Heart function following heart transplantation (HTx) is influenced by numerous factors. It is typically evaluated using transthoracic echocardiography, but reference values are currently unavailable for this context. The primary aim of the present study was to derive echocardiographic reference values for chamber size and function, including cardiac mechanics, in clinically stable HTx patients.
The study enrolled 124 healthy HTx patients examined prospectively. Patients underwent comprehensive two-dimensional echocardiographic examinations according to contemporary guidelines. Results were compared with recognized reference values for healthy subjects.
Compared with guidelines, larger atrial dimensions were seen in HTx patients. Left ventricular (LV) diastolic volume was smaller, and LV wall thickness was increased. With respect to LV function, both ejection fraction (62 ± 7%, P < .01) and global longitudinal strain (−16.5 ± 3.3%, P < .0001) were lower. All measures of right ventricular (RV) size were greater than reference values ( P < .0001), and all measures of RV function were reduced (tricuspid annular plane systolic excursion 15 ± 4 mm [ P < .0001], RV systolic tissue Doppler velocity 10 ± 6 cm/sec [ P < .0001], fractional area change 40 ± 8% [ P < .0001], and RV free wall strain −16.9 ± 4.2% [ P < .0001]). Ejection fraction and LV global longitudinal strain were significantly lower in patients with previous rejection.
The findings of this study indicate that the distribution of routinely used echocardiographic measures differs between stable HTx patients and healthy subjects. In particular, markedly larger RV and atrial volumes and mild reductions in both LV and RV longitudinal strain were evident. The observed differences could be clinically relevant in the assessment of HTx patients, and specific reference values should be applied in this context.
Right ventricular enlargement is present in heart transplant patients.
Conventional parameters of right ventricular function are reduced after heart transplantation.
Ventricular longitudinal strain is significantly lower following heart transplantation.
Previous rejection has an impact on left ventricular ejection fraction and longitudinal strain.
Orthotopic heart transplantation (HTx) offers the possibility of long-term survival in patients with end-stage heart disease. Since the first HTx was performed in 1967, improvements in operative techniques and postoperative therapy, including immunosuppression regimes, have led to reduced operative mortality and increased long-term survival. Two separate surgical approaches have been used. In the older biatrial technique, large parts of the native atria in the recipient are retained, and the donor heart is sutured on to the native atria. In the new bicaval technique, the donor heart is connected through bicaval anastomosis ( Figure 1 ). The latter offers physiologic benefit in terms of better preserving the anatomic configuration and has proved superior in maintaining hemodynamic function, reducing the need of pacemaker and improved survival. Post-transplantation annual routine follow-up includes echocardiography and exercise electrocardiographic at most centers today. In clinical practice, echocardiographic assessment of the transplanted heart is complicated by numerous factors that can affect myocardial function, rendering the use of normal values derived from healthy subjects unsatisfactory.
Thoracic surgery, including pericardiotomy, influence contractility and thereby motion of the heart because of disturbed ventricular interdependence. Furthermore, donor age and ischemic time as well as cardiac allograft vasculopathy (CAV) after HTx may additionally influence allograft function and thereby survival. Function of the donor heart, especially the right ventricle, may be affected by the cause of donor death and procedures related to explantation, storage, and implantation in the recipient.
In normal subjects, age-related changes in ventricular function are well described, and echocardiographic normal values are available, but data concerning HTx patients are sparse.
In the past decade, several new echocardiographic techniques, such as Doppler imaging and two-dimensional (2D) speckle-tracking echocardiography (STE)–derived strain, have been introduced and refined. These offer the possibility to better detect minor, but significant, changes in myocardial function and contractility.
Previous echocardiographic studies aimed at assessing ventricular function and cardiac mechanics in HTx patients are sparse, mostly limited to a small number of patients, and conducted at an early time point after HTx. One of the studies used several different vendors and included only patients examined 1 year after HTx, making comparison of the findings difficult. The remaining three studies were limited to cohorts of 20 to 51 patients and were conducted with different software (Vivid E7; GE Healthcare, Little Chalfont, United Kingdom) than our study. However, it has recently been stated that the concerns of intervendor software variability when measuring left ventricular global longitudinal strain (LVGLS) are of subordinate importance. This should allow wider clinical use of LVGLS. The findings of these studies emphasize the need for specific reference values for the transplanted heart, preferably derived from a larger cohort not limited to a specific time point after HTx.
