Inotropes in the Hypoplastic Left Heart Syndrome: Effects in an Animal Model

Inotropes in the Hypoplastic Left Heart Syndrome: Effects in an Animal Model Christopher J. Riordan, MD, Flemming Randsbaek, MS, John H. Storey, MD, William D. Montgomery, MD, William P. Santamore, PhD, and Erie H. Austin III, MD Division of Thoracic and Cardiovascular Surgery, University of Louisville School of Medicine, Louisville, Kentucky Background. Despite substantial changes in the surgical treatment of children born with the hypoplastic left heart syndrome, overall mortality remains high. Although further improvements in outcomes appear to depend on more effective perioperative care, few experimental data exist to guide appropriate pharmacologic therapy in these infants. Because different inotropic agents may have different effects on the ratio of pulmonary to systemic flow (Qp/Q,), we hypothesize that they may not be equally effective at increasing oxygen delivery. Methods. In neonatal piglets (n = 6; 3.5 to 6.5 kg), we placed an innominate artery-to-pulmonary artery shunt, created an atrial septal defect, and then occluded right ventricular outflow. We examined the effects of a number of commonly used inotropic agents, administering high and low concentrations of dopamine (5 and 15/~g kg -~ rain-l), dobutamine (5 and 15/~g kg -~ min-~), and epinephrine (0.05 and 0.1 ~g/min). Results. Dobutamine at 15 ~g. kg -1 min -~ increased the QvJQ, ratio from 1.03 ± 0.6 at baseline to 2.52 ± 0.55 (p < 0.05) and decreased oxygen delivery from 50 ± 4.3 to 36 --- 1.7 ml/min (p < 0.1). The arterial-venous oxygen difference increased as oxygen delivery went down, going from 44% ± 1% to 48% ± 2% (p < 0.1). Epinephrine at 0.1 /zg kg -1 rain -1 decreased the QvJQs ratio from 1.23 ± 0.21 to 0.82 ± 0.08 (p < 0.05) and increased oxygen delivery from 40 ± 9.7 to 56 ± 1.7 mumin (p < 0.05). Systemic venous oxygen saturation increased from 36% --- 4.8% to 50% ± 8.6% (p < 0.05). Although dopamine decreased the Qp/Q, ratio and increased oxygen delivery, these changes were not statistically significant. Conclusions. Dopamine, dobutamine, and epinephrine all increased cardiac output but had substantially different effects on the Qp/Q, ratio and on oxygen delivery, possibly due to differential effects on systemic and pulmonary vascular resistances. This suggests that inotropic agents may not be equally beneficial in the clinical setting. Systemic venous oxygen saturation and the arteriovenous oxygen difference may help determine if a given inotrope improves oxygen delivery. (Ann Thorac Surg 1996;62:83--90) A lthough dramatic advances have been made in the surgical treatment of the hypoplastic left heart syndrome, this relatively common set of defects remains highly lethal [1]. Surgical therapy is complex and often involves multiple staged procedures [2]. This lesion remains the most common cardiac cause of neonatal death, and some operative series still report mortality rates of 40% to 45% [3]. Infants with the hypoplastic left heart syndrome often have pronounced hemodynamic instability, and this is a major contributor to the high mortality rate seen. Although improvements in perioperative care could potentially lead to better outcomes, experimental models of the hypoplastic left heart syndrome are needed to help develop better management strategies. Our group recently developed an animal model of the univentricular circulation that allows for systematic examination of therapeutic interventions [4]. We have previously used this model to describe the effects of