Article Text
Abstract
Background and aims Right ventricular (RV) functional assessment in premature infants includes basal longitudinal strain (RV BLS), RV systolic tissue Doppler velocity (RV s′), tricuspid annular plane systolic excursion (TAPSE) and RV fractional area change (FAC). A hyperdynamic left ventricle (LV) may influence RV measures of displacement (TAPSE) and velocity (RV s′) but not measures of relative change of length (RV BLS) or area (FAC). We aimed to explore this hypothesis in preterm infants with a patent ductus arteriosus (PDA).
Methods We measured LV function (ejection fraction (LV EF); left ventricular output) and RV function (RV BLS; RV s′; TAPSE; FAC) on days 1, 2 and 5–7 in infants <29 weeks. The cohort was divided based on PDA presence by days 5–7. LV and RV function measurements were compared between the groups using two-way analysis of variance with repeated measures.
Results 121 infants with a mean (SD) gestation and birth weight of 26.8 (1.4) weeks and 968 (250) g were enrolled. By days 5–7, the PDA remained open in 83 (69%), with evidence of hyperdynamic LV function. There was no difference in RV s’ (5.3 (0.9) vs 5.1 (1.0) cm/s, p=0.3) or TAPSE (6.2 (1.3) vs 6.1 (1.2) mm, p=0.7) between infants with and without a PDA, but infants in the PDA group had lower RV FAC (41 (8) vs 47 (10) %, p<0.01) and lower RV BLS (−24.2 (5.0) vs −26.2 (4.1) %, p=0.03).
Conclusions LV influence on RV functional parameters must be taken into account when interpreting of RV function using those techniques.
- RV function
- Preterm Infants
- LV function
- Patent Ductus Arteriosus
Statistics from Altmetric.com
What is already known on this topic?
Persistent patent ductus arteriosus in preterm infants often results in a hyperdynamic left ventricle (LV).
Hyperdynamic LV can alter measures of right ventricular (RV) function.
Novel measurement techniques of RV function have been validated in preterm infants.
What this study adds?
A hyperdynamic LV influences the assessment of RV function in preterm infants.
The presence of a patent ductus arteriosus at 1 week of age is associated with lower values of RV strain and fractional area of change.
Introduction
Right ventricular (RV) performance is an important prognostic determinant of clinical status and long-term outcome in preterm infants with cardiopulmonary pathology.1 The assessment of RV function with conventional echocardiography is limited because of the complex geometry of the RV,2 and the unique physiological changes that occur during the transitional period in the first week of life.3–7 Emerging echocardiography modalities are now being employed to improve our assessment of RV systolic function in preterm infants.8 These methods include: RV systolic tissue Doppler velocity (RV s′), tissue Doppler-derived basal longitudinal strain (RV BLS); tricuspid annular plane systolic excursion (TAPSE) and fractional area change (FAC). Recent evidence suggests that those parameters can determine disease severity and predict adverse outcomes in term infants with pulmonary hypertension.9 However, their applicability in the clinical setting, the influence of various physiological states and the interaction between left ventricle (LV) and RV function in preterm infants has not been adequately studied.
These techniques have been validated in preterm infants,1 ,10–13 but there is still a shortage of available data that properly characterises the relative strengths and weaknesses, and therefore limits their applicability during the early postnatal period. Furthermore, the influence of LV function on these emerging measures of RV function is an underappreciated concept that has not been adequately studied in premature infants. This relationship must be interpreted in the context of change in both position and shape of certain regions of the RV. Specifically, myocardial wall motion is the change in position of a particular region of myocardium over a particular time period and can be measured by velocity (RV s′) and displacement (TAPSE). In contrast, myocardial wall deformation (RV BLS) and change in RV cavity area (FAC) describe the change in shape of a particular region (RV wall or RV cavity area) occurring from a baseline in diastole to its new shape in systole. Only deformation and FAC analysis can discriminate between active and passive movements of myocardial segments, and may actually provide a more accurate representation of RV function than TAPSE and RV s′. This is important to consider when applying the measurements in the preterm population, particularly in instances where passive RV movement may be increased due to a hyperdynamic LV, which often occurs in the presence of a haemodynamically significant patent ductus arteriosus (PDA).