The primary aim of this study was to investigate whether transplanted hearts exhibit differences regarding ventricular size, function, and cardiac mechanics compared with current published reference values for healthy non-HTx subjects and to describe echocardiographic reference values in clinically stable HTx patients applicable in clinical practice. Furthermore, we sought to describe possible differences related to allograft age, existing CAV, or previous severe rejection.
Echocardiographic imaging and all offline calculations were performed by two experienced senior cardiac sonographers. At the time of examination all patients were in sinus rhythm or a regular paced rhythm ( n = 5). Patients were enrolled at their routine follow-up visits between 2012 and 2015 and were included only once. A total of 137 patients were enrolled, but after exclusion, 124 patients remained available for analysis. Exclusion criteria were biopsy-proven rejection at time of examination ( n = 1), atrial fibrillation ( n = 1), more than mild mitral or aortic valvular regurgitation ( n = 2), previous myocardial infarction or left ventricular (LV) ejection fraction < 45% ( n = 5), and insufficient image quality ( n = 4). A total of 23 patients were operated with the biatrial technique (≥15 years since HTx) and 101 patients with the bicaval technique (<15 years since HTx).
Patients were examined using 2D echocardiography with a Philips iE33 ultrasound system equipped with an S5-1 transducer (Philips Healthcare, Eindhoven, the Netherlands). At our center, coronary angiography is routinely conducted before and 1 year after HTx, followed by computed tomographic angiography in alternating years thereafter. Coronary angiography is conducted if computed tomographic angiography raises suspicion of CAV. In our study, 23 patients were diagnosed with mild CAV and nine patients with moderate CAV on the basis of their medical records. At the time of enrollment, no clinical manifestations of rejection were present. Measurements, valve assessment, and calculation of standard echocardiographic parameters were performed according to American Society of Echocardiography guidelines. The study complied with the Declaration of Helsinki and was approved by the local scientific ethics committee in Lund, Sweden (210/114, 210/442, and 2011/777) after informed consent was obtained.
Systolic and diastolic LV volumes were calculated using the modified biplane Simpson method in apical two- and four- chamber views. RV basal diameter, mid diameter, and area were obtained from the nonforeshortened apical four-chamber view. When applicable, values were correlated to body surface area (BSA) according to guidelines. Results describing atrial size are subclassified according to whether surgical anastomosis was by the atrial technique or the caval technique.
Conventional Assessment of LV and RV Function
Ejection fraction was calculated according to modified Simpson biplane method of disks in apical two- and four-chamber views. Tissue Doppler obtained from the LV lateral wall was used to calculate the E/e′ ratio.
RV function was assessed using tricuspid annular plane systolic excursion (TAPSE), RV fraction area change and Doppler tissue imaging–derived tricuspid lateral annular systolic velocity (S′). All measurements were performed according to recommendations.
Three-beat cine-loop clips were recorded during unforced end-respiratory apnea. Grayscale parasternal and apical views were recorded for speckle-tracking, and STE was assessed offline using commercial software (CMQ, QLAB 10.1 version 1.0, Philips iE33; Philips Healthcare). Acquisition of 2D images for STE was performed with care (mindful to avoid foreshortening) and adjusted to appropriate gain settings. Sector width was adjusted to allow complete myocardial visualization, and frame rate was optimized to a minimum of 50 Hz. Strain evaluation was done according to the software manufactures recommendations. The region of interest of the endocardium was manually traced using a point-and-click approach in a single frame at end-diastole. Strain values were automatically generated by the software and presented for each segment together with the mean global value for each view (the parasternal short-axis and apical views).
Circumferential strain was obtained from the short-axis view, aiming to visualize three levels of the left ventricle, apical, papillary muscle, and basal, thereby assessing LV global circumferential strain (LVGCS). LVGLS was obtained from the three conventional apical views.
RV global longitudinal strain was measured in the four-chamber view and calculated by the software as the mean peak systolic strain of all seven segments. RV lateral free wall strain (RVfree) was manually calculated by averaging three regional peak systolic strains along the entire RV free wall (basal, mid, and apical). Strain is expressed as the percentage change in length from the original length. Given this, LVGLS, LVGCS, RV global longitudinal strain, and RVfree are defined as negative values (shortening of the myocardial fibers) and are expressed as percentage changes.