respirator and Presented at the Forty-second Annual Meeting of the Southern Thoracic Surgical Association, San Antonio, TX, Nov 9-11, 1995. Address reprint requests to Dr Austin, Division of Thoracic and Cardiovascular Surgery, Department of Surgery, University of Louisville School of Medicine, Louisville, KY 40292. ventilator manipulations on univentricular physiology = [5]. In this study we investigated the consequences of inotropic interventions. Inotropic agents are potentially beneficial in infants with the hypoplastic left heart syndrome, but their actions have not been fully characterized. Data from experiments in adults may not accurately describe the actions of these medications in younger patients, as numerous studies have demonstrated that inotropes have altered efficacy in children [6]. In addition, the few studies that have examined the actions of inotropic agents in the young have focused on defects other than the hypoplastic left heart syndrome, and may not reflect how the drugs act in this anomaly [7]. The arrangement of the systemic and pulmonary circulations in this defect, with their resistances in parallel rather than in series, may cause inotropic medications to have altered actions. We used our porcine model of the univentricular heart to measure the hemodynamic changes that accompanied the administration of high and low concentrations of dopamine, dobutamine, and epinephrine. We also used the model to examine changes in the ratios of pulmonary to systemic flow (Qp/Qs ratio) and on oxygen delivery, as we hypothesized that the different agents may have 1996 by The Society of Thoracic Surgeons 0003-4975/961515.00 Published by Elsevier Science Inc PII S0003-4975(96)00297-4

84 RIORDAN ET AL Ann Thorac Surg INOTP, OPES IN HYPOPLASTIC LEFT HEART SYNDROME 1996;62:83-90 Table 1. Hemodynamic Effects of Inotropes" FiO2 1.0 FiO 2 0.5 Cardiac Cardiac MAP PAP Output MAP PAP Output Treatment (ram Fig) (ram Hg) (ml/min) (ram Hg) (ram Hg) (ml/min) Epinephrine Baseline 28 ± 1.26 18 + 1.26 700 ± 45 33 ± 3.1 24 ± 2.3 799 ± 217 0.05 p,g- kg -1 rain -1 31 ± 4.4 21 ± 3.5 865 ± 98 b......... 0.1/~g kg -x min -1 33 +- 2.1 20 ± 3.2 870 ± 49 b 36 -+ 2.14 26 ± 3.6 925 + 93* Dobutamine Baseline 32-1.8 20 ± 2.3 785 ± 46 31 ± 1.4 24 ± 4.2 682 -+ 77 5 ~g kg -~ min -~ 32 ± 1.8 22 ± 1.8 785 ± 68......... 15/~g kg -x min -1 30 ± 2.1 22 ± 1.9 789 + 57 37 ± 2.8 28 -+ 3.1 877 ± 98* Dopamine I~aseline 31 ± 1.4 19 _+ 2.4 650 +_ 69 33 + 2.4 27 ± 4.8 737 ± 43 5/~g kg ~1 rain -x 30 ± 2.6 23 ± 3.3 711 +_ 49......... 15/~g, kg 1. min-~ 32 ± 2.8 24 ± 4.1 738 ± 79 39 ± 2.6 30 ± 5.1 853 ± 77* a Epinephrine increases cardiac output at both levels of inspired oxygen, whereas dopamine and dobutamine only do so at the lower level of inspired oxygen, b p.~ 0.05 compared with baseline. FiO2 ~ inspired oxygen fraction; MAP = mean systemic arterial pressure; PAP ~ mean pulmonary arterial pressure. different effects on the Qp/Q~ ratio and therefore on oxygen delivery. Material and Methods Animals Neonatal pigs (n = 6; age, 1 to 2 weeks; weight, 3.5 to 6.5 kg) were used. All animals were examined intraoperatively to rule out the presence of a patent ductus arteriosus. All animals were cared for in accordance with the guidelines published by the National Institutes of Health ("Guide for the Care and Use of Laboratory Animals," NIH publication 85-23, revised 1985). AdditionaUy, all aspects of animal care were in accordance with the standards of the Institutional Animal Care and Use Committee of the University of Louisville. Procedure The details of the operative