Hypertrophic cardiomyopathy (HCM) is a primary autosomal dominant disorder of the myocardium caused by mutations in sarcomeric contractile proteins.1 In addition to left ventricular (LV) diastolic dysfunction, patients with HCM have subclinical systolic LV dysfunction2 that may ultimately result in overt LV systolic dysfunction in approximately 1% of patients per year.3 Non-invasive cardiac imaging techniques have a pivotal role in detecting the disease and selecting or guiding appropriate therapy.4 However, the heterogeneous character of both the phenotype and prognosis of HCM patients5 warrants an ongoing search for new non-invasive imaging techniques that offer parameters that might provide further insight into pathophysiology or be of prognostic value. In previous small tagged magnetic resonance imaging (MRI) studies, LV rotation and twist were studied in HCM patients, with discrepant results.6 7 Speckle tracking echocardiography (STE) is a new, emerging echocardiographic image modality that is able to quantify LV twist.8 9 The current study sought (1) to assess LV rotation parameters in patients with HCM compared to normal control subjects, and (2) to examine whether the pattern of LV hypertrophy affects LV rotation parameters in patients with HCM, using STE.

METHODS

Study participants

The study population consisted of 70 consecutive non-selected patients with HCM (mean age 42 (SD 16) years, 52 men) with a typical sigmoidal or reverse septal curvature.1 These patients were compared to healthy age-matched and gender-matched control subjects in sinus rhythm, without hypertension or diabetes, and with normal left atrial dimensions, LV dimensions and LV function. HCM was characterised morphologically and defined by a hypertrophied, non-dilated LV in the absence of another systemic or cardiac disease that is capable of producing the magnitude of wall thickening seen.10 Informed consent was obtained from all subjects and the institutional review board approved the study.

Echocardiography

Two-dimensional greyscale harmonic images were obtained in the left lateral decubitus position using a commercially available ultrasound system (iE33, Philips, Best, The Netherlands), equipped with a broadband (1–5 MHz) S5-1 transducer (frequency transmitted 1.7 MHz, received 3.4 MHz). All echocardiographic measurements were averaged from three heartbeats. From the M-mode recordings the following data were acquired: left atrial size, LV end-diastolic septal and posterior wall thickness and LV end-diastolic and end-systolic dimension. LV ejection fraction was calculated from LV volumes by the modified biplane Simpson rule in accordance with the guidelines.11 LV mass was assessed with the two-dimensional area-length method, as previously described.12 LV outflow tract gradient was measured with continuous-wave Doppler in the apical five-chamber view. LV outflow tract obstruction was defined as a gradient ⩾30 mm Hg.13 From the mitral-inflow pattern, peak early (E) and late (A) filling velocities, E/A ratio, and E-velocity deceleration time were measured. Tissue Doppler was applied end-expiratory in the pulsed-wave Doppler mode at the level of the inferoseptal side of the mitral annulus from an apical four-chamber view. To acquire the highest wall tissue velocities, the angle between the Doppler beam and the longitudinal motion of the investigated structure was adjusted to a minimal level. The spectral pulsed-wave Doppler velocity range was adjusted to obtain an appropriate scale.

To optimise STE, images were obtained at a frame rate of 60–80 frames/s. Parasternal short-axis images at the LV basal level (showing the tips of the mitral valve leaflets) with the cross-section as circular as possible were obtained from the standard parasternal position, defined as the long-axis position in which the LV and aorta were most in line with the mitral valve tips in the middle of the sector. To obtain a short-axis image at the LV apical level (just proximal to the level with end-systolic LV luminal obliteration) the transducer was positioned one or two intercostal spaces more caudal as previously described by us.14 From each short-axis image, three consecutive end-expiratory cardiac cycles were acquired and transferred to a QLAB workstation (Philips, Best, The Netherlands) for offline analysis.

Speckle tracking analysis

Analysis of the datasets was performed using QLAB Advanced Quantification Software version 6.0 (Philips, Best, The Netherlands), which was recently validated against magnetic resonance imaging (MRI) for assessment of LV twist.15 To assess LV rotation, six tracking points were placed manually (after gain correction) on the mid-myocardium on an end-diastolic frame in each parasternal short-axis image. Tracking points were separated about 60° from each other and placed at 1 o’clock (30°, anteroseptal insertion into the LV of the right ventricle), 3 o’clock (90°), 5 o’clock (150°), 7 o’clock (210°), 9 o’clock (270°, inferoseptal insertion into the LV of the right ventricle) and 11 o’clock (330°) to fit the total LV circumference.