All continuous data conforming to a normal distribution are presented as mean ± SD and 95% CI. Assumptions of normality were confirmed by visual inspection of histograms. Reference range are derived from 95th percentiles. Echocardiographic measures were compared with previously reported distributions in the general population using Welch’s unequal-variance t tests because equal variance in these two contexts could not be assumed. Distributions from the Normal Reference Ranges for Echocardiography study were used for comparison of standard size and function parameters. RV function parameters, which were not available in the Normal Reference Ranges for Echocardiography study, were obtained from current echocardiographic guidelines, whereas all LV strain parameters were obtained from a meta-analysis. The correlations of recipient age, allograft age, hemodynamic measurements, and blood pressure with echocardiographic parameters were explored using Spearman rank correlation coefficients because theses variables were not normally distributed. The influence of the different surgical techniques with regard to biatrial or bicaval anastomosis were compared for differences in left atrial size using the Mann-Whitney U test. The influence of hypertension, diabetes, and CAV on systolic ventricular function was explored using Mann-Whitney U test. Findings were considered statistically significant at P < .05. Statistical analysis was performed using SPSS Statistics version 22.0 (IBM, Armonk, NY).
The study enrolled a total of 137 individual HTx patients who were examined prospectively (2013–2015), according to an extended echocardiographic protocol in conjunction with yearly routine checkups at Skane University Hospital (Lund, Sweden). In our study population, 78 patients had no history of treatment requiring rejection (defined as grade ≥ 3A). A total of 29 patients had one episode, five patients had two episodes, and one patient had four episodes of previous rejection requiring treatment. All baseline patient characteristics of the 124 HTx patients are shown in Table 1 .
|Parameters before HTx||Mean ± SD||Range|
|Biatrial ( n = 23)||Bicaval ( n = 101)||Total||Total|
|Donor age at HTx (y)||37 ± 14||45 ± 16||43 ± 16||14–72|
|Recipient age at HTx (y)||37 ± 19||49 ± 15||47 ± 17||17–69|
|Invasive PVR (Wood units), recipient||2.7 ± 1.2||3.0 ± 1.6||2.9 ± 1.5||0.3–10.8|
|Invasive PASP (mm Hg), recipient||46 ± 15||44 ± 14||44 ± 14||16–86|
|Index donor − recipient height (%)||98||102||101||NA|
|Index donor − recipient weight (%)||88||100||99||NA|
|Ischemic time (min)||172 ± 77||202 ± 48||201 ± 50||85–316|
|Parameters at examination|
|Time since transplantation (y)||17 ± 5||4.1 ± 3.6||7 ± 6||1–25|
|Treatment-requiring (grade > 3A) rejection||8||31||39||0–4|
|Mean time between last rejection and inclusion (y)||17.1||6.7||9.6||1–25|
|Recipient BSA (m 2 )||1.9 ± 0.2||1.9 ± 0.3||1.9 ± 0.2||1.4–2.7|
|Recipient age (y)||55 ± 19||53 ± 15||54 ± 15||18–76|
|SBP (mm Hg)||136 ± 18||136 ± 17||136 ± 17||95–182|
|DBP (mm Hg)||76 ± 20||81 ± 9||80 ± 12||42–110|
|Heart rate (beats/min)||80 ± 12||85 ± 11||84 ± 11||60–120|
|QRS duration (msec)||133 ± 30||119 ± 26||122 ± 27||64–180|
|Invasive PVR (Wood units) ( n = 39)||NA||1.4 ± 0.7||1.4 ± 0.7||0.4–2.9|
|Invasive PASP (mm Hg) ( n = 39)||NA||27.5 ± 11.2||27.5 ± 11.2||7–64|
|Invasive MPAP (mm Hg) ( n = 39)||NA||18 ± 7.7||18 ± 7.7||8–43|
Mean atrial volume was the only parameter differing depending on which surgical technique was used. In the biatrial group ( n = 23), left atrial volume was 53 ± 23 mL/m 2 compared with 41 ± 16 mL/m 2 in the bicaval group ( n = 101; P < .001), and right atrial volume was 38 ± 21 versus 29 ± 14 mL/m 2 ( P < .001), respectively. Indexed atrial size in patients operated with the currently used bicaval technique were significantly larger ( P < .001) than reference values for both the left and right atria. Full data regarding atrial size are shown in Tables 2 and 3 .
