Author + information
- Received February 6, 2012
- Revision received March 14, 2012
- Accepted March 29, 2012
- Published online September 4, 2012.
- Giancarlo Todiere, MD⁎,
- Giovanni Donato Aquaro, MD⁎,⁎ (, )
- Paolo Piaggi, MS, PhD†,
- Francesco Formisano, MD‡,
- Andrea Barison, MD§,
- Pier Giorgio Masci, MD⁎,
- Elisabetta Strata, MD⁎,
- Lorenzo Bacigalupo, MD‡,
- Mario Marzilli, MD∥,
- Alessandro Pingitore, MD, Phd¶ and
- Massimo Lombardi, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Giovanni Donato Aquaro, Gabriele Monasterio CNR-Tuscany Foundation, Via Moruzzi 1, 56124 Pisa, Italy
Objectives This study sought to assess the rate of progression of fibrosis by 2 consecutive cardiac magnetic resonance (CMR) examinations and its relation with clinical variables.
Background In hypertrophic cardiomyopathy (HCM) myocardial fibrosis, detected by late gadolinium enhancement (LGE), is associated to a progressive ventricular dysfunction and worse prognosis.
Methods A total of 55 HCM patients (37 males; mean age 43 ± 18 years) underwent 2 CMR examinations (CMR-1 and CMR-2) separated by an interval of 719 ± 410 days. Extent of LGE was measured, and the rate of progression of LGE (LGE-rate) was calculated as the ratio between the increment of LGE (in grams) and the time (months) between the CMR examinations.
Results At CMR-1, LGE was detected in 45 subjects, with an extent of 13.3 ± 15.2 g. At CMR-2, 53 (96.4%) patients had LGE, with an extent of 24.6 ± 27.5 g. In 44 patients, LGE extent increased significantly (≥1 g). Patients with apical HCM had higher increments of LGE (p = 0.004) and LGE-rate (p < 0.001) than those with other patterns of hypertrophy. The extent of LGE at CMR-1 and the apical pattern of hypertrophy were independent predictors of the increment of LGE. Patients with worsened New York Heart Association functional class presented higher increase of LGE (p = 0.031) and LGE-rate (p < 0.05) than those with preserved functional status.
Conclusions Myocardial fibrosis in HCM is a progressive and fast phenomenon. LGE increment, related to a worse clinical status, is more extensive in apical hypertrophy than in other patterns.
Clinical presentation of hypertrophic cardiomyopathy (HCM) is very heterogeneous, varying from asymptomatic benign forms to malignant arrhythmogenic or severe progressive expressions, resulting in end-stage HCM (1–3). Myocardial fibrosis is a pathological hallmark of HCM and is considered a substrate for ventricular arrhythmias and for progression to systolic dysfunction (4–7).
Cardiac magnetic resonance (CMR) allows noninvasive detection and quantification of myocardial fibrosis by late gadolinium enhancement (LGE) (8). The presence of LGE is known to be associated with the presence of risk factors for sudden cardiac death, heart failure symptoms, and evolution to ventricular dilation and dysfunction (3).
Recently a study by O'Hanlon et al. (9) demonstrated the prognostic role of LGE in HCM as a predictor of the combined endpoint of cardiovascular death, unplanned cardiovascular admission, sustained ventricular tachycardia or ventricular fibrillation, or appropriate implantable cardioverter-defibrillator (ICD) discharge. Furthermore, Bruder et al. (10) demonstrated LGE as predictor of cardiac death showing higher odds ratio than the conventional risk factors for sudden death.
All these studies demonstrated the prognostic role of the presence of LGE in HCM. However, the rate of progression of LGE in HCM was not prospectively investigated. Objectives of this study were: 1) to assess the amount and rate of progression of LGE by 2 consecutive CMR examinations; and 2) to evaluate the relationship between the rate of progression of LGE and clinical variables.
We enrolled a total of 70 consecutive patients with a diagnosis of HCM in sinus rhythm without contraindications for CMR. All these patients underwent to a first CMR examination (CMR-1). After CMR-1, 15 patients were excluded: 2 patients for known significant coronary artery disease, 8 patients for ICD implantation, 2 patients for permanent atrial fibrillation (occurred after CMR-1), 1 patient for septal myectomy, and finally 2 patients for nonsufficient quality of images. The final population that completed the serial CMR examinations comprised 55 patients. The study was approved by the internal ethical committee of our institute, and all subjects gave their written informed consent. We have certified that they comply with the Principles of Ethical Publishing (11).
