This was a retrospective cohort study of premature infants admitted to the University of Iowa Stead Family Children’s Hospital Neonatal Intensive Care Unit (NICU) between September 2018 and July 2020. The study was approved by the Institutional Review Board. Clinical and echocardiography data were collected from the electronic medical record (Epic 2021) and Neonatal Hemodynamics program database.
Eligibility Criteria
Neonates were eligible for the study if they were born at < 27 weeks gestation, had at least one dose of surfactant on or after postnatal day 7, and had a complete TnECHO within 24–48 hours following surfactant administration. Neonates with structural heart disease and those who died within 48h of receiving their qualifying dose of surfactant were excluded.
Clinical characteristics: Information were collected on the following: (i) neonatal demographics including gestational age, sex, birth weight, race, and ethnicity; (ii) maternal data and pregnancy history including maternal age, gravidity, parity, multiple gestation, history of preeclampsia, gestational diabetes, intrauterine fetal demise, intrauterine growth retardation (IUGR), chorioamnionitis, and antenatal steroids; (iii) delivery room course including mode of delivery, need for delivery intervention, resuscitation and complications, APGAR scores at 1, 5 and 10 min, and mode of ventilation upon admission; (iv) surfactant administration including type, dose, postnatal age (days) at timing of dose, and total number of doses given on or after postnatal day 7.
Surfactant Response
Patient ventilation characteristics including fraction of inspired oxygen (FiO2) and mean airway pressure were recorded at 0, 1, 4, 12, 24, and 48 hours after the administration of each surfactant dose. Respiratory severity score (RSS = mean airway pressure x FiO2)(12) was calculated at each of these time points. The response to each dose of surfactant was categorized according to the magnitude of change in RSS over the subsequent 48 hours as positive [≥ 15% improvement], negative [≥ 15% deterioration], or non-response [< 15% change]. For individual patients that received multiple surfactant doses, patients were classified according to the most predominant response (for example, if 4 doses were administered and 3 out of 4 showed positive response, then the patient would be classified as a surfactant responder). Finally, systolic, and diastolic blood pressures, heart rate, and hemoglobin were recorded at each of these time points over 48 hours after administration. All cardiorespiratory data were collected by data abstractors (MB, AC) who were blinded to PDA status.
Echocardiography assessment
TnECHO evaluation was completed according to a standardized imaging protocol by fully accredited neonatologists with hemodynamic expertise.(13) Studies were performed using the Vivid E90 cardiovascular ultrasound system (GE Medical Systems, Milwaukee, WI, USA) with a 12-MHz high-frequency phased-array transducer probe. Standard two-dimensional, M-mode, color Doppler, pulsed wave (PW) Doppler, and continuous wave (CW) Doppler images were obtained. Analyses of left heart volume loading, left (LV) and right (RV) ventricle systolic function, shunt physiology, pulmonary hemodynamics, and cardiac output, were performed. All echocardiography analyses were performed using a dedicated workstation (EchoPAC version BT10; GE Medical Systems, Milwaukee, WI, USA) by a single trained investigator (RG) who was blinded to the clinical information to minimize bias. Measurements were performed according to published methodology.(14) Three consecutive cardiac cycles were evaluated and averaged for each measurement.
Echocardiograms completed within 24–48 hours of each dose of surfactant were analyzed. The PDA was interrogated via suprasternal ductal view (color and 2D) and measured at the narrowest diameter via a sweep from aorta to pulmonary artery. Peak and mean gradient and shunt direction were recorded via a PW Doppler of the duct in this same view. Blood flow through the mitral valve was assessed from the apical four-chamber view, with a 2- to 3-mm sample volume PW Doppler placed immediately distal to the tips of the valve leaflets. Peak E (early diastolic phase) and A (late diastolic phase) wave were obtained.(15) Isovolumic relaxation time (IVRT) was obtained by placing the sample volume midway between left ventricular outflow and mitral inflow in the apical four-chamber view. Pulmonary vein flow was assessed in the apical four-chamber, with color flow imaging to identify relevant vein; subsequently PW Doppler was obtained by placing the sample volume placed at 1 cm depth into the right (or left) upper pulmonary vein. Left atrium (LA) to aortic ratio (LA:Ao) was obtained in the parasternal long axis via M-mode with the cursor placed at the plane of the aortic valve hinges to include the maximal diameter of the LA and in a plane perpendicular to the aortic wall at the level of the aortic valve. The leading-edge method of measurement was used. Left ventricular output (LVO, expressed in mL/min/kg) was calculated by multiplying the aortic cross-sectional area [calculated as: (aortic radius2 × π)] multiplied by velocity time integral (VTI) and heart rate and indexed to weight (in kg).(16, 17) To calculate VTI, a PW Doppler sample volume was placed at the level of the aortic valve hinge points, perpendicular to the aortic valve in the apical five-chamber view, with the angle of insonation parallel to the LV outflow tract. The area under the waveform of the aortic systolic beat was traced to obtain the VTI and the heart rate. The annulus of the aortic valve was measured, from the parasternal long axis view, between hinge points with the valve open at the end of ejection. Diastolic flow in the descending aorta, celiac artery, and middle cerebral artery (MCA) were obtained via color PW Doppler. Diastolic flow in the descending aorta was obtained via a suprasternal long axis aortic view with the sample volume placed at the level of the diaphragm. Diastolic flow in the celiac artery was obtained with a subcostal sagittal view and the diastolic flow in the MCA was obtained via a cranial axial view with the sample volume measure in line with the flow.