( N = 124)
( n = 23)
( n = 101)
|PLAP (mm)||46 ± 9||48 ± 11||45 ± 8||NS|
|LA volume (mL)||79 ± 30||96 ± 47||75 ± 23||<.001|
|LA volume/BSA (mL/m 2 )||42 ± 16||53 ± 23||39 ± 13||<.001|
|RA volume (mL)||56 ± 25||71 ± 39||52 ± 19||<.001|
|RA volume/BSA (mL/m 2 )||29 ± 14||38 ± 21||27 ± 11||<.001|
|Total, mean ± SD||Range (2.5th to 97.5th percentile)||95% CI of mean|
|Atria bicaval group|
|PLAP||46 ± 9||29–61||43–47|
|LA volume (mL)||76 ± 30||29–121||71–79|
|LA volume/BSA (mL/m 2 )||41 ± 16||13–65||36–42|
|RA volume (mL)||54 ± 25||14–90||48–56|
|RA volume/BSA (mL/m 2 )||29 ± 14||6–49||25–29|
|IVSd (mm)||10.9 ± 2.2||7–15||9.6–11.4|
|LVPWd (mm)||10.5 ± 1.9||7–14||9.2–10.6|
|LVEDd (mm)||46 ± 6||34–57||44–48|
|LVEDd/BSA (mm/m 2 )||24 ± 4||17–32||23–26|
|LVSd (mm)||29 ± 5||20–40||27–31|
|LVSd/BSA (mm/m 2 )||16 ± 3||10–23||14–17|
|LV mass (g)||160 ± 50||87–276||144–175|
|LV mass index (g/m 2 )||90 ± 25||50–142||77–93|
|LVOT diameter (mm)||22 ± 4||18–25||18–23|
|LVEDV (mL)||88 ± 24||47–136||78–97|
|LVEDV/BSA (mL/m 2 )||47 ± 14||27–71||42–51|
|LVESV (mL)||35 ± 12||17–65||30–40|
|LVESV/BSA (mL/m 2 )||19 ± 7||9–33||16–21|
|RV outflow tract (mm)||31 ± 6||25–39||29–33|
|RV basal diameter (mm)||37 ± 6||25–52||35–39|
|RV basal diameter/BSA (mm/m 2 )||19 ± 4||13–25||18–22|
|RV mid diameter (mm)||33 ± 6||20–45||31–35|
|RV mid diameter/BSA (mm/m 2 )||18 ± 6||11–25||16–20|
|RV longitudinal diameter (mm)||68 ± 12||31–80||63–73|
|RV longitudinal diameter/BSA (mm/m 2 )||36 ± 7||20–49||27–41|
|RVEDA (cm 2 )||20 ± 5||12–28||18–22|
|RVEDA/BSA (cm 2 /m 2 )||11 ± 4||6–16||10–11|
|RVESA (cm 2 )||12 ± 3||7–18||11–13|
|RVESA/BSA (cm 2 /m 2 )||6 ± 1||4–9||6–7|
Compared with published normal values, mean diastolic LV diameter was larger ( P < .05), and interventricular septal and posterior wall thickness as well as LV mass were all increased ( P < .0001). Mean LV diastolic volume was smaller than reference values ( P < .05), but no difference was found in mean LV systolic volume ( P = NS). Both mean RV inflow (37 ± 6 mm) and RV mid diameter (33 ± 6 mm) were enlarged compared with reference values ( P < .0001). Furthermore, indexed RV area measures showed enlargement of both RV diastolic area ( P < .01) and RV systolic area ( P < .0001) compared with reference values. Complete data regarding LV and RV size in absolute values and indexed to BSA are shown in Table 3 . Comparisons with reference values are shown in Table 4 .