preparation were described elsewhere [4]. Briefly, each piglet was anesthetized with ketamine (30 mg/kg intravenously) and acepromazine (2 mg/kg intravenously). Continuous anesthesia was maintained with pentobarbital (5 mg kg -1 h -~ intravenously). Both groins were dissected, and systemic arterial blood pressure was monitored with a 5F micromanometertipped catheter (Millar Inc, Houston, TX) placed into the middescending thoracic aorta via the femoral artery. The other femoral artery was cannulated to measure arterial blood gases. A fluid-filled catheter was placed into the midinferior vena cava via the femoral vein to monitor central venous pressure. With the animal in supine position, a median sternob omy was performed. The pericardium was tacked up to form a weu. One thousand units of heparin was given intravenously, and a 6-ram reinforced Gore-Tex (W. U Gore & Assoc, Flagstaff, AZ) graft was placed from the innominate artery to the common trunk of the pulmonary artery. This was anastomosed end-to-end to the innominate artery and end-to-side to the pulmonary artery. After the graft was completed, transit time flow probes (Triton, Inc, San Diego, CA) were placed on the proximal aorta, to measure left ventricular outflow, and around the innominate artery just proximal to the graft, to measure flow through the graft, ie, pulmonary artery blood flow. A 5F micromanometer tip catheter (Millar, Inc) was passed from the infundibular portion of the right ventricle into the common trunk of the pulmonary artery to continuously monitor pulmonary artery pressure. A 4F Rashkind septostomy catheter (Bard Inc, Billerica, MA) was then passed transvenously via the remaining femoral vein to create a nonrestrictive atrial septal defect. The Rashkind catheter was then advanced into the right ventricle. The balloon was inflated and the catheter was slowly withdrawn, thereby snaring the chordae tendineae of the tricuspid valve. By repeatedly tearing the chordae tendineae, we rendered the tricuspid valve incompetent. By occluding the right ventricular outflow tract, we completed the univentricular circuit. All systemic venous return was routed across the atrial septal defect into the left atrium. Pulmonary flow was maintained by that portion of the left ventricular outflow that exited the innominate artery and traversed the 6-mm Gore-Tex graft to the pulmonary artery. Right ventricular distention was prevented by having rendered the tricuspid valve incompetent. All pressure and flow measurements were displayed on a Hewlett-Packard monitor and simultaneously recorded on a Gould recorder (model TA-11; Gould, Inc, Valley View, OH). Arterial blood gases were measured on a blood and electrolyte analyzer (Nova Biomedical, Norwood, MA). Venous oxygen saturations were measured using an oximetric catheter (Opticath; Abbott Labs,

Ann Thorac Surg RIORDAN ET AL 85 1996;62:83-90 INOTROPES IN HYPOPLASTIC LEFT HEART SYNDROME North Chicago, IL) placed in the midabdominal inferior vena cava and were continuously recorded on a Abbott Labs recorder. Protocol Once the univentricular preparation was completed and stable conditions had been present for at least 10 minutes, baseline measurements were obtained. Baseline ventilator settings consisted of inspired oxygen fraction (FiO2) of 1.0 and positive end-expiratory pressure of 0 cm H20. Individual inotropic agents were then administered. The drugs used were epinephrine at concentrations of 0.05 and 0.1/~g" kg- min -1, dobutamine at 5 and 15/~g kg -1. min -1, and dopamine at 5 and 15/~g" kg -1" min -a. The