Thirty-eight acromegalic patients were included in the present study. Thirteen were excluded due to inadequate image quality. The mean age of the remaining patients was 57.2±13.6 years and seven were male. Their data were compared to an age- and gender-matched control population, which consisted of 34 healthy volunteers (mean age 52.7±4.9 years, 15 male). The diagnosis of acromegaly was based on current clinical standards: typical clinical features, elevated HGH levels, and elevated IGF-1 levels insuppressible with an oral glucose tolerance test.5 The acromegalic group was further divided based on the activity of the disease: if serum HGH and/or IGF-1 concentration was over the diagnostic threshold, acromegaly was considered to be active.5,8 Our study was approved by the human research ethics committee of our institution and complied with the Declaration of Helsinki. All of the patients and healthy volunteers gave informed consent. Our acromegalic datasets came from the MAGYAR-Path Study (Motion Analysis of the heart and Great vessels bY three-dimensionAl speckle-tRacking echocardiography in Pathological cases), which was developed to assess the diagnostic and prognostic value of 3D-STE-derived parameters (‘Magyar’ means ‘Hungarian’ in the Hungarian language).

All healthy volunteers and acromegalic patients underwent a complete two-dimensional (2D) transthoracic echocardiographic study using a Toshiba ArtidaTM imaging system (Toshiba Medical Systems, Tokyo, Japan) with a PST-30SBP phased-array transducer (1-5 MHz). The examinations were performed according to current clinical standards. In all cases LV dimensions, volumes and ejection fraction (LVEF) and left atrial (LA) dimensions were measured and complete 2D Doppler and tissue Doppler studies were performed.9

For the 3D-STE examinations, the same Toshiba Artida™ system was used, with a PST-25SX matrix- array transducer (Toshiba Medical Systems, Tokyo, Japan).7 In each case, a pyramid-shaped full-volume 3D dataset was obtained. The complete dataset consisted of six wedge-shaped subvolumes, which were recorded during six RR intervals and a single breath-hold. The subvolumes were set to be as narrow as possible to improve spatial resolution, thus improving later endocardial border delineation. The recorded datasets were analyzed offline using the supplied 3D Wall Motion Tracking software, version 2.7 (Toshiba Medical Systems, Tokyo, Japan). From the datasets, the software automatically reconstructed apical 4-chamber and 2-chamber views, then the reader specified the three cross-sectional planes manually, using the guide planes to help standardize LV measurements. After setting up the planes, the reader then specified the septal and lateral parts of the mitral annulus and the LV apex, after which the software automatically detected the endocardial border and created a 3D LV cast (Figure 1).

Figure 1.

Three-dimensional speckle-tracking echocardiographic study of an acromegalic patient. Apical 4-chamber view (A) and apical 2-chamber view (B) are automatically selected by the software. Cross-sectional planes are at the apical (C3), midventricular (C5) and basal (C7) left ventricular (LV) levels. The three-dimensional LV model (D) and the corresponding volumetric parameters (E) are shown along with segmental LV radial strain curves (F).

(0,39MB).

3D-STE was used to measure LV end-diastolic volume (LVEDV), LV end-systolic volume (LVESV), LVEF and LV mass based on the above-mentioned LV cast.

3D-STE strain measurements were performed on the LV cast using the 16-segment LV model.10 In each segment, three unidirectional strain parameters were assessed: radial (RS) (thickening and thinning of the segment), longitudinal (LS) (lengthening and shortening of the segment) and circumferential (CS) (widening and narrowing of the segment). Based on these strain parameters, the software calculated the following complex strains: area (AS) (combination of LS and CS) and 3D (3DS) (combination of RS, LS and CS).

All data are reported as mean ± standard deviation. Significance was established for p values less than 0.05. The Shapiro-Wilks test was used to check for normal distribution. Levene's test for equality of variances was used to test homogeneity of variance. For normally distributed datasets, the two-tailed Student's t test was used for comparisons, however if the dataset did not follow a normal distribution, the Mann-Whitney-Wilcoxon test was used. For categorical variables, Fisher's exact test was performed. To establish significant relationships between independent variables, Pearson's correlation coefficients were calculated. The statistical analysis was performed using RStudio (RStudio Team, RStudio: Integrated Development for R. RStudio, Inc., Boston, MA, 2015). For offline data extraction and analysis, a commercial software package was used (MATLAB 8.6, The MathWorks Inc., Natick, MA, 2015).

Significant differences were identified between acromegalic patients and healthy controls regarding the prevalence of hypertension (p<0.0001), hypercholesterolemia (p<0.0001) and diabetes (p=0.01) (Table 1).

The active acromegaly subgroup consisted of 14 patients (mean age: 58.6±14.6 years, five males). Their mean serum HGH level was 5.5±3.9 ng/ml, IGF-1 level was 502.3±366.0 ng/ml and IGF-1 index (IGF-1 level/upper limit of normal) was 2.0±1.1. The inactive acromegaly group included 11 patients (mean age: 54.0±12.9, 2 males). Their mean serum HGH level was 4.5±8.1 ng/ml, IGF-1 level was 216.9±129.0 ng/ml and IGF-1 index was 0.9±0.6.

Significant differences were seen in LA diameter (p=0.02), LV end-diastolic diameter (LVEDD) (p=0.006) and volume (LVEDV) (p=0.002), interventricular septal thickness (p=0.02), and LV posterior wall thickness (p=0.007) between the acromegalic group and healthy controls. Body surface area (BSA) was also significantly different between controls and acromegalic patients (p=0.01). Furthermore, late transmitral flow velocity A(p=0.05) and E/A ratio (p=0.004) also differed significantly between the two groups (Table 1).

3D-STE-derived LVEDV (p=0.0003) and LVESV (p=0.005) proved to be significantly different between all the acromegalic patients and controls. LVEDV (p=0.006) and LVEF (p=0.02) differed significantly between the active acromegaly subgroup and controls, while LVEDV (p=0.01) and LVESV (p=0.01) were significantly different between the inactive disease subgroup and controls. LVEF (p=0.02) was significantly higher in active than in inactive disease (Tables 2 and 3).

Global and mean segmental RS (p=0.01 and p=0.009, respectively) were significantly higher in acromegalic patients compared to controls. Among regional strain parameters, basal and apical RS (p=0.04 and p=0.0004, respectively) were significantly different between acromegalic patients and controls. Patients with active disease had significantly higher global and mean segmental RS (p=0.03 and p=0.03, respectively) compared to controls; in addition, basal and apical RS (p=0.04 and p=0.002, respectively) and basal CS (p=0.01) were significantly higher compared to controls. In patients with inactive disease, no significant differences were detected in LV strain parameters compared to the control group. Between active and inactive disease groups, only basal CS (p=0.02) was found to be significantly different (Tables 2 and 3).

In the overall acromegaly group, there was a significant negative correlation between HGH levels and apical 3DS (r=-0.40, p=0.05). No significant correlation was detected between any hormone level and LV strains in the active disease group. In the inactive disease group, IGF-1 index correlated significantly with midventricular RS (r=0.60, p=0.05), global and mean segmental RS (r=0.66, p=0.03 and r=0.66, p=0.03, respectively) and global 3DS (r=0.61, p=0.05). Furthermore, there was a significant correlation between mean segmental RS and IGF-1 levels (r=0.61, p=0.05) in this group.