OUP user menu

Increased amount of interstitial fibrosis predicts ventricular arrhythmias, and is associated with reduced myocardial septal function in patients with obstructive hypertrophic cardiomyopathy

Vibeke M. Almaas, Kristina H. Haugaa, Erik H. Strøm, Helge Scott, Christen P. Dahl, Trond P. Leren, Odd R. Geiran, Knut Endresen, Thor Edvardsen, Svend Aakhus, Jan Peder Amlie
DOI: http://dx.doi.org/10.1093/europace/eut028 1319-1327 First published online: 20 February 2013

Abstract

Aims Reduced echocardiographic strain is associated with ventricular arrhythmias in hypertrophic cardiomyopathy (HCM) patients. The aim of this cross-sectional study was to investigate which type of histological fibrosis contributes to ventricular arrhythmias and reduced septal longitudinal strain, in obstructive HCM-patients with or without additional coronary artery disease (CAD) and/or hypertension (HT).

Methods and results Sixty-three HCM-patients (mean age 57 ± 13 years) were included. Strain by speckle tracking echocardiography was performed prior to either percutaneous transluminal septal ablation (n = 37) or septal myectomy (n = 26). In 24 patients myectomy specimens were available (histology population) and allowed determination of %area of interstitial and replacement fibrosis. Twenty-nine (46%) patients had concomitant CAD and/or HT, and 15 (24%) experienced ventricular arrhythmias defined as documented ventricular tachycardia or arrhythmogenic suspected syncope. The patients with ventricular arrhythmias had lower septal longitudinal strain compared with those without arrhythmias (−9.0 ± 4.0 vs. −13.6 ± 5.6%, P = 0.006). In the histology population reduced septal longitudinal strain correlated to interstitial (R2 = 0.36 P = 0.003), but not to replacement fibrosis (R2 = 0.03 P = 0.43). By logistic regression analyses, interstitial fibrosis predicted ventricular arrhythmias (OR 1.16, 95% CI 1.02–1.32, P = 0.03), while replacement fibrosis did not (OR 1.22, 95% CI 0.93–1.59, P = 0.15).

Conclusion Total amount of fibrosis was a marker of ventricular arrhythmias in obstructive HCM-patients. Interstitial fibrosis seemed to be more important compared with replacement fibrosis in arrhythmogenesis, and was related to reduced septal myocardial function. These findings suggest that interstitial fibrosis may play an important role as the arrhythmogenic substrate, and that strain echocardiography can help detection of patients at risk.

  • Echocardiography
  • Hypertrophic cardiomyopathy
  • Myocardial fibrosis
  • Ventricular arrhythmias

What's new

Different types of histological fibrosis in patients with hypertrophic cardiomyopathy (HCM) have been described. Previous studies comparing histopathology and phenotype of HCM have quantified total amount of fibrosis without discriminating between subtypes of microscopic fibrosis. We hypothesized that subtypes of microscopic fibrosis may have an impact on arrhythmogenesis in patients with HCM.

  • We are the first to investigate the impact of different types of histological fibrosis on ventricular arrhythmias and on myocardial function in obstructive HCM-patients with or without additional coronary artery disease and/or hypertension.

  • Increased amount of both interstitial and replacement type of fibrosis were markers of ventricular arrhythmias.

  • Only interstitial fibrosis predicted ventricular arrhythmias, had an excellent sensitivity and specificity for detecting ventricular arrhythmias, and was strongly correlated to reduced myocardial function.

Interstitial fibrosis may be more important than replacement fibrosis in arrhythmogenesis, and strain echocardiography may help detection of patients at risk.

Introduction

Patients with hypertrophic cardiomyopathy (HCM) have increased risk of ventricular arrhythmias and sudden cardiac death (SCD).13 Autopsy studies from HCM-patients who have suffered SCD have shown that fibrosis comprise an average of 15% of the hypertrophied ventricular septum, eight-fold greater than in patients with heart disease of other aetiologies. These findings suggest that myocardial fibrosis play an important role in mortality and arrhythmogenesis in HCM-patients.4,5 Histologically, different types of fibrosis in HCM-patients have been described including subendocardial, perivascular, replacement, and interstitial type of fibrosis.615 Recent studies have shown that a majority of patients with HCM have myocardial fibrosis detected on late enhancement cardiac magnetic resonance (LE-CMR) imaging,10,16 and that LE is associated with ventricular arrhythmias.1720 However, CMR is expensive and not widely available. Myocardial fibrosis will lead to regionally reduced function. Strain echocardiography is an accurate measure of regional function and seems to identify fibrosis detected on LE-CMR.21 Furthermore, it has been indicated that reduced myocardial function by strain may be associated with non-sustained ventricular tachycardia (NSVT), indicating a relationship between myocardial fibrosis, reduced function, and arrhythmogenesis in patients with HCM.22,23