To further examine this, we aimed to assess RV function using RV s′, TAPSE, RV BLS and FAC in preterm infants with and without a PDA and an accompanying hyperdynamic LV. We hypothesised that a hyperdynamic LV in the setting of a haemodynamically significant PDA would increase RV measures of displacement (TAPSE) and velocity (RV s′) but not measures of relative change of length (RV BLS) or area (FAC) when compared with infants without a PDA.
Methods
Study population
This was a prospective observational study of infants of <29 weeks' gestation recruited from the neonatal intensive care unit at the Rotunda Hospital, Dublin, Ireland between January 2013 and December 2015. We have recently published individual reference values of RV fractional area of change and strain imaging using a proportion of infants from this cohort, but have not reported the effect of LV function on novel quantitative measures of RV function.14 ,15 Ethical approval was granted from the local Ethics Committee and written informed consent was obtained from parents of all study participants before enrolment. Infants were excluded if they had a suspected or definite chromosomal abnormality or congenital heart disease other than a PDA or a patent foramen ovale, those that died within 1 week of age, or if a scan was not performed between days 5 and 7 because of unavailability of study investigators. Medical and/or surgical treatment of a PDA was not instituted during the study period in this cohort. In our centre, PDA closure is only attempted beyond the second week of life in infants on prolonged invasive ventilation. At the time of the days 5–7 scan, none of the infants were in receipt of inotropes, or had culture-positive sepsis. Perinatal variables and baseline characteristics were collected from the infants' medical records. Heart rate, blood pressure, mean airway pressure and mode of ventilation were noted during the echocardiography scans.
Echocardiographic assessment
Echocardiography was performed at a median (IQR) of 10 hours (7–13) (day 1), 43 hours (38–46) (day 2) and 143 hours (125–161) (days 5–7) using the GE Vivid I or Vivid S6 ultrasound system with a 10S or 12S probe (at a frequency of 8 MHz). All studies were performed using a standard functional echocardiography protocol.16 The images and video loops were stored as raw Digital Imaging and Communications in Medicine (DICOM) data and archived for analysis to be performed offline using the EchoPac software (GE, V.112, revision 1.3).
We performed the following echocardiography measurements: PDA diameter in two-dimension measured at the pulmonary end;14 diastolic to systolic ratio of flow across the PDA;17 left atrium to aortic root ratio (LA:Ao); LV end-diastolic diameter in millimetres (LVEDD); ejection fraction (EF) using the Simpson's biplane method; LV output (LVO) in mL/kg/min. In addition, a number of RV function measurements were obtained: TAPSE was measured from the apical four-chamber view using M-mode.11 FAC was measured using a focused RV four-chamber view.1 RV s′ was measured using pulse wave at the base of the RV free wall.18 Finally, RV peak systolic BLS (RV BLS) was obtained from the apical four-chamber view using tissue Doppler-derived deformation imaging as previously described.15 ,19 ,20 Our group and others have previously demonstrated the feasibility and reproducibility of all the RV measures described in this study.1 ,12 ,15
Statistical analysis
The cohort was divided into two groups based on the presence or absence of a PDA on days 5–7. Values were presented as means (SD) or medians (IQR) and compared using an independent t-test or a Mann-Whitney U test as appropriate. Categorical values were presented as count (per cent) and compared using χ2 or Fischer's exact test as appropriate. Two-way analysis of variance with repeated measures was used to compare change overtime between the two groups. Linear regression was used to assess the independent effect of LV function on RV function measurements while adjusting for gestation, birth weight, mode of ventilation and mean airway pressure. SPSS (V.22) was used to conduct the analysis. A p value <0.05 was considered to be statistically significant.
Results
One hundred and twenty-one infants with a mean (SD) gestation and birth weight of 26.8 (1.4) weeks and 968 (250) g were included in this study. On day 1, the PDA was present in 117 (97%) infants with a median diameter of 2.4 (2.0–2.9); the PDA was present in 102 (84%) on day 2 with a median diameter of 2.8 (2.2–3.2); by days 5–7, the PDA remained open in 83 (69%) infants (PDA group), and spontaneously closed in the remainder (no PDA group). Infants in the PDA group were of a slightly lower gestation and birth weight (table 1).