All the patients enrolled underwent clinical, electrocardiographic, and echocardiographic evaluation at the time of CMR-1 and CMR-2. The diagnosis of HCM and assessment of the left ventricular (LV) outflow gradient were based on previously reported echocardiographic criteria (12). The conventional primary prevention risk markers for sudden death in HCM were evaluated: family history of sudden death, extreme LV wall thickness (>30 mm), unexplained (nonvasovagal) syncope, nonsustained ventricular tachycardia on ambulatory electrocardiographic Holter recordings (>4 ventricular beats at a heart rate >120 beats/min), and abnormal “flat” systolic arterial pressure during exercise stress test (13). A complete clinical evaluation was performed at the enrollment and repeated at CMR-2. By clinical interrogation, each patient was classified in a New York Heart Association (NYHA) functional class on the basis of the presence and the severity of dyspnea. The history of other symptoms (syncope, chest pain, palpitation) was also recorded. A 12-lead resting electrocardiogram was recorded on the day of the CMR examination. An imaging stress test was performed in all the patients. When the stress test was positive, evaluation of coronary anatomy was performed by angiography or coronary computed tomography angiography. Patients with significant coronary artery disease were excluded from the study.
CMR was performed with a dedicated 1.5-T (Signa Hdx, General Electrics Healthcare, Milwaukee, Wisconsin) with an 8-channel cardiac phased array coil. Short-axis cine images from the mitral plane valve to the LV apex were acquired using a steady-state free precessing FIESTA (fast imaging employing steady-state acquisition) pulse sequence with the following parameters: 30 phases, slice thickness 8 mm, no gap, 8 views per segment, number of excitation 1, field of view 40 cm, phase field of view 1, 224 × 224 matrix, voxel dimensions 1.78 × 1.78 × 8 mm, reconstruction matrix 256 × 256, 45° flip angle, repetition time/echo time equal to 3.5/1.5, and a bandwidth of 125 KHz. In both the first and second examinations, LGE images were acquired 10 min after the administration of Gd-DTPA (Magnevist, Schering-AG, Berlin-Wedding, Germany) with a dosage of 0.2 mmol/kg in short-axis views. An inversion recovery T1-weighted gradient echo sequence was used with the following parameters: field of view 40 mm, slice thickness 8 mm, no gap between each slice, repetition time 4.6 ms, echo time 1.3, 20° flip angle, matrix 224 × 224, reconstruction matrix 256 × 256, number of excitation 1. The appropriate inversion time was set to null normal myocardium (range 250 to 300 ms) and it was the same for both CMR examinations.
Analysis of CMR images was performed, using a commercially available research software package (Mass 6.1, Leiden, the Netherlands). Left ventricular mass was measured by the analysis of the cine short-axis images. The endocardial and epicardial contours of LV myocardium were manually traced in the end-diastolic and the end-systolic phases. End-diastolic volume index, end-systolic volume index, mass, and mass index were measured as previously described (14,15). Maximal LV end-diastolic wall thickness was measured as previously described (16). Comparing the 2 CMR examinations, a significant change in LV mass was defined for a measured difference ≥5 g. The extent of LGE was measured using a previously validated method (17). Briefly, endocardial and epicardial contours in each image were manually traced to identify LV myocardium in each image. To obtain the standard deviation of the signal noise of the images a region of interest was placed in the background of the image, near the patient's thoracic wall. The mean signal intensity and standard deviation were measured in this region of interest.
Myocardial voxels with signal intensity higher than the average signal intensity of the region of interest plus 6 standard deviations were considered enhanced (18,19). The percentage of enhanced voxels in the entire LV myocardium was measured. Extent of LGE was expressed in grams and percentage of LV mass. A significant increase of LGE was defined when the extent of LGE increased of ≥1 g at CMR-2. This threshold was chosen because in considering the dimensions of the voxel, 1 g of LGE was equivalent to >50 voxels. The rate of progression of LGE (LGE-rate) was defined as the ratio between the increment of LGE extent in grams and the time (months) between the 2 examinations. Quantification of LGE was performed in a random fashion by 2 investigators who were blinded to: 1) the clinical information of the patients; and 2) the value of LGE extent measured at CMR-1.
Categorical variables were compared by Pearson chi-square test or Fisher exact test as appropriate; the McNemar and the Cochran tests were employed for analyzing changes in patients' proportions among classes (e.g., NYHA functional class) between CMR-1 and CMR-2. Statistical tests used to compare groups included paired/unpaired Student t test for difference in mean values and Mann-Whitney U test or Wilcoxon test for skewed variables. The Wilcoxon signed rank test was employed for paired measurements while the Wilcoxon rank sum test (Mann-Whitney U test) was used for 2 independent groups.