We investigated the impact of different types of histological fibrosis on ventricular arrhythmias and on myocardial function obstructive HCM-patients undergoing septal myectomy. Furthermore, we wanted to explore the relationship between reduced myocardial function by strain echocardiography and ventricular arrhythmias in a HCM-population including patients with concomitant coronary artery disease (CAD) and/or hypertension (HT). We hypothesized that strain echocardiography may contribute to risk stratification in HCM-patients.

Methods

Patient population

Between November 2001 and September 2009, 104 HCM-patients were referred to our institution considering septal reduction therapy with either percutaneous transluminal septal myocardial ablation (PTSMA) or surgical septal myectomy, and subsequently included in this cross-sectional study at Oslo University Hospital, Rikshospitalet, Norway. At Scandinavian institutions PTSMA is the preferred treatment option for patients without concomitant CAD or valvular heart disease.24 Eighty-four out of 104 patients met the criteria for septal reduction therapy with resting left ventricular outflow tract gradient (LVOTG) ≥30 mmHg or exercised-induced LVOTG ≥50 mmHg, and New York Heart Association functional class >2 despite maximal medical treatment.2 The diagnostic criteria for HCM were end diastolic wall thickness >15 mm and a non-dilated left ventricle (LV) with ejection fraction (EF) ≥50%.1 Exclusion criteria were hypertrophy of other causes (n = 15), history of myocardial infarction (n = 4), and prior intervention with PTSMA (n = 1) or septal myectomy (n = 0). One patient was excluded because of unavailable off-line echocardiographic analyses. Patients with CAD (coronary artery stenosis ≥50% evaluated at coronary angiography) without myocardial infarction, mild, or moderate valvular heart disease unrelated to HCM, and patients on antihypertensive therapy were not excluded.

The study population comprised 63 patients with obstructive HCM, of which 29 patients had concomitant CAD and/or HT. Clinical examination, registration of year of symptom onset, registration of ventricular arrhythmias, and risk stratification for SCD were assessed prior to intervention with either PTSMA (n = 37) or myectomy (n = 26). Of these, myectomy specimens and histology analyses were available in 24 patients (histology-population). The examinations necessary for risk stratification [clinical examination, echocardiography, 24–48 h electrocardiogram (ECG) (Holter) and bicycle exercise ECG] were usually done during 2 days at the first appointment at the outpatient clinic.

The PTSMA and septal myectomy procedures were performed as described previously.2527 In the histology population, isolated myectomy was performed in 11 patients, and combined surgery in 13. Combined surgery included additional coronary artery bypass graft surgery (n = 10) and/or additional valvular heart surgery (n = 5). One patient treated with PTSMA had concomitant CAD not eligible for revascularization.

The study complies with the Declaration of Helsinki, and the research protocol was approved by the Regional Committee for Medical Research Ethics, and informed consent of the subjects has been obtained.

Definition of ventricular arrhythmias and risk stratification

Ventricular arrhythmias were registered prior to intervention and defined by ≥1 of the following criteria: (i) prior cardiac arrest with documented ventricular arrhythmias, (ii) documented NSVT at Holter monitoring (three or more complexes with heart rate ≥120 and duration <30 s), (iii) unexplained syncope (suspected arrhythmogenic syncope and/or syncope during exercise). Risk stratification was performed according to current guidelines for primary and secondary prevention of SCD.1,2 Premature SCD in the family was defined as sudden unexpected death without any known heart disease and/or HCM-related sudden death in first- and/or second-degree relatives, and age <40 years.