Infant characteristics at birth
On days 5–7, infants in the PDA group had a median (IQR) PDA diameter of 2.7 (2.2–3.2) mm and a non-restrictive flow pattern with a median systolic to diastolic ratio of 2.7 (1.7–4.4). In addition, infants in the PDA group demonstrated evidence of LV volume overload and hyperdynamic LV function compared with the no PDA group, with a lower diastolic blood pressure and a higher LA:Ao, LVEDD, EF and LVO (all p<0.01, table 2). Figure 1 illustrates the divergence in those parameters of LV overload and function between the two groups over the first week of age (figure 1).
Clinical and echocardiography parameters of left ventricular overload in the two groups on days 5–7
Markers of LV volume overload and function in the two groups over the first week of age. Infants in the PDA group (grey line) had higher LV end-diastolic diameter, LA:Ao, LVO and EF from as early as day 2 (*p<0.05, two-way analysis of variance with repeated measures). LV, left ventricle; LA:Ao, left atrial to aortic root ratio; EF, ejection fraction; LVO, left ventricular output; PDA, patent ductus arteriosus.
In the entire cohort, there was an overall increase in TAPSE (4.8 (1.2) vs 5.6 (1.3) vs 6.2 (1.3) mm, p<0.001), RV s′ (3.6 (0.9) vs 4.5 (1.0) vs 5.3 (0.9) cm/s, p<0.001), FAC (30 (9) vs 39 (8) vs 42 (9) %, p<0.001) and RV BLS (22.3 (4.7) vs 24.0 (4.7) vs 24.8 (4.9) %, p=0.001) over the three time points. On days 5–7, coinciding with LV volume overload and hyperdynamic LV function in the PDA group, there was no difference in RV s′ (5.3 (0.9) vs 5.1 (1.0) cm/s, p=0.3) or TAPSE (6.2 (1.3) vs 6.1 (1.2) mm, p=0.7) between infants with and without a PDA. During the same time period, infants in the PDA group had lower RV FAC (41 (8) vs 47 (10) %, p<0.01) and lower RV BLS (−24.2 (5.0) vs −26.2 (4.1) %, p=0.03). Figure 2 illustrates the divergence in RV BLS and RV FAC between the two groups by days 5–7 (figure 2). On linear regression analysis, the presence of a PDA on days 5–7 was independently associated with lower RV BLS (β=−2.3, p=0.02) and lower RV FAC (β=−0.06, p<0.01) when adjusting for birth weight, gestation mean airway pressure and mode of ventilation.
RV function measurement in the two groups over the first week of age. Infants with a PDA (grey line) had lower RV BLS and RV FAC on days 5–7 (*p<0.05, two-way analysis of variance with repeated measures). TAPSE, tricuspid annular plane systolic excursion; RV s′, right ventricular tissue Doppler systolic velocity; RV BLS, right ventricular tissue Doppler basal longitudinal strain; RV FAC, right ventricular fractional area change. PDA, patent ductus arteriosus.
Discussion
This prospective study sought to determine the influence of a hyperdynamic LV on measurements of RV function in preterm infants in the first week of life. We observed that the presence of a PDA and hyperdynamic LV was associated with lower values of RV strain and FAC, even after adjusting for gestation, birth weight and ventilation status. Interestingly, RV functional measurements of velocity (RV s′) and displacement (TAPSE) were not affected by a hyperdynamic LV in infants with and without a PDA at 1 week of life.