The Pearson correlation coefficient was employed for quantifying the relationship between Gaussian distributed variables; for skewed variables, the logarithmic transformation was applied (LGE extent in grams, LGE extent in percentage of LV mass, and increment of LGE extent). Simple and multiple linear regression analyses were then conducted for quantifying the effect of parameters at CMR-1 in relation to LGE changes at CMR-2. The Kolmogorov-Smirnov test was employed to assess normality of data distribution and for the residuals of regression models. A p value <0.05 was considered statistically significant.
Data are presented as mean ± SD, median and interquartile range (IQR: 25th to 75th percentile), and as proportions with percentage, as indicated. Statistical analyses were performed using MATLAB (The MathWorks, Natick, Massachusetts).
The final population of patients with HCM who completed the serial CMR examinations included 55 patients. The average interval between CMR-1 and CMR-2 was 719 ± 410 days.
Clinical characteristics of the study population are summarized in Table 1.
Thirty-seven patients were males. Age ranged from 14 to 83 years (mean 42.4 ± 17.7 years).
Of the 55 patients, 35 (63.6%) were in NYHA functional class I, 19 (34.5%) in class II, and 1 (1.8%) in class III at the time of CMR-1 evaluation. HCM was obstructive in 5 patients (9%). Ten patients had ≥2 arrhythmic risk factors. Three patients had a stress imaging test positive for inducible ischemia despite a coronary artery angiography without significant stenosis in the epicardial coronary arteries.
Measurements obtained at CMR-1 are summarized in Table 1. At CMR-1, LV mass index was upper the range of normality in 43 (78.2%) of the study patients. The mean maximal end-diastolic wall thickness was 20.8 ± 5.4 mm (range 15 to 38 mm) and was ≥30 mm in 3 patients (5.5%). Myocardial hypertrophy mainly involved the interventricular septum and/or the anterior free wall in 32 (58.2%) patients of the study population. Hypertrophy was confined in the ventricular apex in 10 patients (18.2%) and to the inferior and/or inferolateral wall in 3 patients (5.5%). Finally, hypertrophy was diffuse in 10 patients (18.2%). Mean LV end-diastolic volume index was 75.5 ± 17.5 ml/m2. Mean LV ejection fraction was 68.9 ± 9.6%, and was ≤50% in only 1 patient (ejection fraction = 37%).
LGE was detected in 45 (81.8%) of the study patients. The median of LGE extent was 8 g (IQR: 3 to 18 g) with a non-Gaussian distribution (p < 0.001). On average, LGE was identified in 2.3 ± 1.9 (range 0 to 6 segments) of the 16 conventional myocardial segments into which the myocardium was subdivided.
Measurements obtained at the CMR-2 are summarized in Table 1. LV mass augmented in 30 patients (54.5%), remained substantially unchanged in 21 (38.2%), and decreased in 4 patients (7.3%) (Fig. 1). The mean extent of LGE increased from 13.3 ± 15.2 g at CMR-1 to 24.6 ± 27.5 g at CMR-2 (p < 0.001). LV ejection fraction (average 68.5 ± 9.4%) did not change significantly from CMR-1 (p = 0.923).
LGE extent increased (≥1 g) in 44 patients (80%). The median LGE increment was 6 g (IQR: 2 to 14 g) (Fig. 2). LGE extent was unchanged (difference of LGE extent <1 g) in 6 patients and decreased in 2. Among the 11 patients without detectable LGE at CMR-1, 8 patients showed ex novo LGE at CMR-2 (Fig. 3). Thus at CMR-2, LGE was globally found in 53 (96.4%) patients. In 15 patients (27.3%), the number of myocardial segments with LGE significantly increased from CMR-1 to CMR-2, while in 25 patients (45.5%), the increase of LGE was confined in the same segments of CMR-1. LGE-rate was 0.54 ± 0.98 g/month (range 0 to 6 g/month).
During the time interval between the CMR examinations, NYHA functional class improved from II to I in 3 patients who started medical therapy after CMR-1, and worsened in 13 patients (10 from NYHA functional class I to II, 3 from NYHA functional class II to III). Therefore, at CMR-2, 23 patients were in NYHA functional class II and 4 in NYHA functional class III.
LGE increment and clinical correlates
The increment of LGE extent (logarithmic values) between the 2 CMR scans was inversely related to the age at the enrollment (r = −0.309, p = 0.024). The LGE increment was not correlated to the time interval between them (r = 0.044, p = 0.758), and no difference was observed between males and females (p = 0.592).