Group Allocation: Patients were classified as either PDA [≥ 1mm] or NO PDA [< 1mm]. We chose an arbitrary threshold of 1 mm based on prior data which showed that patients with a trivial sized PDA have a comparable echocardiography profile in terms of markers of shunt volume; in essence, they physiologically behave in a comparable manner to patients with “NO PDA”. The Iowa PDA score was also calculated to provide an objective measure of shunt severity and hemodynamic significance.(18) Elements of the PDA score included mitral valve E-wave velocity, IVRT, pulmonary vein D-wave velocity, LA:Ao, LVO, and the presence of diastolic flow reversal in the descending aorta, celiac artery, or middle cerebral artery (Supplemental Table 1). Points of 0, 1, or 2 are given based on these measured elements. The sum of points is added to the size of the PDA (in millimeters) divided by the patient’s weight to calculate the final PDA score. A score of ≥ 6 signifies a presumed hsPDA.
Approach to PDA Care
Comprehensive TnECHO screening is performed on all infants born less than 27 weeks gestation between 12–18 postnatal hours. If a small PDA or a PDA with low volume shunt is identified, the neonate is observed, and repeat TnECHO is obtained between day 5–7 or sooner if clinically indicated. If a PDA with a moderate to high volume shunt is diagnosed, the infant is treated with intravenous acetaminophen 15 mg/kg q6hr. The infant is reassessed after 3 days of acetaminophen, and if the shunt remains hemodynamically significant, acetaminophen is continued for a total of 7 days. If the PDA remains hemodynamically significant after postnatal day 7, intravenous indomethacin 0.2 mg/kg q12hr for 3 doses is administered. Infants may receive up to 2 courses of indomethacin after which they are referred for definitive closure (surgery or percutaneous device closure) if the PDA remains hemodynamically significant. Repeat TnECHO evaluation is performed within 24 hours after each course of medical therapy. In addition, the clinical team may request TnECHO evaluation at any time if there are clinical concerns.
Approach to Surfactant Use
First-intention High Frequency Jet Ventilation (LifePulse 204, Bunnell Incorporated, Salt Lake City, Utah, USA) is standard of care for all preterm infants born less than 27 weeks GA according to a previously reported approach.(19) The first dose of endotracheal surfactant [poractant 2.5 ml/kg divided in two aliquots] is given prophylactically, within 1–2 hours of birth, to all intubated infants born < 26 weeks GA. Otherwise surfactant is administered to infants with clinical and radiological features of RDS, defined as respiratory distress sufficient to require > 30% oxygen delivered by positive pressure using either nasal continuous positive airway pressure (CPAP) or an endotracheal (ET) tube and a chest radiograph with any of the following features [ground glass granular appearance, air bronchograms, loss of cardiac or diaphragmatic borders, complete whiteout]. Surfactant is administered by gentle hand bagging using peak inspiratory pressure (PIP) not to exceed 20 cm.(19) If a second dose is needed in the transitional period, poractant 1.25 mg/kg is given in a dose divided in two aliquots. Repeat surfactant dosing is considered after postnatal day 4–7 if there is clinical suspicion of surfactant inactivation or dysfunction from atelectotrauma, pulmonary hemorrhage, sepsis, pneumonia, or maturation associated “surfactant slump”.(6) Surfactant inactivation or dysfunction is considered when the infant is requiring more than 50% oxygen with a chest radiograph demonstrating diffuse atelectasis, ground glass appearance, or air bronchograms. Calfactant is typically used for premature infants who are unresponsive to two doses of poractant, or in premature infants with concern for surfactant inactivation or dysfunction. Calfactant is given in a dose of 3 ml/kg divided into two aliquots. Following surfactant administration, blood gases are monitored 15 minutes post surfactant and then at 1–3-hour intervals depending on the clinical condition of the patient. Routine chest radiographs are obtained at 1 and 4 hours post surfactant to avoid hyperinflation. Ventilator settings are adjusted to maintain arterial carbon dioxide between 40–55 mmHg in the first 7 postnatal days.
Outcomes
The primary outcome was response (positive, negative, or non-responder) to surfactant according to net change in RSS within 48 hours of surfactant administration. Secondary outcomes included PDA score, duration of exposure to PDA, and number of doses/courses of indomethacin or acetaminophen. Additionally, the frequency of neonatal morbidities [interventricular hemorrhage, necrotizing enterocolitis (Stage II or greater), pulmonary hemorrhage, bronchopulmonary dysplasia, and sepsis] and mortality was recorded.
Statistical analysis
As the relationship between PDA and late surfactant responsiveness is not known a sample size of convenience was chosen; specifically, we included all patients who received late surfactant during the defined study time-period. Univariate analysis was performed to compare demographic and outcome data between “PDA” vs “NO PDA” groups. In addition, repeat measures ANOVA was performed to evaluate time-dependent changes in clinical and echocardiography characteristics according to each individual surfactant dose (i.e., each dose was classified as a positive, negative, or non-response). Changes in RSS over time were evaluated in each sub-group. A multivariate logistic regression was built to investigate factors associated with positive (model 1) or negative (model 2) response to late surfactant. Variables with a p value < 0.1 on univariate analysis were included in the model. An odds ratio with 95% confidence interval were used and results were considered significant if p < 0.05. An ROC curve was generated for PDA score as a predictor of negative response to late surfactant.