Echocardiography

Transthoracic echocardiography was performed at a median (25%, 75% percentiles) of 98 (27, 181) days prior to intervention using Vivid 7 (GE Healthcare, Horten, Norway). Two-dimensional and Doppler echocardiographic measurements were performed according to standards.28 Maximal wall thickness (MWT) was assessed from parasternal short-axis plane primarily at the level of papillary muscle.29 We used Spirito–Maron index and Wigle’s score to measure the extent of hypertrophy.29,30 Peak resting and exercise-induced LV outflow tract gradient (LVOTG) were measured by continuous Doppler at rest, and during exercise (bicycle) when resting LVOTG was <30 mmHg. Strain by speckle tracking echocardiography was measured as previously described (Figure 1).31 Longitudinal systolic strain was obtained from three apical views (apical long axis, apical two-chamber, and apical four-chamber) at (mean ± SD) 61 ± 21 frames/s. Global longitudinal strain was obtained by averaging peak longitudinal strain in a 16-segment model (6 basal, 6 mid, and 4 apical segments). Mean of peak longitudinal strain of basal and mid septal segments from the apical four-chamber (AP4CHA) view was defined as septal longitudinal strain (Figure 1).

Figure 1

Longitudinal strain by speckle tracking echocardiography from a patient with obstructive HCM. (A) Apical short axis; (B) apical two-chamber; (C) apical four-chamber. Global longitudinal strain was obtained by averaging the maximum systolic shortening in a 16-segment model (6 basal, 6 mid, and 4 apical segments). Septal longitudinal strain was defined as mean of two septal segments (basal and mid segments from apical four-chamber view).

The standard echocardiographic measurements were analysed by a single observer (V.M.A.) and the strain echocardiographic measurements by two observers (K.H.H. and V.M.A.) blinded for clinical data and outcome.

Histopathology of myectomy specimens

The tissue specimens were fixed in 10% formalin, paraffin-embedded, and sectioned for light microscopy. The specimen was stained with haematoxylin–eosin–saffron and acid fucsin orange-G (AFOG), highlighting connective tissue. A small proportion of each specimen was fixed in glutaraldehyde, embedded in epon, and processed for electron microscopy to exclude storage disease. The type of fibrosis was classified as interstitial, replacement, perivascular, or subendocardial (Figure 2). Thin strands of fibrous tissue encircling the myocytes were classified as interstitial fibrosis, and confluent fibrosis substituting the myocytes, was classified as replacement fibrosis.13 Perivascular fibrosis and myocyte disarray were assessed as being present or not. Subendocardial fibrosis was excluded from the analyses. Two observers (E.H.S. and H.S.) analysed the AFOG-stained sections simultaneously, and made a visual estimate of the percentage area (%area) affected by interstitial and replacement fibrosis of the total specimen.32 The observers were blinded for clinical data and outcome. A repeated blinded quantification of replacement fibrosis was performed after 12 weeks.

Figure 2

Different types of histological fibrosis in patients with obstructive hypertrophic cardiomypathy. Sections from the surgical specimens removed from the basal part of the ventricular septum in patients undergoing septal myectomy were stained with acid fuchsin orange-G. Blue colour indicates fibrosis/connective tissue. (A) Interstitial fibrosis (×200); (B) replacement fibrosis (×200); (C) perivascular fibrosis (×100); (D) subendocardial fibrosis (×40).

Genetic analyses

A total of 86 polymerase chain reaction products together spanning the translated exons with flanking intron sequences of the sarcomere genes myosin-binding protein C (MYBPC3), β-myosin heavy chain (MYH7), regulatory and essential light chains of myosin (MYL2 and MYL3), and cardiac troponin T (TNNT2) and I (TNNI3) were subjected to DNA sequencing using Version 3.1 of BigDye terminator cycle-sequencing kit and a Genetic Analyzer 3730 from Applied Biosystems.

Statistics

Continuous data are presented as mean ± SD or as median (25%, 75% percentiles). Comparisons of means and medians were performed by the use of unpaired Student's t-test and Mann–Whitney U test, respectively (SPSS version 18, SPSS Inc). Comparisons of proportions were done by the Pearson χ2 or Fischer's exact test. Univariate linear regression analyses were used to identify explanatory variables of reduced septal longitudinal strain, and to correlate total amount of fibrosis and the different types of fibrosis to age and gender. Multivariate linear regression analysis was performed by including significant variables (P < 0.05) from the univariate analyses. In the multivariate analysis of the total population (n = 63) modifying factors as age at intervention, sex, CAD and HT were forced in. In the multivariate analysis of the histology population (n = 24) MWT was forced in. Univariable logistic regression analyses were used to identify predictors of arrhythmic events, and multivariable logistic regression analyses was performed including significant variables (P < 0.05) from the univariable analyses. The ability of global and septal longitudinal strain and amount of histological fibrosis to discriminate patients with ventricular arrhythmias were assessed using the area under the curve (AUC) by receiver operating characteristic (ROC) analyses. Optimal sensitivity and specificity was defined as the value from the ROC curve closest to the upper left corner. Reproducibility was expressed by intraclass correlation coefficient. P values < 0.05 were considered statistically significant.