Patients with apical HCM had a higher increment of LGE (median 11, IQR: 3 to 22 g vs. median 5, IQR: 2 to 11 g; p = 0.004) and LGE-rate (1.40 ± 1.78 g/month vs. 0.33 ± 0.53 g/month; p = 0.001) than those with other patterns of hypertrophy, despite a nondissimilar increase in LV mass (32.0 ± 33.9 g vs. 11.8 ± 31.0 g; p = 0.142); in addition, there was no significant difference in the relationship between LGE at CMR-1 and LGE rate in the 2 groups (interaction term p = 0.818). Patients with apical HCM had also a higher increment of LGE indexed by the LGE at CMR-1 (2.8 ± 3.0 vs. 0.9 ± 0.9; p = 0.002) than other patterns. Patients with worsening of NYHA functional class had higher increase of LGE extent (median 13, IQR: 3 to 18 g vs. median 2, IQR: 1 to 8 g; p = 0.031) and higher LGE-rate (1.15 ± 1.57 g/month vs. 0.40 ± 0.51 g/month; p = 0.049) than those with preserved or improved clinical functional status.
A significant direct relationship between the increase of LV mass index and the increment of LGE was observed (r = 0.504, p < 0.001); furthermore, the increment of LGE was related to the extent of LGE at CMR-1 (r = 0.498, p < 0.001). Differently, LGE increase was not related to the LV mass index at CMR-1 (p = 0.352), ejection fraction (p = 0.068), and the maximal end-diastolic wall thickness (p = 0.077).
The apical pattern of hypertrophy and the extent of LGE at CMR-1 were significant independent predictors of the increment of LGE between the examinations, in a multivariate model containing age as a covariate: the regression model overall explained almost 35% of the variability of LGE increase (Table 2). On average, having an apical pattern of hypertrophy would result in approximately 19-g increase of LGE and a 10-g LGE extent at CMR-1 would correspond to an average increase of approximately 4 g at CMR-2.
HCM is an evolutive disease, so it is intuitive to consider fibrosis a progressive phenomenon, however, the rate of progression of LGE was not previously evaluated. In this study we assessed for the first time the rate of progression of LGE by repeated CMR examinations in HCM patients. The main results were: 1) after an average of 2 years the prevalence of LGE increase from 81.8% to 96.4% of subjects and the extent of LGE increased in the majority of them; 2) subjects with apical HCM had greater increment of LGE and higher LGE-rate than those presenting other patterns of hypertrophy; and 3) the increment of LGE was higher in patients with worsening NYHA functional class.
The presence of LGE in patients with HCM may be considered relevant in terms of prognostic stratification, as recent reports demonstrated that after a clinical follow-up of 3 years, patients with LGE had worse prognoses than those without LGE (9,10). However, the specificity of LGE as prognostic marker in HCM should be rediscussed because it is detected in most of the patients at the first evaluation and in almost all of them after few years, as shown by these results. Then, the clinical and prognostic role of other features of LGE as its global extent, the pattern of distribution and the rate of progression should be evaluated by further studies in HCM patients.
In this study we evaluated the increment of LGE between 2 CMR examinations performed in a time range of 719 ± 410 days and calculated a new index, the LGE-rate, defined as the ratio between the increment of LGE in grams and the time in months between the 2 CMR examinations, which represents the rate of progression of LGE over time. LGE-rate was used to compare the progression of LGE extent in patients with a different time gap between the CMR examinations. However, results showed that the increment of LGE was not related to the time interval between the 2 CMR examinations, and the rate of progression of LGE was highly heterogeneous with a spectrum of values of LGE-rate ranging from 0.06 to 2.53 g/month.
Although at CMR-1 the extent of LGE was not different in patients with different patterns of hypertrophy (Fig. 4), patients with apical HCM had higher increment of LGE and higher LGE-rate than other patterns of HCM at CMR-2 (Fig. 5). Moreover, at multiple regression analysis the apical pattern together with the extent of LGE at the first examination were independent predictors of the increment of LGE extent. In a case report, Gebker et al. (20) described a large increment of LGE extent in patients with apical HCM between 2 CMR examinations after a time interval of 2 years. Several mechanisms may account for the higher progression of fibrosis in the apical pattern of LV hypertrophy. The physiological rarefaction of capillary density at the ventricular apex may contribute to the mismatch between oxygen demand and supply in hypertrophic apical segments. In fact, Moon et al. (21) demonstrated perfusion defect at the hypertrophied segment, representing abnormal myocardial capillary density, in apical HCM patients (22). In patients with apical HCM the development of apical aneurysm may be the extreme consequence of the increased fibrosis, shrinkage, and severe wall thinning (21,23). Yet, myocardial infarction appears to be a common complication of apical HCM during long-term follow-up, even without significant coronary artery disease (24,25). Finally, a genetic predisposition may condition the rate of progression of fibrosis in apical as well in other patterns of HCM.