order of drug administration varied between experiments. Drugs were administered at an FiO 2 of 1.0 and 0.5. Each agent was given for at least 10 minutes, until equilibrium conditions were present. Measurements were then taken. Before administering another drug, we obtained repeat baseline measurements. These readings were taken after administration of the initial drug had been stopped for at least 15 minutes and an equilibrium condition had again been reached. Data Analysis and Statistics Pulmonary flow and aortic flow were measured directly. Systemic flow was calculated as aortic flow minus pulmonary flow, as all pulmonary flow was derived from the innominate artery via the innominate artery-to-pulmonary artery graft. The Qp/Qs ratio was then determined by dividing pulmonary flow by the derived systemic flow. Pulmonary vascular resistance (PVR) was determined as (MPA - CVP)/Qp, where MPA = mean pulmonary artery pressure, CVP = central venous pressure, and Qp = pulmonary flow. Central venous pressure was used in these calculations because an earlier study using this model demonstrated CVP to be equal to left atrial pressure, given the nonrestrictive atrial septal defect incorporated in the preparation. Systemic vascular resistance (SVR) was defined as (MAP - CVP)/Q~ where MAP = mean arterial pressure and O~ -- systemic flow (calculated from directly measured values). Oxygen delivery was calculated as systemic arterial content times systemic arterial flow, where systemic arterial oxygen content = (138 hemoglobin systemic medal oxygen saturation) + (0.0031 partial pressure of arterial oxygen). 2.~ 1.6 I~ 1.2 ~ 0.8 0.4 0 3 2.5 ~ 2 ~ 1.5 1 ~ 0.5 0 Epinephrine vs Qp/Qs FiO2 100 % FiO2 50 % Base 0.05 0.I Base Epinephrine (fig/rain) Dobutamine vs Qp/Qs FiO2 100 % * FiO2 500/0 Dobutamine (pg/kg/min) Dopamine vs Qp/Qs, I~i~9 1 ~ 1~i~9 ~fl o/~ Bme S 1S Base ls Dolmmine (pg/kg/min) Fig 1. Effects of inotropes on the pulmonary-to-systemic flow (Qp/ Q~) ratio. Epinephrine decreases Qr/Q~, whereas dobutamine increases it. Dopamine does not lead to significant changes, although its actions appear similar to those of epinephrine. (FiO2 = inspired oxygen fraction; *p < 0.05 compared with baseline.) Results were analyzed using nonlinear regression analysis and goodness of fit test using a computerized statistical package (Excel; Microsoft Corp, Redwood, WA). One-way analysis of variance was used to analyze differences between means, with significance set at p less than 0.05. Results Hemodynamics Table 1 shows the hemodynamic effects of each agent. Epinephrine tended to increase mean systemic arterial pressure, but not to a statistically significant degree. No agent increased pulmonary artery pressure to a significant degree, although dopamine tended to increase it. All three agents increased cardiac output, although only epinephrine induced statistically significant changes at an FiO 2 of 1.0, whereas all three agents did at an FiO 2 of 0.5. O.l

86 RIORDAN ET AL Ann Thorac Surg INOTROPES IN HYPOPLASTIC LEFT HEART SYNDROME 1996;62:83-90 Epinephrine vs Resistance 800 - FiO2 100% FiO2 50% 7~oo ~ 200 Base 0.05 0.1 Base 0.1 Epinephrine (pg/min) Dobutamine vs Resistance ~ 1000 T FiO2100% ~ FIO250% " I, i i aj ~ 600 400 ~soo. 0 np~ Dobutamine (pg/kg/min) Dopamine vs Resistance FiO2 1 O0 FiO2 50 % ~1~0 * ~. Dopamine (ug/kg/min) Fig 2. Effects of inotropes on resistances. Dobutamine at high doses increased systemic vascular resistance, whereas dopamine increased pulmonary vascular resistance. (FiO2 = inspired oxygen fraction; *p < 0.05 compared