Results

Clinical characteristics

Fifteen out of 63 patients (24%) had experienced ventricular arrhythmias, while 48 had no recorded events (Table 1). Of these, two of the patients had survived aborted cardiac arrest with documented ventricular fibrillation (of which one patient also had experienced arrhythmogenic suspected syncope), seven had NSVT (of which two also had experienced arrhythmogenic suspected syncope), and six had arrhythmogenic suspected syncope. In all, 29 (46%) patients had additional CAD (n = 16), HT (n = 21), or both (n = 8). Ventricular arrhythmias were not significantly more frequent in patients with additional CAD and/or HT compared with those without (P = 0.09).

View this table:
Table 1

Patients characteristics of 63 obstructive HCM-patients with or without additional CAD and/or HTa, and with or without ventricular arrhythmias

With additional CAD and/or HT (n = 29)Without additional CAD and/or HT (n = 34)Pb valueNo ventricular arrhythmias (n = 48)Ventricular arrhythmias (n = 15)Pc value
Age at symptom onset (years)57 ± 1539 ± 15<0.00149 ± 1342 ± 190.16
Age at intervention (years)63 ± 1151 ± 13<0.00158 ± 1353 ± 130.24
Female (n)15 (52%)19 (56%)0.7424 (50%)10 (67%)0.26
BSA (m²)1.9 ± 0.21.9 ± 0.2 0.541.9 ± 0.21.9 ± 0.20.22
Familial history of HCM (n)6 (21%)12 (35%)0.2011 (23%)7 (47%)0.10
Sarcomere gene mutation (n)5 (17%)19 (56%)0.00215 (32%)9 (60%)0.05
Additional CAD and/or HTa (n)25 (52%)4 (27%)0.09
Beta blockers (n)27 (93%)26 (77%)0.0942 (88%)11 (73%)0.23
Calcium antagonists (n)9 (31%)12 (35%)0.7216 (33%)5 (33%)1.00
ACE inhibitors/A2 antagonists (n)15 (52%)1 (3%)<0.00113 (27%)3 (20%)0.74
Ventricular arrhythmias (n)4 (14%)11 (32%)0.09
Prior cardiac arrest with documented VF (n)1 (3%)1 (3%)1.0002 (13%)0.05
NSVT, n = 59 (n)1/26 (4%)6/33 (18%)0.120/457/14 (50%)<0.001
Unexplained syncope (n)2 (7%)7 (21%)0.1609 (60%)<0.001
Premature SCD in the family (n)7 (24%)6 (18%)0.537 (15%)6 (40%)0.06
Pathologic blood pressure response during exercise, n = 56, (n)5/24 (21%)11/32 (34%)0.2713/44 (29%)3/12 (25%)1.00
LV hypertrophy ≥30 mm (n)03 (9%)0.2403 (20%)0.01
  • Values are mean ± SD and n (%) when appropriate.

  • HCM, hypertrophic cardiomyopathy; CAD, coronary artery disease; HT, hypertension; BSA, body surface area; ACE inhibitors, angiotensin-converting enzyme inhibitors; A2 antagonists, angiotensin 2 antagonists; VF, ventricular fibrillation; NSVT, non-sustained ventricular tachycardia; SCD, sudden cardiac death; LV, left ventricle.

  • aTreatment for HT.

  • bComparisons between HCM-patients with or without additional CAD and/or HT.

  • cComparisons between HCM-patients with or without ventricular arrhythmias.

In the histology population (n = 24), 6 patients (25%) had experienced ventricular arrhythmias (1 aborted cardiac arrest with ventricular fibrillation, 2 NSVT, and 3 syncope), while 18 (75%) had no recorded events. Of the 24, 15 (63%) had additional CAD (n = 13), HT (n = 10), or both (n = 8) with no differences in VT frequency (P = 0.64).