|Parameter||HTx patients||Normal reference||P|
|Atria bicaval group|
|PLAP||46 ± 9||19 ± 2||<.001|
|LA volume (mL)||76 ± 30||47 ± 13||<.001|
|LA volume/BSA (mL/m 2 )||41 ± 16||26 ± 6||<.001|
|RA volume (mL)||54 ± 25||38 ± 14||<.001|
|RA volume/BSA (mL/m 2 )||29 ± 14||21 ± 7||<.001|
|IVSd (mm)||10.9 ± 2.2||8.6 ± 1.6||<.0001|
|LVPWd (mm)||10.5 ± 1.9||8.8 ± 1.5||<.0001|
|LVEDd (mm)||46 ± 6||44 ± 5||.05|
|LVSd (mm)||29 ± 5||30 ± 5||NS|
|LV mass (g)||160 ± 50||127 ± 37||<.0001|
|LVEDV (mL)||88 ± 24||93 ± 25||<.05|
|LVEDV/BSA (mL/m 2 )||47 ± 14||51 ± 11||<.001|
|LVESV (mL)||35 ± 12||34 ± 11||NS|
|LVESV/BSA (mL/m 2 )||19 ± 7||19 ± 5||NS|
|RV outflow tract (mm)||31 ± 6||32 ± 5||NS|
|RV basal diameter (mm)||37 ± 6||34 ± 6||<.0001|
|RV mid diameter (mm)||33 ± 6||28 ± 6||<.0001|
|RV longitudinal diameter (mm)||68 ± 12||68 ± 8||NS|
|RVEDA (cm 2 )||20 ± 5||17 ± 4||<.0001|
|RVEDA/BSA (cm 2 /m 2 )||11 ± 4||10 ± 2||<.01|
|RVESA (cm 2 )||12 ± 3||9 ± 3||<.0001|
|RVESA/BSA (cm 2 /m 2 )||6 ± 1||5 ± 1||<.0001|
Conventional Assessment of Ventricular Function
Mean LV fractional shortening was 35 ± 9%, and ejection fraction was 62 ± 7%, which is lower ( P < .01) than the normal reference value but still within the normal range. Concerning conventional parameters of diastolic function, mean E/A ratio was 1.8 ± 0.6, and mean E/e′ ratio was 7.1 ± 3.0. In our material, all conventional parameters measuring systolic RV function were decreased ( P < .0001) compared with current guidelines. Mean TAPSE was 15 ± 4 mm, mean S′ was 10 ± 6 mm/sec, and mean RV fractional area change was 40 ± 8%. Parameters of ventricular function are presented in Table 5 , and comparisons with reference values are shown in Table 6 . Results for LV and RV function are also illustrated in Figures 2 and 3 , respectively. No correlation between diabetes or hypertension (defined as systolic blood pressure > 140 mm Hg or diastolic blood pressure > 90 mm Hg) and conventional parameters of systolic function was found ( P = NS, data not shown).
|Parameter||Total, mean ± SD||Range (2.5th to 97.5th percentile)||95% CI of the mean|
|FS (%)||35 ± 9||20–54||33–40|
|LVEF (%)||62 ± 7||48–78||60–65|
|S′ mean (cm/sec)||8.2 ± 2.2||6.1–10.5||6–10.4|
|S′ septal (cm/sec)||6.6 ± 1.6||4.2–12.9||5.8–7.2|
|S′ lateral (cm/sec)||8.6 ± 2.2||6.4–10.8||6.4–10.8|
|S′ anterior (cm/sec)||7.9 ± 1.9||6–9.9||6–9.8|
|S′ inferior (cm/sec)||7.4 ± 1.9||5.5–9.3||5.5–9.3|
|MV E (cm/sec)||80 ± 21||50–120||75–87|
|MV A (cm/sec)||46 ± 13||26–90||41–51|
|MV deceleration time (msec)||156 ± 31||101–120||146–165|
|E/A ratio||1.8 ± 0.6||0.8–3.2||1.7–2.1|
|e′ (lateral) (cm/sec)||8 ± 3.1||5.5–11.1||7.2–9.1|
|E/e′ (lateral)||7.1 ± 3.0||3.1–14.7||6.4–8.4|
|LVSV (mL)||62 ± 16||47–79||55–66|
|LVGCS (%)||−22.9 ± 6.3||9–22||21–25|
|LVGLS (%)||−16.5 ± 3.3||12–35||15–18|
|RVFAC (%)||40 ± 8||20–54||37–43|
|TAPSE (mm)||15 ± 4||8–23||14–16|
|S′ RV (cm/sec)||10 ± 6||5.7–15.3||9–11|
|RIMP||0.29 ± 0.18||0.10–0.96||0.24–0–35|
|IVA (cm/sec 2 )||2.2 ± 1.0||0.21–4.31||1.8–2.5|
|RVOT TVI (cm)||11.1 ± 3.2||9.7–12.6||10.4–11.9|
|RVGLS (%)||−15.3 ± 4.1||8–22||14–18|
|RVfree (%)||−16.9 ± 4.2||7–24||15–18|