with baseline.) Pulmonary-to-Systemic Flow Ratio Figure 1 shows the influence of each agent on the Qp/Qs ratio. The upper panel demonstrates that epinephrine decreased Q1,/Qs from a baseline of 1.23 --- 0.21 to 0.82 + 0.08 at a dose of 0.1 /~g kg -1 min -1. Epinephrine decreased Qp/Qs at both high and low FiO2s. The decrease in Qp/Qs was significant at an FiO 2 of 1.0 (p < 0.05). In contrast to epinephrine, dobutamine increased the Qp/Qs ratio (Fig 1, middle panel). At an FiO 2 of 1.0, the Qp/Qs ratio increased from 1.03 + 0.16 at baseline to 2.52 + 0.55 at 15 /~g kg -~ min 1 of dobutamine. This increase was statistically significant (p < 0.05). Dobutamine also increased Qp/Qs at an FiO 2 of 0.5 (p < 0.05). The bottom panel of Figure 1 shows that although dopamine decreased Qp/Qs at both levels of FiO2, the decrease failed to reach statistical significance. Resistance Figure 2 shows the relationship of the various inotropic agents and SVR and PVR. Epinephrine produced no significant changes in SVR (Fig 2, upper panel). Pulmonary vascular resistance increased, but not to a statistically significant level. This was true at both levels of FiO 2. Dobutamine administration had no significant effect on PVR (Fig 2, middle panel). There was a statistically significant increase in SVR (p < 0.1) at an FiO 2 of 1.0. No significant changes in SVR were seen at an FiO2 of 0.5. There was no significant change in SVR due to dopamine (Fig 2, lower panel), but PVR was increased significantly at both levels of FiO a (p < 0.1). Oxygen Delivery Figure 3 shows the effects of the inotrope-induced changes in the Qp/Qs ratio on oxygen delivery. The upper panel shows that epinephrine increased oxygen delivery at an FiO 2 of 1.0 (p < 0.05). Epinephrine also increased oxygen delivery at an FiO 2 of 0.5, but this did not reach statistical significance. Once again, dobutamine had effects opposite to those of epinephrine: dobutamine decreased oxygen delivery at an FiO 2 of 1.0 (Fig 3, middle panel). Although this decrease was statistically significant, no significant changes were seen at an FiO 2 of 0.5. Dopamine had no 6O ~4o ~e o 6O ~.~ "H 40 o 6O ~ 4O,.. 20 ~ o Epinephrine vs Oxygen Delivery Fi.O2 100% FiO2 50% Base 0.05 0.1 Base 0.1 Epinephrine (pg/kg/min) Dobutamine vs Oxygen Delivery Base 5 15 Base! 5 Dobutamine (pg/kg/min) Dopamine vs Oxygen Delivery FiO2 100% FiO2 50% Dopsmine (pl~g/min) Fig 3. Effects on oxygen delivery. Epinephrine increased oxygen delivery at an inspired oxygen fraction (FiO2) of 1.0 but did not reach significance at an Fi02 of 0.5. Dobutamine decreased oxygen delivery at an Fi02 of 1.0. Dopamine had no significant effects. (*p < 0.05 compared with baseline.)

Arm Thorac Surg RIORDAN ET AL 87 1996;62:83-90 INOTROPES IN HYPOPLASTIC LEFT HEART SYNDROME 0.70 "~ Epinephrine vs PVR/SVR. 0.60 0.50 ~ 0.40 0.30 0.20 0.10 0.00 Base 0.05 0.1 Base 0.1 Epinephrine (mc~kg/min) 0.70 Dobutamine vs. PVR/SVR 0.6... t FiO2 10% IJIII FiO2 50% 0.10 0.00 DobUtamine (mcg/kg/min) 1.00 Dopamine vs PVI~SVR 0 Iat /i Fi02100% Fi02 50% ~ 0.