Echocardiographic parameters

Patients with ventricular arrhythmias had higher MWT and reduced global and septal longitudinal strain compared with those without ventricular arrhythmias (P = 0.01, P = 0.02 and P = 0.006, respectively) (Table 2). In the total population there were no differences of the echocardiographic parameters in the patients with or without additional CAD and/or HT (data not shown). Interobserver intraclass correlation for strain measurements, calculated from a 16 LV segments model in 10 patients, was 0.91.

View this table:
Table 2

Echocardiographic parameters in 48 obstructive HCM-patients without and 15 obstructive HCM-patients with ventricular arrhythmias

No ventricular arrhythmias (n = 48)Ventricular arrhythmias (n = 15)P value
MWT (mm)20 ± 424 ± 50.01
Mean Wigle's score (1–10)6.4 ± 2.47.4 ± 2.00.14
Spirito–Maron Index68 ± 1375 ± 160.14
EF (%)64 ± 963 ± 80.64
Peak LVOTGa (mmHg)66 ± 2265 ± 220.96
Global longitudinal strain (%)−14.7 ± 3.4−12.2 ± 3.70.02
 Septal longitudinal strain (%)−13.6 ± 5.6−9.0 ± 4.00.006
  • Values are mean ± SD.

  • HCM, hypertrophic cardiomyopathy; MWT, maximum wall thickness; EF, ejection fraction; LVOTG, left ventricle outflow tract gradient.

  • aResting or exercise-induced LVOTG.

Histological parameters

Extent of total fibrosis, interstitial, and replacement fibrosis were greater in patients with ventricular arrhythmias compared with those without (Table 3). Only two patients had higher %area of replacement than interstitial fibrosis, and none of these experienced ventricular arrhythmias. There were no differences of the histological parameters in the patients with or without additional CAD and/or HT (data not shown).

View this table:
Table 3

Histological parameters of 18 obstructive HCM-patients without and 6 obstructive HCM-patients with ventricular arrhythmias

No ventricular arrhythmias (n = 18)Ventricular arrhythmias (n = 6)P value
Disarray (n)14 (78%)5 (83%)1.00
Small vessel disease (n)11 (61%)2 (33%)0.36
Histological fibrosis present (n)17 (94%)6 (100%)1.00
%Area of total amount of fibrosis (%)10 (6, 15)47 (30, 71)0.001
%Area of interstitial fibrosis (%)7 (2, 12)44 (23, 61)0.001
%Area of replacement fibrosis (%)3 (1, 3)7 (3, 14)0.02
  • Values are n (%) and median (25%, 75% percentiles) when appropriate.

  • HCM, hypertrophic cardiomyopathy.

There was a significant correlation between increasing amount of total fibrosis and lower age at intervention (R2 = 0.17, P = 0.04), with similar relations for interstitial and replacement type of fibrosis (R2 = 0.15, P = 0.06, and R2 = 0.10, P = 0.13, respectively). Gender was not correlated to extent of histological fibrosis.

Re-test reproducibility, expressed as intraclass correlation, was 0.98.

Correlation analyses of myocardial septal function

Ventricular arrhythmias, MWT, and the presence of a sarcomere mutation were significantly correlated to reduced septal longitudinal strain by univariate linear regression analyses in the total population (n = 63) (Table 4). In the multiple regression analyses of septal longitudinal strain in the total population including age at intervention, sex, presence or not of a sarcomere mutation, ventricular arrhythmias, MWT, CAD, and HT in the model (R2 = 0.40 P = 0.001), MWT was independently associated to reduced septal longitudinal strain (P = 0.005). In the histology population (n = 24) %area of total amount of fibrosis and interstitial fibrosis, but not replacement fibrosis, correlated significantly with reduced septal longitudinal strain (Table 4). In the multiple regression analyses of septal longitudinal strain in the histology population, including ventricular arrhythmias, MWT, and interstitial fibrosis in the model (R2 = 0.56 P = 0.008), interstitial fibrosis was independently associated to reduced septal longitudinal strain (P = 0.05).