If a tracking point showed poor speckle tracking by visual assessment, the position of the tracking point was manually changed on an end-diastolic frame in a circumferential direction towards one of the other tracking points, but not more than 1 hour (30°). When speckle tracking was still insufficient, the position of the tracking point could be changed additionally in the direction of the endocardium. Because all tracking points are needed for optimal measurement of global LV rotation, a subject was considered insufficient for analysis of global LV rotation by STE and excluded from further analysis when, despite these changes, one or more tracking points still did not track well. In addition, patients in whom the short-axis image at the LV apical level could not be obtained from an intercostal space more caudal than the standard position, were also considered insufficient for analysis of global LV rotation by STE because measured rotation is not representative of true rotation.14

Data were exported to a spreadsheet program (Excel, Microsoft Corporation, Redmond, WA, USA) to determine LV peak systolic rotation during ejection (Rot_max), the early peak of LV systolic rotation during the isovolumic contraction phase (Rot_early), time to Rot_max (from R wave to Rot_max), time to Rot_early (from R wave to Rot_early), instantaneous LV peak systolic twist (Twist_max, defined as the maximal value of instantaneous apical Rot_max − basal Rot_max), time to Twist_max (from R wave to Twist_max), and LV untwisting at 5%, 10% and 15% of diastole. The degree of untwisting was expressed as a percentage of maximum systolic twist: untwisting  =  (Twist_max − Twist_t)/Twist_max × 100%, where Twist_t is twist at time t. To adjust for intra-subject and inter-subject differences in heart rate, the time sequence was normalised to the percentage of systolic duration. End-systole was defined as the point of aortic valve closure. In each study it was verified that heart rate for the cardiac cycle in which the timing of aortic valve closure was assessed, was the same as the cardiac cycle used for analysis of LV twist.

Statistical analysis

Measurements are presented as mean (SD). Variables were compared using the Student t test, ANOVA, or χ² test when appropriate. Relations between different parameters were assessed by correlation analysis. A p value of <0.05 was considered statistically significant. Intraobserver and interobserver variability for LV twist in our centre are 6% (6%) and 9% (5%), respectively.

RESULTS

Feasibility of obtaining LV rotation parameters in HCM patients

In 20 HCM patients (29%) image quality of the LV basal level was insufficient for complete STE analysis despite allowed changes in tracking point position. The LV apical level was excluded from analysis in 23 patients (33%) because of either the inability to obtain a short-axis image at the LV apical level from an intercostal space more caudal than the standard position (8%) or because of insufficient image quality (25%). Ten of 26 patients (38%) with a sigmoidal curvature and 17 of 44 patients (39%) with a reverse septal curvature were excluded. The clinical characteristics and LV dimensions and function of the excluded patients were not different from the final study population. In 43 patients (61%) both the LV basal and apical levels were available, facilitating analysis of all LV rotation parameters. These LV rotation parameters in HCM patients (mean age 43 (15) years, 31 men) were compared to 43 age­matched and gender-matched healthy controls.

Characteristics of the study population

In table 1, the clinical and echocardiographic characteristics of the final study population are shown. LV mass, maximal LV wall thickness, left atrial, interventricular septal and LV posterior wall dimensions were increased, whereas LV end-diastolic and end-systolic dimensions were decreased in HCM patients compared to control subjects (all p<0.001). Em septal was lower in HCM patients than in control subjects (6.1 (1.8) cm/s vs 9.7 (2.1) cm/s, p<0.001).

