Between January 2006 and December 2009, we included consecutive patients >18 years old with severe symptomatic AS (aortic valve area <1 cm2 or mean transaortic gradient ≥40 mmHg) referred for AVR at the Cardiothoracic Surgery Department of Centro Hospitalar São João, Porto, Portugal. We excluded patients with aortic regurgitation >II/IV or other significant valve diseases (>mild), significant coronary artery disease (lesions >50% on coronary angiography), EF≤40%, or previous heart surgery. All patients had to be in sinus rhythm at the time of inclusion. From the initial 141 patients included in a prospective cohort, at the time of AVR, 56 random patients underwent myocardial biopsies for collagen volume fraction evaluation and were considered for this prospective analysis. Clinical and echocardiographic follow-up was provided for all patients. Mean clinical follow-up was 5.0±2.2 years and mean final echocardiographic follow-up was 4.0±1.8 years.

The diagnosis of hypertension was performed based on the clinical records provided by the attending physician. Renal insufficiency was determined when creatinine clearance was <60 ml/kg calculated using the Cockcroft-Gaul formula.

Clinical endpoints were defined as all-cause death and a composite of all-cause death or non-fatal cardiovascular hospitalization (for HF, myocardial infarction, stroke, new-onset atrial fibrillation or advanced AV block requiring hospitalization).

All surgery was performed using standard procedure for AVR. Patients were placed on cardiopulmonary bypass and cardiac arrest was induced and maintained with cold blood cardioplegia. The majority of patients received a bioprosthesis (64.3%) and prosthesis sizes used were <21 mm in 12.5%, 21 mm in 39.3%, 23 mm in 32.1% and 25 mm in 16.1%. At the time of surgery, patients underwent myocardial biopsies from the LV interventricular septum.

Echocardiographic examination was performed by a trained cardiologist and recorded on digital support. All recordings were examined by a cardiologist experienced in echocardiographic imaging at an accredited independent echocardiography laboratory (Hospital Clínico San Carlos in Madrid, Spain), blinded to patient details. Studies were performed using Phillips IE-33 equipment with a S5-1 transducer and M-mode, two-dimensional, pulsed, continuous, color-flow and tissue Doppler capabilities. Correct orientation of imaging planes, cardiac chambers dimensions and function measurements were performed according to the European Association of Echocardiography (EAE)/American Society of Echocardiography (ASE) recommendations.7

Left ventricle mass was estimated according to the joint recommendations of the ASE and EAE using Devereux's formula for ASE measurements in diastole: LV mass=0.8×(1.04×([LV internal dimension+posterior wall thickness+interventricular septal thickness]3−[LV internal dimension]3)+0.6 g.7 Left ventricular hypertrophy was defined by an LV mass index >115 g/m2 in men and >95 g/m2 in women.

Relative wall thickness (RWT) was calculated for the assessment of LV geometry using the formula 2× posterior wall thickness/LV diastolic diameter. Increased RWT was present when this ratio was >0.42.7

LA volume was measured in LV end systole in the frame preceding mitral valve opening. The volume was measured using the biplane area length method and corrected for body surface area. Aortic valve area was estimated using quantitative Doppler continuity equation.

Mitral inflow was assessed in the apical four-chamber view using pulsed wave Doppler with the sample volume placed at the tips of mitral leaflets during diastole. From the mitral inflow profile, the peak flow velocity of early filling (E wave), peak flow velocity of atrial contraction (A wave), the E/A ratio, and early filling deceleration time (DT) were measured. Doppler tissue imaging (DTI) of the mitral annulus was obtained from the apical four-chamber using a sample volume placed in the septal mitral valve annulus. The peak systolic annular velocity (Sm) and early diastolic septal velocity (e’) were determined, and the E/e’ ratio was derived.

As a measure of global LV load, we calculated the valvulo-arterial impedance: Zva=(SAP+MG)/SVI, where SAP is the systolic arterial pressure and MG is the mean transvalvular pressure gradient and SVI is stroke volume index.8

Light microscopic quantification of fibrosis has previously been described and validated.9 Fibrosis analysis of myocardial biopsies was performed using picrosirius-red–stained, 4-μm-thick-sections of tissue (±5 sections of each sample). Images of these sections were acquired with a projection microscope (×50). Subsequent image analysis with Slidebook 4.0 software (3I, Denver, Colo) was performed to determine the extent of reactive interstitial fibrosis, which was expressed as collagen volume fraction (%). Areas of reparative and perivascular fibrosis were excluded. Myocardial fibrosis was calculated as the sum of all connective tissue areas divided by the sum of connective tissue and muscle areas averaged over four to six representative fields of the section in 18 male and 38 female AS patients. In our laboratory, the normal value of fibrosis for LV myocardial biopsy material is 5.4±2.2%.