View this table:
Table 4

Univariate linear regression analyses of septal longitudinal strain in the total population of 63 obstructive HCM-patients, and in the histology population of 24 patients

Septal longitudinal strain
Total population (n = 63)R2P
Age at intervention (years)0.060.06
Sex0.040.15
BSA (m2)0.040.11
Sarcomere gene mutation (yes/no)0.100.01
Ventricular arrhythmias (yes/no)0.120.006
MWT (mm)0.28<0.001
Peak LVOTGa (mmHg)0.030.20
CAD (yes/no)0.040.14
HTb (yes/no)0.050.11
Atrial fibrillation (yes/no)0.010.67
Histology population (n = 24)
 %Area of total amount of fibrosis0.310.006
 %Area of interstitial fibrosis0.360.003
 %Area of replacement fibrosis0.030.43
  • HCM, hypertrophic cardiomyopathy; BSA, body surface area; MWT, maximum wall thickness; LVOTG, left ventricle outflow tract gradient; CAD, coronary artery disease; HT, hypertension.

  • aResting or exercise induced LVOTG.

  • bTreatment for hypertension.

Prediction of ventricular arrhythmias

Univariable logistic regression analyses of the total population showed that reduced septal longitudinal strain, increased MWT, and genotype-positive status were predictors of ventricular arrhythmias (Table 5). A strong correlation was found between septal longitudinal strain and MWT (R2 = 0.28, P < 0.001) and these parameters were therefore analysed separately in the multivariable model. Multivariable analyses including septal longitudinal strain and genotype-positive status showed that septal strain was an independent predictor of ventricular arrhythmias (OR 1.45, 95% CI 1.02–2.05, P = 0.04), while genotype-positive status was not (OR 1.63, 95% CI 0.08–32.02, P = 0.75). The MWT did not predict ventricular arrhythmias when analysed together with the presence or not of a sarcomere mutation (OR 1.27, 95% CI 0.82–1.96, P = 0.28).

View this table:
Table 5

Univariable logistic regression analyses of ventricular arrhythmic events in the total population of 63 obstructive HCM-patients, and in the histology population of 24 patients

Arrhythmic events
Total population (n = 63)OR, 95% CIP
Age at intervention (years)0.97, 0.93–1.020.24
Sex2.18, 0.65–7.320.21
BSA (m2)0.15, 0.01–3.010.21
Sarcomere gene mutation (yes/no)0.30, 0.09–1.010.05
MWT (mm)1.20, 1.03–1.400.02
Peak LVOTGa (mmHg)1.00, 0.97–1.030.96
CAD (yes/no)2.68, 0.53–13.440.23
HTb (yes/no)2.40, 0.60–9.670.22
Global longitudinal strain (%)1.18, 0.97–1.420.09
Septal longitudinal strain (%)1.20, 1.04–1.370.01
Histology population (n = 24)
 %Area of total amount of fibrosis1.19, 1.02–1.390.03
 %Area of interstitial fibrosis1.16, 1.02–1.320.03
 %Area of replacement fibrosis1.22, 0.93–1.590.15
  • HCM; hypertrophic cardiomyopathy, BSA; body surface area, MWT; maximal wall thickness, LVOTG; left ventricular outflow tract gradient, CAD; coronary artery disease, HT; hypertension.

  • aResting or exercise-induced LVOTG.

  • bTreatment for hypertension.

Univariable logistic regression analyses of the histology population showed that reduced septal longitudinal strain (OR 1.47, 95% CI 1.04–2.06, P = 0.03), and increased %area of total amount of fibrosis and interstitial fibrosis were predictors of ventricular arrhythmias, while replacement fibrosis was not (Table 5).

By ROC analyses of the total population, global longitudinal strain had a sensitivity of 71%, a specificity of 67%, and an AUC of 0.70 (95% CI 0.52–0.87) (optimal cut-off value ≥−12.2%), to detect patients with ventricular arrhythmias. Septal longitudinal strain had even better AUC of 0.73 (95% CI 0.60–0.87) (optimal cut-off value ≥−13.3%) with a sensitivity of 86% and a specificity of 50% to detect patients with ventricular arrhythmias.

ROC analyses of the histology population of 24 patients showed that total amount of fibrosis had a sensitivity of 100%, a specificity of 94%, and an AUC of 0.97 (95% CI 0.91–1.0) (optimal cut-off value ≥22.5%), interstitial fibrosis had a sensitivity of 83%, a specificity of 94%, and an AUC of 0.96 (95% CI 0.88–1.0) (optimal cut-off value ≥21.5%), and replacement fibrosis had a sensitivity of 83%, a specificity of 83%, and an AUC of 0.81 (95% CI 0.56–1.0) (optimal cut-off value ≥3.75%), to detect patients with ventricular arrhythmias.