LV rotation parameters in HCM patients vs control subjects

Compared to control subjects, HCM patients had increased basal Rot_max (−5.5° (2.3°) vs −3.4° (1.7°), p<0.001), and comparable apical Rot_max (7.3° (3.1°) vs 7.0° (2.2°), p = NS), resulting in increased Twist_max (12.4° (4.0°) vs 9.9° (2.7°), p<0.01). In a high proportion of HCM patients, counterclockwise basal Rot_early and clockwise apical Rot_early were absent (63% and 40% respectively), whereas in all but two control subjects both parameters were measurable. In the HCM patients with available basal and apical Rot_early, reduced values were seen compared to control subjects (1.2° (0.9°) vs 1.8° (1.0°), p<0.05 and −0.4° (0.3°) vs −0.9° (0.7°), p<0.01, respectively). Untwisting at 5% (10% (10%) vs 17% (14%), p<0.05), 10% (25% (21%) vs 35% (21%), p<0.05), and 15% (39% (20%) vs 50% (20%), p<0.05) of diastole was decreased in HCM patients compared to control subjects (table 2).

Relation between the pattern of LV hypertrophy and LV twist in HCM patients

According to septal morphology, HCM patients could be divided into 16 patients (37%) with sigmoidal septal curvature, and 27 (63%) with reverse septal curvature (table 2). No difference in clinical characteristics could be identified between these two groups (70% vs 74% male, mean age 40 (14) vs 44 (12) years, both p = NS). However, there was a striking difference in apical Rot_max (9.4° (2.8°) vs 6.0° (2.6°), p<0.01) and Twist_max (15.3° (3.2°) vs 10.6° (3.3°), p<0.01) between patients with sigmoidal and reverse septal curvature (fig 1), whereas the other LV rotation parameters were comparable. The extent of LV hypertrophy, reflected by either LV mass or maximal LV segmental thickness, was significantly correlated with Twist_max (r = −0.40, p<0.01 and r = −0.34, p<0.05, respectively). This significant negative correlation of LV mass and Twist_max remained present only in the subgroup of patients with reverse septal curvature (r = −0.36, p<0.05), whereas this correlation was lost in patients with sigmoidal septal curvature (r = 0.03, p = NS). In HCM patients with LV outflow tract obstruction at rest, apical Rot_max and Twist_max were increased (8.7° (2.9°) vs 6.5° (3.1°), p<0.05, and 14.2° (4.0°) vs 11.5° (3.9°), p<0.05, respectively), whereas basal Rot_max was comparable (−5.5° (2.4°) vs −5.4° (2.2°), p = NS) when compared to HCM patients without LV outflow tract obstruction. LV outflow tract obstruction was more often present in patients with sigmoidal septal curvature compared to reverse septal curvature (56% vs 22%, respectively, p<0.05).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Schematic left ventricular systolic rotation curves (based on averaged values of peak rotation during the isovolumic contraction phase (from the subgroup in which these data were available) and the ejection phase, and the timing of these parameters) in hypertrophic cardiomyopathy (HCM) patients, subdivided according to septal morphology, and control subjects, highlighting the differences of left ventricular apical peak systolic rotation during ejection in the subgroups of HCM patients. LV  =  left ventricle.

DISCUSSION

The major findings of this study are (1) increased basal Rot_max in HCM patients with a sigmoidal or reverse septal curvature, and (2) increased apical Rot_max and Twist_max in HCM patients with a sigmoidal septal curvature, whereas it is normal in HCM patients with a reverse septal curvature.

Phenotype-functional relation in HCM patients

HCM is a relatively common genetic cardiac disorder with a well known phenotypic and genotypic heterogeneity.10 Recently, septal morphology was linked to the underlying genetic substrate and best predicted the presence of a myofilament mutation.16 Our study is the first to relate LV rotation parameters to the phenotype of HCM, identifying a possible genotype-phenotype-functional relation. Basal Rot_max was to a similar extent increased in HCM patients with a sigmoidal and reverse septal curvature. However, apical Rot_max and Twist_max were only increased in HCM patients with a sigmoidal septal curvature.

LV twist originates from the dynamic interaction between oppositely wound subepicardial and subendocardial myocardial fibre helices and has an important role in LV ejection and filling.17 18 The direction of LV twist is governed by the epicardial fibres, mainly because of their longer arm of movement.19 LV twist tends to equalise sarcomere shortening between endocardial and epicardial layers of the LV,20 thereby serving as a compensatory mechanism to prevent substantial transmural inhomogeneities of sarcomere shortening in patients with increased LV wall thickness.

