Phosphodiesterase Inhibitor-Based Vasodilation Improves Oxygen Delivery and Clinical Outcomes Following Stage 1 Palliation

Abstract

Background: Systemic vasodilation using α-receptor blockade has been shown to decrease the incidence of postoperative cardiac arrest following stage 1 palliation (S1P), primarily when utilizing the modified Blalock-Taussig shunt. We studied the effects of a protocol in which milrinone was primarily used to lower systemic vascular resistance (SVR) following S1P using the right ventricular to pulmonary artery shunt, measuring its effects on oxygen delivery (DO2) profiles and clinical outcomes. We also correlated Fick-based assessments of DO2 with commonly used surrogate measures.

Methods and results: Neonates undergoing S1P were treated according to best clinical judgment prior to (n=32) and following (n=24) implementation of a protocol that guided operative, anesthetic, and postoperative management, particularly as it related to SVR. A majority of the subjects (n=51) received a modified right ventricular to pulmonary artery shunt. In a subset of these patients (n=21), oxygen consumption (VO2) was measured and used to calculate SVR, DO2, and oxygen debt. Neonates treated with the protocol had significantly lower SVR (P=0.02), serum lactate (P<0.001), and Sa-vO2 difference (P<0.001) and a lower incidence of CPR requiring extracorporeal membrane oxygenation (E-CPR, P=0.02) within the first 72 postoperative hours. DO2 was closely associated with SVR (r2=0.78) but correlated poorly with arterial (SaO2) and venous (SvO2) oxyhemoglobin concentrations, the Sa-vO2 difference, and blood pressure.

Conclusions: A vasodilator protocol utilizing milrinone following S1P effectively decreased SVR, improved serum lactate, and decreased postoperative cardiac arrest. DO2 correlated more closely with SVR than with Sa-vO2 difference, highlighting the importance of measuring VO2 in this population.

Clinical trial registration: URL: http://www.clinicaltrials.gov. Unique identifier: NCT02184169.

Keywords: hypoplastic left heart syndrome; hypoxia; oxygen; oxygen consumption; phosphodiesterase inhibitor.

© 2016 The Authors. Published on behalf of the American Heart Association, Inc., by Wiley Blackwell.

Figures

Figure 1

Details of the treatment protocol, including salient preoperative, intraoperative, and postoperative features of the HLHS treatment algorithm, implemented in January 2014. AVVR indicates atrioventricular valve regurgitation; CPB, cardiopulmonary bypass; CVP, central venous pressure; DO 2, systemic oxygen delivery; ECLS, extracorporeal life support; EGA, estimated gestational age; ETT, endotracheal tube; HLHS, hypoplastic left heart syndrome; MAP/MABP, mean arterial blood pressure; mBTS, modified Blalock‐Taussig shunt; MUF, modified ultrafiltration; OHT, orthotopic heart transplantation; PAB, pulmonary artery band; PD, peritoneal dialysis; PDA, patent ductus arteriosus; PGE, prostaglandin E1; PRBC, packed red blood cells; Qp/Qs, ratio of pulmonary to systemic blood flow; S1P, stage 1 palliation; SvO2, systemic venous oxyhemoglobin saturation; TR, tricuspid valve regurgitation; TXA, tranexamic acid; VO 2, oxygen consumption.

Figure 2

A, The inotrope score, which reflects the doses of dopamine, dobutamine, and epinephrine, was significantly higher prior to protocol implementation. B, The vasoactive inotrope score, which additionally reflects the use of milrinone, norepinephrine, and vasopressin, was significantly higher following protocol implementation. C, The volume of blood products administered was similar between groups, as were chest drain output (D) and urine output (E) per 12‐hour shift. Gray bars, lines, dots, and error bars reflect preimplementation data (n=32); black represents postimplementation data (n=24). Data are means, errors are SEM. ***P<0.001, **P<0.01.

Figure 3

Patients treated following implementation of the protocol exhibited lower systemic vascular resistance (P=0.02, A), lower serum lactic acid levels (P<0.001, B) and a lower incidence of CPR requiring extracorporeal life support (E‐CPR, P=0.02, C) compared to those treated prior to the algorithm. A, Because SVR was calculated only in patients with measured oxygen consumption, A includes n=10 patients in the preimplementation group (solid gray line), n=11 patients in the postimplementation group (solid black line), and a convenience sample of dTGA patients (n=6, dotted gray); data are means, error are SEM. B and C, Gray lines, bars, and error bars reflect preimplementation data (n=32); black represents postimplementation data (n=24), and stripes represent dTGA patients (n=6).

Figure 4

Oxygen consumption (A), systemic blood flow (B), systemic oxygen delivery (C), cumulative oxygen debt (D), mean arterial blood pressure (E), central venous pressure (F), rectal temperature (G), arterial (H, SaO2) and venous (I, SvO2) oxyhemoglobin saturations from patients prior to (gray) and following (black) implementation of the treatment algorithm surrounding the S1P. A through D, Oxygen consumption and its dependent variables are shown for all patients in the OxyCAHN study (n=10 pre‐ and n=11 postimplementation). E through I, n=32 pre‐ and n=24 postimplementation. A convenience sample (n=6, dotted gray) of infants undergoing ASO for dTGA/IVS is provided for comparison. Data are means, errors are SEM. P values, β, and standard errors for these variables are listed in Table 3.

Figure 5

Relationship of systemic oxygen delivery to (A) systemic vascular resistance, (B) arterial oxyhemoglobin saturation (SaO2), (C) venous oxyhemoglobin saturation (SvO2), (D) arteriovenous oxyhemoglobin saturation (Sa‐vO2) difference, and (E) mean arterial blood pressure for all patients studied in the OxyCAHN study (n=21). Data are hourly median values, black lines represent median spline through predicted values for DO 2. Adjusted r2 is shown for each model. Note that r2 values are not adjusted for repeated measures. MABP indicates mean arterial blood pressure.