Categorical variables were expressed as percentages and continuous variables as mean ± standard deviation unless otherwise specified. Continuous variables were compared between groups using an unpaired t-test (for normally distributed variables) or the Mann–Whitney U-test (for non-normally distributed variables). For comparison between baseline and follow-up, a paired Student's t-test was applied or a Wilcoxon test (for non-normally distributed variables). Chi-square test (or Fisher exact test) was used to compare categorical variables. Univariable binary logistic regression models (Wald method, p=0.05/0.20 for covariate inclusion/exclusion) were used in conjunction with the area under the curve (ROC) to assess the best cutoff-point (highest sensitivity and specificity) of a continuous variable to predict a particular outcome of all-cause death and all-cause death and cardiovascular hospitalization. The Kaplan-Meier and Cox models were used to assess survival times after surgery for all-cause death, for non-fatal cardiovascular hospitalization, and for the composite of all-cause death or cardiovascular hospitalization, and the log-rank test was used to compare survival curves. All reported probability values are two-tailed, and p<0.05 was considered statistically significant. Analyses were performed with the IBM® SPSS® Statistics software package (version 22.0) (SPSS Inc, Chicago, IL, USA).

At the end of 5.0±2.2 years of follow-up, there were seven deaths (12.5%) and four non-fatal cardiovascular hospitalizations (7.1%, two patients for HF, one for de novo atrial fibrillation and one for biologic prosthesis dysfunction).

In patients with an event (all-cause death or cardiovascular hospitalization) there were no significant differences in baseline clinical characteristics (Table 3) compared with those who did note experience events. On the preoperative echocardiogram, these patients had worse longitudinal function, with a lower peak systolic annular velocity (Sm) (Table 3), and no differences in AS severity, LV mass or EF.

There was a positive association between the level of fibrosis, assessed by CVF, and clinical outcomes (Figure 2). Patients who died had a significantly higher degree of fibrosis at the time of surgery (27.1±20.7% vs. 15.4±11.8%, p=0.035), and the same trend was observed for non-fatal cardiovascular hospitalization (24.8±17.7% vs. 6.7±3.2%, p=0.114). Myocardial fibrosis levels were also higher in those with the composite outcome of all-cause death or non-fatal cardiovascular hospitalization (24.0±18.2% vs. 15.3±12.0%, p=0.038).

Figure 2.

Collagen volume fraction level in patients with clinical outcomes. A: all-cause death, B: non-fatal cardiovascular hospitalization, C: all-cause death or non-fatal cardiovascular hospitalization. CVF: collagen volume fraction.

(0.19MB).

After multivariate Cox regression analysis, CVF was the only independent predictor of all-cause death (hazard ratio (HR) 1.88; 95% confidence interval (CI): 1.08-3.29, for each 10% increase; p=0.026) and all-cause death or cardiovascular hospitalization (HR 1.73; 95%CI:1.03-2.911, for each 10% increase; p=0.038) (Table 4).

A cut-off of 15.4% for CVF was used to calculate probability of survival free of events. This value was chosen after ROC curve analysis to assess the best CVF cutoff-point to predict the outcome of all-cause death (area under curve (AUC) 0.75, p=0.036) and all-cause death and cardiovascular hospitalization (AUC 0.92, p=0.038). Comparative characteristics of patients with CVF≥15.4% vs. CVF<15.4% can be seen in Table 5 and Table 6. Patients with CVF≥15.4% had a lower probability of survival (37.5% vs. 97.0%, p=0.001), lower probability of survival free of non-fatal cardiovascular hospitalization (84.4% vs. 94.1%, p=0.018) and lower probability of survival free of the composite of all-cause death or non-fatal cardiovascular hospitalization (0% vs. 91.2%, p<0.001) (Figure 3).

Figure 3.

Collagen volume fraction amount and long-term survival free of events after aortic valve replacement. CVF: collagen volume fraction.

(0.18MB).