On the clinical point of view, a significant correlation between LGE and impaired functional class (NYHA functional class ≥II) was already demonstrated in previous reports (26,27). In addition, the current study showed a relation between the rate of progression of fibrosis and the clinical status: patients with worsened NYHA functional class had a higher increment of LGE and LGE-rate than those with unchanged functional status. This result may be explained by the previous observation that impaired diastolic function was related to the extent of LGE in HCM (26) and, consequently, a positive relation between the LGE-rate and the worsening of diastolic function may be hypothesized to justify the association between worsened NYHA and LGE-rate.
The main limitation of this study was the small size of the population. This was expected, considering the percentage of HCM patients undergoing to ICD implantation after the first CMR and those with atrial fibrillation not permitting the acquisition of CMR images because of a nonoptimal electrocardiogram triggering.
Another limitation was that the coronary artery angiography was performed only in patients with a positive exercise stress test. However, 36 patients with a negative stress test were <45 years of age and 17 of them were <30 years of age. The Framingham Risk score demonstrated a <10% risk for coronary artery disease in all the remaining 16 patients (28). Furthermore, the pattern of distribution of LGE in all the patients in the 2 CMR examinations was not ischemic-like; it was intramural, patchy, and nonrespecting of the coronary vessel territory (29). Yet, in this study, genetic analysis for the screening of sarcomeric mutations was not performed, and further investigations are needed to assess whether the rate of progression of fibrosis may be conditioned by a genetic predisposition.
The progression of fibrosis in HCM is very fast, though very heterogeneous, and is faster in apical hypertrophy than in other patterns and it is related to the worsening of the clinical status. Therefore CMR can be applied as a useful and safe tool for longitudinal follow-up evaluation of HCM. However, the clinical and prognostic impact of the LGE-rate, evaluated by repeated CMR examinations, needs to be assessed by further studies, assessing also whether multiple CMR examinations over time could be more useful than a “single shot” approach in the clinical setting.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Todiere and Aquaro contributed equally to this work.
- Abbreviations and Acronyms
- cardiac magnetic resonance
- hypertrophic cardiomyopathy
- implantable cardioverter-defibrillator
- interquartile range
- late gadolinium enhancement
- left ventricular
- New York Heart Association
- Received February 6, 2012.
- Revision received March 14, 2012.
- Accepted March 29, 2012.
- American College of Cardiology Foundation
- Elliott P.M.,
- Poloniecki J.,
- Dickie S.,
- et al.
- Biagini E.,
- Coccolo F.,
- Ferlito M.,
- et al.
- Moon J.C.,
- Reed E.,
- Sheppard M.N.,
- et al.
- O'Hanlon R.,
- Grasso A.,
- Roughton M.,
- et al.
- Bruder O.,
- Wagner A.,
- Jensen C.J.,
- et al.
- Maron B.J.,
- McKenna W.J.,
- Danielson G.K.,
- et al.
- Goldberger J.J.,
- Cain M.E.,
- Hohnloser S.H.,
- et al.
- Rickers C.,
- Wilke N.M.,
- Jerosch-Herold M.,
- et al.
- Adabag A.S.,
- Maron B.J.,
- Appelbaum E.,
- et al.
- Gebker R.,
- Neuss M.,
- Paetsch I.,
- Nagel E.
- Matsubara K.,
- Nakamura T.,
- Kuribayashi T.,
- Azuma A.,
- Nakagawa M.
- Maron M.S.,
- Finley J.J.,
- Bos J.M.,
- et al.
- Eriksson M.J.,
- Sonnenberg B.,
- Woo A.,
- et al.
- Maron M.S.,
- Appelbaum E.,
- Harrigan C.J.,
- et al.
- Wilson P.W.,
- D'Agostino R.B.,
- Levy D.,
- Belanger A.M.,
- Silbershatz H.,
- Kannel W.B.
- Soriano C.J.,
- Ridocci F.,
- Estornell J.,
- Jimenez J.,
- Martinez V.,
- De Velasco J.A.