(~ ~ 0.411 0.20 0.00 Dopamine (mcg/kg/m/n) Fig 4. Effects of inotropes on the pulmonary vascular resistance-tosystemic vascular resistance (PVRISVR) ratio. Epinephrine increases the PVR/SVR ratio, whereas dobutamine decreases it. Dopamine does not alter the ratio. The changes in PVR/SVR are opposite to those in the pulmonary-to-systemic flow ratio: an increase in the PVR/SVR ratio leads to a decrease in the flow ratio. This is a reflection of the parallel arrangement of the circulations. (FiO2 = inspired oxygen fraction.) significant effects on oxygen delivery at either level of FiO 2. Figure 4 plots the changes in the ratio of PVR to SVR as a function of drug administration. This plot demonstrates that there were significant increases in the PVR/SVR ratio when dopamine and epinephrine were given (p < 0.05), whereas no changes were seen with dobutamine. Oxygen Saturations Table 2 shows the influence of the various medications on systemic arterial oxygen saturations, systemic venous oxygen saturation (SvO2), and the arterial-venous oxygen (AVO2) difference. The SaO 2 did not alter significantly with the administration of any agents, despite the large changes in the Qp/Qs ratio that occurred with these drugs. Epinephrine at an FiO 2 of 1.0 increased SvO 2 significantly (p < 0.05). Dobutamine and dopamine had no significant effects on SvO 2. At an FiO2 of 1.0, epinephrine lowered the AVO 2 difference significantly (p < 0.05). At an FiO 2 of 0.5, epinephrine led to no significant changes in the AVO 2 difference. Increasing doses of dobutamine led to significant increases in the AVO 2 difference, whereas dopamine did not alter the AVO 2 difference. Comment The hypoplastic left heart syndrome encompasses a relatively common and potentially highly lethal set of congenital defects. In untreated infants, mortality approaches 100%, with 15% one-day, 70% one-week, and 91% one-month mortalities reported [8]. Although dramatic technical advances have occurred in the operative treatment of these defects, overall survival remains suboptimal, with reported mortality rates ranging from 40% to 45% [9]. In contrast to infants with other univentricular defects, such as tricuspid atresia, children born with the hypoplastic left heart syndrome can have particularly volatile hemodynamics. This lability is present before and immediately after the initial operative intervention (the Norwood procedure). Good outcomes appear to be related to the ability to achieve a balance between the pulmonary and systemic circulations, which maintain parallel resistances after first-stage operative palliation [10]. A number of groups have reported using alterations in respiratory mechanics, including supplemental carbon dioxide and nitrogen, to stabilize hemodynamics and balance the Qp/Qs ratio [11]. Until very recently, new therapeutics such as these have been developed empirically, through "trial and error." Our group and a number of others have completed new animal models of a univentricular heart that raise the possibility of developing and examining perioperative therapies in an experimental framework. Nor- Table 2. Inotropes and Saturations at an Inspired Oxygen Fraction of 1.0 ~ Treatment SaO 2 SvO 2 AvO 2 Epinephrine Baseline 93-0.6 36-4.6 57 _+ 2.3 0.05/~g kg -1 min -1 92 _+ 2.4 47 + 8.3 b 45 + 2.1 b 0.1 p,g kg -1 min -1 91 -+ 2.4 50 + 8.5 b 41 _+ 3.4 b Dobutamine Baseline 94 + 1.1 46 ± 4.2 48 -+ 1.1 5/~g kg -~ min -~ 94 _+ 0.7 49 + 5.2 45 -+ 2.6 15/~g kg -~ min -1 95 ± 0.8 44 ± 5.1 51 ± 0.9 b Dopamine Baseline 94 ± 1.1 47 _+ 2.4 47 ~ 2.5 5/~g kg -~ min -1 94 ± 0.9 48 ± 4.6 46 ~ 3.3 15/~g kg -1 min -~ 93 _+ 1.1 48 ± 7.3 45 ~ 3.4 a Systemic arterial oxygen saturation remained constant throughout the protocol and over a wide range of pulmonary-to-systemic flow ratios. Epinephrine increased systemic venous oxygen saturations. Epinephrine decreased the AVO 2 difference, whereas dobutamine increased the AVO 2 difference, b p < 0.05 compared with baseline. AVO 2 = arteriovenous oxygen difference; SaO 2 = arterial oxygen saturation; SvO 2 = venous oxygen saturation.

