
- Accepted October 17, 2005.
- American College of Cardiology Foundation
- Seiler C.,
- Jenni R.,
- Vassalli G.,
- et al.
- Cannon R.O.,
- Rosing D.R.,
- Maron B.J.,
- et al.
- Camici P.G.,
- Chiriatti G.,
- Lorenzoni R.,
- et al.
- Maron B.J.,
- Wolfson J.K.,
- Epstein S.E.,
- et al.
- Cheitlin M.D.,
- Alpert J.S.,
- Armstrong W.F.,
- et al.,
- A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Clinical Application of Echocardiography)
- Mourad J.J.,
- Hanon O.,
- Deverre J.R.,
- et al.
- Bolognese L.,
- Carrabba N.,
- Parodi G.,
- et al.
- Olivotto I.,
- Cecchi F.,
- Casey S.A.,
- et al.
- Lombardi R.,
- Betocchi S.,
- Losi M.A.,
- et al.
- Choudhury L.,
- Guzzetti S.,
- Lefroy D.C.,
- et al.