Hypertrophic cardiomyopathy mutation

In 24 out of 63 patients (38%), sarcomere mutations were identified. Of these, 12 had a mutation in the MYBPC3 gene, 7 in the MYH7 gene, 1 in the MYL3 gene, 1 in the TNNI3 gene, 1 in the TNNI2 gene, and 2 patients had mutations both in the MYBPC3 and MYH7 gene. Genotype-positive patients had higher frequency of ventricular arrhythmias and lower frequency of additional CAD and/or HT compared with genotype-negative patients (Table 1). Furthermore, genotype-positive patients were younger at symptom onset (41 ± 16 vs. 52 ± 17 years, P = 0.01) and at intervention (50 ± 15 vs. 61 ± 11 years, P = 0.001), and had lower septal longitudinal strain (−10.2 ± 5.2 vs. −13.9 ± 5.4%, P = 0.01) compared with the genotype-negative patients. In the subgroup of 24 genotype-positive patients, septal longitudinal strain was significantly reduced in the patients with ventricular arrhythmias (n = 9) compared with those without (n = 15) (−7.7 ± 2.3%, −11.6 ± 5.8%, P = 0.04).

Four (17%) out of 24 patients (histology population) were genotype-positive. Two patients had a mutation in the MYH7 region (M982T and V606M), one had a mutation in the MYBPC3 region (c.927-2A>G) and one had a mutation in the MYL3 region (A57D). There were no significant differences of the total amount of fibrosis (35(10, 63)% vs. 14(6, 14)% P = 0.27), interstitial (33(8, 59)% vs. 9(3, 18)% P = 0.18) or replacement type of fibrosis (3(1, 4)% vs. 3(1, 6)% P = 0.68) in genotype-positive and genotype-negative patients.

Discussion

This study demonstrates that increased amount of interstitial fibrosis and reduced septal function are interrelated and associated with ventricular arrhythmias in obstructive HCM-patients. Both reduced septal function and increased amount of interstitial fibrosis were predictors of ventricular arrhythmias and had excellent sensitivity and specificity for detecting ventricular arrhythmias. Interestingly, replacement fibrosis was not a predictor of ventricular arrhythmias. These results suggests that interstitial fibrosis, specifically, may play an important role as the arrhythmogenic substrate in obstructive HCM-patients, and that strain echocardiography can help detection of patients at risk.

Myocardial strain by echocardiography accurately quantifies regional myocardial function.33 Despite normal or supernormal EF, HCM-patients have reduced myocardial function evaluated by strain echocardiography.23,31 Studies have shown that reduced longitudinal strain correlates with increased wall thickness and with hyperenhancement indicating fibrosis on CMR.21 In our study, global longitudinal strain was reduced at the same level as reported previously. Correia et al.22 and Di Salvo et al.23 recently showed that reduced septal function by strain echocardiography identified HCM-patients at high risk of NSVT. In our study, septal longitudinal strain was reduced in HCM-patients with ventricular arrhythmias, and reduced septal longitudinal strain was an independent predictor of ventricular arrhythmias. Furthermore, reduced septal longitudinal strain correlated with total amount of fibrosis, and when separating into interstitial and replacement fibrosis, only increased amount of interstitial fibrosis correlated significantly with reduced septal longitudinal strain. These findings indicate that interstitial fibrosis may have a stronger impact on myocardial contractile function than replacement fibrosis, and that septal longitudinal strain measurement may have a role in ventricular arrhythmic risk stratification.

Sudden cardiac death in patients with HCM is caused by sustained VT and/or ventricular fibrillation, and myocardial fibrosis and myocyte disarray have been suggested to be the arrhythmogenic substrate.34 Schumacher et al.35 found marked alteration of local conduction properties in the hypertrophic septal areas in obstructive HCM-patients with asymmetrical septal hypertrophy. Local conduction disturbances are well-known requirements for development of ventricular tachyarrhythmias.