88 RIORDAN ET AL Ann Thorac Surg INOTROPES IN HYPOPLASTIC LEFT HEART SYNDROME 1996;62:83-90 wood's group [12] used a porcine model, constructed with cardiopulmonary bypass and circulatory arrest, to examine the effects of adding supplemental carbon dioxide to the breathing circuit. Hanley and associates [13] reported using a model constructed by performing a Damus-Stansel-Kaye procedure and placing a shunt in fetal lambs to examine the consequences of addition of nitric oxide, carbon dioxide, and lowered oxygen tensions on the univentricular circuit. We have recently used our model to characterize the consequences of lowering FiO2, adding positive end-expiratory pressure, and adding supplemental carbon dioxide. Although pharmacologic agents, and in particular inotropes, constitute a major part of the therapy used to stabilize infants with the hypoplastic left heart syndrome, there are few experimental data to guide the application of these drugs. Experimental data may be useful as the actions of these drugs on infants cannot be predicted from data derived from studies performed on adults [6]. Studies that have specifically examined the effects of inotropes in adults and children show a number of differences. Driscoll and associates [14] have demonstrated decreased myocardial sensitivity to dopamine in both human infants and young animal models. Zaritsky and Chernow [15] have shown how this decreased sensitivity translates into a need for higher doses of both dopamine and other inotropes to achieve the same effects in children that are seen in adults. These authors suggest that these findings may reflect incomplete sympathetic innervation and reduced norepinephrine stores in the young. As development continues and innervation increases, responses closer to those found in the adult are seen. Recognizing the potential differences in the actions of drugs due to developmental changes in pharmacology, a number of studies have been performed to examine the effects of inotropes in younger patients, but their results may not be directly applicable to patients with the hypoplastic left heart syndrome. The findings in these studies are often conflicting, and are derived from defects that do not reproduce the unusual physiology of the univentricular circuit. Examining a number of infants at the time of postoperative catheterization, Stephenson and colleagues [7] found that dopamine decreased SVR by a mean of 8% and increased PVR by an average of 11%. The defects encountered in this group included atrial septal defects, tetralogy of Fallot, and pulmonary hypertension. In contrast, a Japanese group [16] examining a number of infants with large ventricular septal defects before repair showed an increase in both PVR and SVR with dopamine infusion. Animal studies have failed to yield any more consistent data. Abdul-Rasool and associates [17] found in a dog model that dopamine increased systemic vascular resistance, whereas dobutamine decreased systemic and pulmonary vascular resistance. Toorup and colleagues [18], in contrast, showed that, in lambs with a left-to-right shunt placed at the atrial level, dopamine decreased SVR, whereas dobutamine decreased both SVR and PVR. Driscoll and co-workers [19], in a study using young dogs, failed to demonstrate any change in SVR with the administration of dobutamine. In addition to yielding conflicting results, these studies have focused on defects, such as atrial or ventricular septal defects, that have different hemodynamics than those encountered in the hypoplastic left heart syndrome. This circulation is distinct in having the pulmonary and systemic resistances in parallel rather than in series. As a result, the ratio of flow through the two systems, the Qp/Qs ratio, is a direct function of the ratio of PVR to SVR. Changes in SVR or PVR induced by pharmacologic agents can have potentially large effects in this circuit, ones that may be quite distinct from those seen in other cardiac defects. Because of these considerations, we thought it was important to use our animal model to examine the actions of inotropic agents on a univentricular heart. The drugs that we chose to examine, epinephrine, dobutamine, and dopamine, have different combinations of actions on al, /31, and /32 receptors. As a result, we thought they might influence SVR, PVR, and the PVR/ SVR ratio differently, leading to varied effects on the Qp/Qs ratio. We found that the agents we examined had dramatic and opposite effects on the Qp/Qs ratio, and led to contradictory effects on oxygen delivery. Although dopamine, epinephrine, and dobutamine all increased total cardiac output, only epinephrine significantly increased oxygen delivery. Dopamine had no significant impact on oxygen delivery, whereas dobutamine decreased it. We have shown previously, in both a theoretical and an animal model that oxygen delivery in the hypoplastic left heart syndrome is a direct function of the Qp/Qs ratio [20, 21]. There appears to be a fairly narrow range of Qp/Qs that leads to maximal oxygen delivery; Qp/Qs ratios either above or below this