Increased basal Rot_max in HCM patients might be explained by loss of counteraction of the subendocardial fibre helix, caused by endocardial ischaemia due to microvascular dysfunction in HCM patients.21 22 Also, larger radius differences in hypertrophic muscle between the subepicardium and subendocardium may increase the dominant action of the subepicardial fibres and increase basal Rot_max.23

The most important finding in this study is increased apical Rot_max and Twist_max in HCM patients with a sigmoidal septal curvature, whereas it is normal in HCM patients with a reverse septal curvature. Not surprisingly, apical Rot_max was increased in HCM patients with a sigmoidal septal curvature. In tagged magnetic resonance imaging studies it has been shown that apical Rot_max is also increased in patients with aortic stenosis.24 25 This may be caused by subendocardial ischaemia with dysfunction of the subendocardial fibres that try to rotate the apex in a clockwise direction. Another potential mechanism may be LV hypertrophy with an increased arm of force over which the subepicardial fibres work, although in HCM patients with a sigmoidal septal curvature we found no correlation between LV mass or maximal LV segmental thickness and apical Rot_max or Twist_max. Of note, patients with a sigmoidal septal curvature more often had LV outflow tract obstruction. Extravascular compressive forces caused by these gradients may lead to microvascular dysfunction and subendocardial ischaemia in HCM patients.22

The normal apical Rot_max in HCM patients with a reverse septal curvature is more difficult to explain since the above-mentioned factors that increase apical Rot_max are also present in these patients, although these patients less often had LV outflow tract obstruction. Importantly, LV rotation is also dependent on helical morphology and intrinsic myocardial contractility. The myofibre helix angle changes continuously from the subendocardium to the subepicardium, typically ranging from +60° at the subendocardium to −60° at the subepicardium.26 Taber et al19 showed that Twist_max approximately doubles with a change in the subendocardial and subepicardial helix angle from +90° to +60° and −90° to −60°, respectively. In patients with a reverse septal curvature the helical fibre configuration will not be optimal because of the distorted apical morphology. Possibly, as LV hypertrophy becomes too extensive in the apical segments, myocardial fibre disarray becomes so encompassing that it predominates factors likely to increase apical Rot_max. Moreover, effectiveness of LV wall contraction depends on its curvature: the more convex towards the LV cavity the wall is, the less it contracts.27 This might all contribute to the relatively decreased apical Rot_max in HCM patients with a reverse septal curvature compared to a sigmoidal septal curvature.

The natural history of HCM is typically variable.28 Considering the possible genotype-phenotype-functional relation, classifying HCM patients according to LV rotation characteristics might provide subgroups of HCM patients with a less heterogeneous prognosis. Clinical studies are needed to test this hypothesis.

The assessment of LV diastolic function is currently based on load-dependent pulsed-Doppler indices of LV filling and less load-dependent tissue Doppler velocities, which only describe events occurring after mitral valve opening.29 In a recent study by Takeuchi et al23 in hypertensive patients, it was shown that LV untwisting assessed by STE may be a novel parameter for evaluating LV relaxation. Diastolic dysfunction is a major pathophysiological abnormality in HCM.30 We found delayed untwisting to be a rather uniform characteristic of patients with HCM regardless of the extent and site of LV hypertrophy, which is in agreement with the results of a study by Spirito and Maron investigating the relation between the extent of LV hypertrophy and Doppler echocardiographic indexes of LV diastolic filling in HCM patients.31

STE offers novel non-invasive indices to assess LV systolic and diastolic function in patients with HCM. Our study showed the important influence of the pattern of hypertrophy on LV twist in HCM, which provides further insight into the pathophysiology of this disease.

Comparison with previous studies

In previous tagged MRI studies in HCM patients, only a very limited number of patients were included and the results were discrepant.6 7 Young et al6 found increased basal and apical Rot_max resulting in increased Twist_max, whereas Maier et al7 found reduced apical Rot_max and Twist_max. Notomi et al described increased Twist_max in seven HCM patients using tissue Doppler imaging.32 Carasso et al found higher circumferential and lower longitudinal strain, but normal Twist_max using velocity vector imaging in 72 HCM patients.33 Unfortunately, data on the site or extent of LV hypertrophy were not reported in any of these studies. Our finding of a significant influence of septal morphology on apical Rot_max and Twist_max in HCM patients can potentially explain previous discrepancies in reported Twist_max in HCM patients.