Histological studies have shown that the amount of myocardial fibrosis correlates with increased wall thickness, and Unverferth et al. found that the amount of fibrosis was highest in the hypertrophied septal region.4,5,36 In the current study, the multiple regression analyses of septal longitudinal strain showed that MWT was independently associated with reduced septal longitudinal strain. In 1979 Anderson et al.8 described different types of histological fibrosis in HCM-patients. Most of the studies comparing histopathology and phenotype of HCM, quantify the total amount of fibrosis without discriminating between the subtypes of microscopic fibrosis.4,5 We used a routine connective tissue staining method to visualize the different types of myocardial fibrosis. We found that the extent of interstitial and replacement fibrosis was higher in HCM-patients with ventricular arrhythmias compared with those without. However, the regression analyses showed that interstitial fibrosis predicted ventricular arrhythmias, and not replacement fibrosis. Our findings, therefore, indicate that interstitial fibrosis is more important than replacement fibrosis in arrhythmogenesis. Increasing attention has been paid to structure and function of the myocardial interstitial space.37 Increased amount and changed structure of interstitial fibrosis has been associated with SCD in children and young adults with previously asymptomatic HCM.12 None of these patients had myocardial replacement fibrosis. The authors indicate an association between increased amount of interstitial fibrosis and LV arrhythmias, although the exact mechanisms are unclear. Tamarappoo et al.38 suggests that increased amount of and changes in the composition of extracellular matrix may create slow electrical propagation, and thereby create potential sites for reentry. Regions with interstitial fibrosis, where the fibrosis encircles the preserved myocytes, may be more vulnerable for electrical and mechanical disturbances than regions with replacement fibrosis. Future studies should therefore aim to differentiate between interstitial and replacement type of fibrosis.

The patients without additional CAD and/or HT were younger at symptom onset and at intervention, and there were a higher proportion of genotype-positive patients compared with those with additional disease. Nevertheless, MWT, peak LVOTG, global and septal longitudinal strain, and histological parameters were not different. However, the proportion of patients with ventricular arrhythmias was not different in the patients with and without additional CAD and/or HT. Among those with CAD and/or HT there were a higher proportion of patients with medical treatment for HT with angiotensin-converting enzyme inhibitors, A2 antagonists, and diuretics. However, there was no significant effect of treatment for HT in the regression analyses.

In the current study, we found a higher proportion of ventricular arrhythmias in patients with sarcomere mutations compared with genotype-negative patients. Genotype-positive patients were younger at symptom onset and at intervention. These findings indicate an earlier onset and higher arrhythmic risk in genotype-positive HCM-patients and are in line with previous findings.39,40 Although robust genotype–phenotype correlations have not yet been established, HCM-patients with sarcomere mutations seem to have a more severe phenotype than patients without an identifiable mutation indicating that the aetiology of HCM play a role for onset and outcome of the disease.41 Our study adds to the current knowledge by showing that myocardial septal strain was reduced in the genotype-positive patients. Reduced myocardial septal strain was a risk factor for ventricular arrhythmias in both genotype-positive and -negative patients. The reasons for reduced myocardial septal strain in the genotype-positive patients compared with the genotype-negative patients are unknown. This study did not include sufficient number of patients to conclude whether type of fibrosis differed between genotype-positive and -negative patients and, therefore could explain a reduced septal strain in genotype-positive patients, and thus contribute to a higher risk of ventricular arrhythmias.

Limitations

The current study included only obstructive HCM-patients referred to a tertiary centre. The results of this study are relevant for this subset of HCM-patients, which constitute ∼25% of a total HCM-population, and the results may not be representative for a general HCM-population. Comorbidity with CAD and/or HT may influence myocardial contraction and development of fibrosis. However, the extent of histological fibrosis was not different in the group with or without CAD and/or HT, and the current HCM-population of 63 patients did not differ from other HCM-populations with respect to age, sex, MWT, EF, and frequencies of familial HCM and a family history of SCD.10,20,23 Furthermore, the incidence of ventricular arrhythmias of 24% was similar to the reported incidence (17–26%) in other HCM-populations.17,19,20

Conclusion

In the current study, total amount of fibrosis in myectomy specimens was a marker of ventricular arrhythmias in obstructive HCM-patients. Interstitial fibrosis seemed to be more important compared with replacement fibrosis in arrhythmogenesis. Reduced septal longitudinal strain was an independent predictor of ventricular arrhythmias and interrelated with increased amount of interstitial fibrosis. Thus, reduced septal contraction by echocardiography may serve as an additional risk stratification tool when evaluating arrhythmic risk in HCM-patients.

Funding

This work was supported by Inger and John Fredriksen's Heart Foundation.

Acknowledgements

We appreciate the help from Maj-Britt Skaale, Department of Cardiology, Oslo University Hospital, Rikshospitalet, for processing rhythm registration data.

Conflict of interest: None declared.

References

View Abstract