Patients with residual LVH after surgery had a lower survival free of a combined event (all-cause death or non-fatal cardiovascular hospitalization), when compared with those with normal LVM (70.8% vs 92.6%, p=0.037) (Figure 4). Moreover, patients with Sm of <4.9 cm/s had significantly worse survival (81.6% vs. 100.0%, p=0.021) and worse survival free from death or non-fatal cardiovascular hospitalization (55.8% vs. 96.3%, p=0.017; AUC 0.77, p=0.027) (Figure 5).

Figure 4.

Residual left ventricular hypertrophy (LVH) and long-term survival free of events after aortic valve replacement. LVM: left ventricular mass; LVH: left ventricular hypertrophy.

(0.19MB).

Figure 5.

Peak systolic annular velocity (Sm) and long-term survival free of events after aortic valve replacement.

(0.18MB).

Fibrosis is an early morphological alteration in patients with AS and has been cited as one of the reasons for impaired LVH regression after AVR.13 Once established, fibrosis is a major determinant of diastolic and systolic dysfunction and it is one of the structural substrates for arrhythmogenicity, thus playing a major role in sudden death and the progression of HF.4,14 While myocyte hypertrophy is dependent on load, fibrosis seems also to be regulated by non-hemodynamic factors, such as neurohormones.15

In LVH associated with AS, there is an increased production of collagen and a shift toward inhibition of collagen degradation.16–18 When compared with controls, myocardial biopsies of AS patients have higher expression of collagens and an up-regulation of tissue inhibitor of metalloproteinase-1 (TIMP1) and tissue inhibitor of metalloproteinase-2 (TIMP2) mRNA, favoring inhibition of collagen degradation, which significantly correlates with the degree of fibrosis.18 Experimental studies have described total regression of matrix metallopeptidase (MMP) and TIMP gene expression as well as an association between changes in LVMI and MMP/TIMP gene expression after corrective surgical therapy and LV hypertrophy regression.19 The renin-angiotensin-aldosterone system seems to be a key factor in this process. Mechanical stretch induces local production of angiotensin II, which in turn stimulates the release of multiple growth factors and cytokines from cardiac fibroblasts that act in an autocrine and paracrine fashion, affecting the progression of hypertrophy and remodeling.20–22

The noninvasive evaluation of the level of myocardial fibrosis has been object of several studies, using surrogate echocardiographic parameters or CMR imaging.

Cardiac magnet resonance imaging is currently the gold standard method for measuring LV wall thickness, mass, volume and EF. This technique shows structural changes in the myocardium, including replacement fibrosis with LGE and expansion of extracellular volume using T1 mapping.23 LGE is usually associated with more advanced forms of myocardial disease, worse systolic function and higher LV end-diastolic volumes,24,25 and, therefore, worse prognosis.5 A recent meta-analysis of 2032 patients found that LGE has a substantial value in all-cause mortality and cardiovascular mortality prediction.26 However, LGE has the limitation of only identifying regional differences in replacement myocardial fibrosis, and can miss diffuse interstitial fibrosis, which is an earlier event in disease progression. Nevertheless, T1 mapping, by measuring the expansion of the extracellular volume, reflects in part alterations in diffuse myocardial fibrosis, which is a reversible early form of fibrosis.13 Chin et al. showed, in a large prospective cohort, that the use of T1 mapping in combination with LGE leads to the categorization of patients in groups with different prognostic values.27 However, further studies are needed to demonstrate whether this technique can assess the transition from hypertrophy to HF in AS.

In severe AS, the degree of myocardial fibrosis is crucial in the transition from compensated hypertrophy to HF.4 As fibrosis worsens, LV filling pressures increase and, later on, EF decreases. However, EF is reduced only in end-stage disease since it is related to global radial function,28 and fibrotic changes in AS hypertrophic hearts are initially subendocardial and affect basal segments (where regional wall stress is highest). This will principally have an impact on longitudinal function, which is not well represented by EF.11 Mitral ring displacement and strain imaging seem to have overcome EF limitations in evaluation of early changes in systolic function. In a recent study, in patients with symptomatic AS and severe fibrosis, radial function was relatively preserved, while mitral ring displacement, a surrogate of overall longitudinal function of the septum, was reduced and predicted functional improvement.11 Furthermore, using strain imaging, evidence of subclinical systolic LV dysfunction with decreased LV strain and strain-rate can be seen in patients with severe AS and preserved LVEF.29,30

In our study, we used peak systolic annular velocity (Sm) to assess longitudinal function and found that patients with Sm<4.9 cm/s had worse survival and survival free of death or non-fatal cardiovascular hospitalization. Our data support the prognostic importance of assessing early parameters of systolic function in AS.