The presence of a PDA in the preterm population, particularly if persistent by the first week of age is associated with the development of pulmonary overcirculation, and systemic hypoperfusion.21 Zonnenberg and de Waal22 performed a systematic review of the literature on the impact of the PDA on left ventricular function and included 34 studies that used left heart dimensions by echocardiography to define a haemodynamically significant PDA. We know that the increasing left to right shunting across the duct becomes more significant as the pulmonary pressures fall, thereby increasing the pressure gradient between the systemic and pulmonary circulations. This enhanced pulmonary blood flow from the significant PDA results in an increase in LV preload, with left atrial and ventricular dilatation, an increase in LV EF and increased LVO, all resulting in an overall ‘hyperdynamic’ LV.23 The lower blood flow in the systemic circulation leads to a reduction in RV preload and, in theory, reduced RV output into the lungs. However, this is difficult to ascertain as measurement of RV output in the presence of a large PDA is problematic as the PDA flow interferes with the assessment of the RV outflow pulsed wave Doppler. In this study, we observed a decrease in both FAC and strain in the presence of a PDA at 5–7 days of age, suggesting that the decrease in RV function is a result of reduced RV preload.14
This is the first study to properly assess the influence of the PDA and a hyperdynamic LV on RV function in preterm infants. Both TAPSE and RV s′ assess the displacement of the tricuspid valve annulus. TAPSE measures annular movement from base to apex and RV s′ measures the velocity of that annular movement.24 ,25 Previous serial studies over the first week of life have shown increasing values for both TAPSE and RV s′.13 ,15 Our results replicated this maturational increase, despite our hypothesis that RV function would be impaired in infants with a PDA at days 5–7. The lack of difference between TAPSE and RV s′ in infants with and without a PDA by days 5–7 can be partly attributed to the increase in RV passive movement in the presence of a hyperdynamic LV without a true increase in function. TAPSE and RV s′ may be falsely elevated as they interrogate motion at a single point in the RV free wall with reference to the ultrasound transducer. Therefore, this movement can be influenced by translational motion (hyperdynamic LV pulls an abnormally functioning RV towards the apex) and this single point assessment may not fully capture true changes in RV mechanics.14
In contrast, FAC and the RV BLS were lower at days 5–7 in the PDA group, possibly providing a more accurate assessment of RV function in this setting. FAC assesses the overall change in cavity dimension, and is preload-dependent and afterload-dependent.14 RV BLS measured using tissue Doppler assesses longitudinal shortening of the RV basal segment during systole from its baseline in diastole.26 As both these measurements use the baseline of an area (RV cavity) and a segment (RV free wall) within the heart, the change in those dimensions during systole may better reflect RV function independent of the influence of the LV. Under these circumstances, both FAC and RV BLS would more accurately reflect the true RV function. As mentioned earlier, the relationship between a PDA and those parameters remained significantly independent of a significant possible confounder (ventilation status) further highlighting the strength of this relationship.
Limitations
In this study, we were unable to assess whether RV function was predominantly determined by the presence of a PDA (and its potential role in lowering systemic blood flow and venous return to the RV) or by the increased LV function measured using EF. Due to the collinearity between PDA presence and EF, examining the individual effect of those two factors is difficult in this setting. None of the infants was in receipt of inotropes, or were septic at the time of the days 5–7 scan. Therefore, it is highly likely that LV function was predominantly determined by PDA presence rather than any other factors.
The definition of a hyperdynamic LV in the premature population has not been definitively established. In the adult population, a hyperdynamic LV is defined as an EF >70%.27 Very few of our infants developed an EF of 70% or greater. However, we elected to refer to LV function in the PDA group as ‘hyperdynamic’ to reflect the significant changes observed in LV indices in infants with a PDA.
Conclusion
A hyperdynamic LV and a persistent PDA will influence the assessment of RV function in preterm infants and affect the ability of certain measures of RV performance to reflect true function. TAPSE and RV s′ appear to be falsely elevated while FAC and RV BLS may be more reliable non-invasive echocardiographic measures of RV function in the setting of hyperdynamic LV and persistent PDA at 1 week of life. Further studies are required to support these results and assess the impact of each measure as a means for assessing efficacy in patient management strategies.
References
Footnotes
This work was presented in part at the European Academy of Paediatric Societies held in Geneva, Switzerland, in October 2016.
Contributors CRB and ATJ performed the echocardiography studies and the image analysis to obtain the functional measurement. CRB wrote the first draft of the manuscript. OF provided the support with the concept and cardiology expertise. NM aided with the writing of the first draft and AELK provided senior oversight to the project and performed the statistical analysis.
Funding AELK is in receipt of an Irish Health Research Board Mother and Baby Clinical Trials Network Grant (HRB CTN 2014-10), an EU FP7/2007-2013 grant (agreement no 260777, The HIP Trial) and a Rotunda Hospital Foundation Research Grant (Reference: FoR/EQUIPMENT/101572).
Competing interests None declared.
Patient consent Obtained.
Ethics approval Rotunda Hospital Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The data from the study are only available to the principal and senior author of the study.