range are associated with decreased levels of oxygen delivery. In the case of inotropic medications, the ability of a drug to increase oxygen delivery seems to be determined not only by its ability to increase cardiac output but also by its effect on the Qp/Qs ratio. If a medication moves the Qp/Qs ratio towards its optimum, then oxygen delivery increases. Conversely, if a drug moves the Qp/Qs ratio away from the optimum, then oxygen delivery decreases, even though that drug may be increasing total cardiac output. This is better demonstrated when comparing the effects of epinephrine and dobutamine in our model. Dobutamine increased cardiac output but decreased oxygen delivery because it moved the Qp/Qs ratio away from its optimum, increasing it from 1.03 0.16 to 2.52. 0.55. Conversely, epinephrine was able to increase oxygen delivery because it not only increased cardiac output but also moved the Qp/Qs ratio closer to an optimum, changing it from 1.23 --- 0.21 to 0.82 --- 0.08. We have previously found, both in our theoretical and animal studies, that an optimum Qp/Qs ratio is seen at a value slightly less than 1. As seen in Table 2, the changes that occurred in oxygen delivery in our model were accompanied by changes in

Ann Thorac Surg RIORDAN ET AL 89 1996;62:83-90 INOTROPES IN HYPOPLASTIC LEFT HEART SYNDROME SvO 2 and AVO 2 difference. As epinephrine increased oxygen delivery, SvO 2 increased while the AVO 2 difference decreased. When dobutamine decreased oxygen delivery, the AVO 2 difference increased. This suggests that determining SvO 2 and the AVO 2 difference may be a useful way to assess the effect of a given drug on oxygen delivery. Our model demonstrates one manner in which the effects of the drugs can be quantitated in the clinical setting. We have previously demonstrated that SvO 2 functions as a good marker for the Qp/Qs ratio in an animal model [21]. In that study we found that SvO2 reaches an optimum at approximately the same range of Qp/Qs values that oxygen delivery does; the plots of SvO 2 versus the Qp/Qs ratio and oxygen delivery versus the Qp/Q~ ratio are very similar. The present study also demonstrated that SvO 2 and the AVO 2 difference were good markers for oxygen delivery. In a clinical setting, SvO 2 should be able to indicate whether an inotrope is successful in increasing oxygen delivery. Failure of an inotropic agent to increase SvO 2 and decrease the AVO 2 difference would indicate that the drug had not increased oxygen delivery, most likely because it had moved the Qp/Qs ratio away from an optimum. Limitations Although our model was based on a single left ventricle performing cardiac work, rather than on a right ventricle, it contained a circulation dependent on a systemic-topulmonary shunt. This function closely approximates the arrangement both in the preoperative hypoplastic left heart syndrome and in the heart after first-stage palliation. This circuit is of a very specific type, making the results of this study most easily extrapolated to the defects that constitute the hypoplastic left heart syndrome. The physiology of the univentricular heart defects, with parallel SVR and PVR, is relatively consistent, however, so that some insight into the full spectrum of these defects may be derived from this study. This protocol was performed at an FiO 2 of 1.0 and 0.5. In clinical practice infants are typically maintained at a substantially lower FiO2, typically 0.21 or less. In our model the Qp/Q~ ratio was substantially less than that seen in human infants for an equivalent FiO2, despite the use of a 6-mm graft in the preparation. As we encountered minimal pressure drops across the graft, the lower Qp/Qs ratio probably reflects differences in the pulmonary vasculature between species, or may be due to the fact that these animals do not undergo cardiopulmonary bypass. The lower Qp/Qs ratio in the preparation did necessitate the use of higher FiO 2 to obtain Qp/Qs ratios in the range seen in humans. This means that this study does not allow the direct determination of an optimal value of FiO 2 to use in humans. The overall trends and actions of the agents seen in this study should still be useful and valid, however. Although our model suggests that there are substantial differences between inotropic agents regarding their effects on the Qp/Q~ ratio and oxygen delivery, it may be that at least some of the data reflect species differences. There is little research to suggest that the actions of the medications that we examined are substantially different in the porcine model and in humans, although little is known about the effect of these medications in a univentricular system. These concerns notwithstanding, our data at least raise the possibility that different inotropic agents may have quite substantial differences in their effects in the neonate with the hypoplastic left heart syndrome. Further studies characterizing these medications in the human are needed. Conclusions In an animal model of the hypoplastic left heart syndrome, epinephrine decreased the Qp/Qs ratio and increased oxygen delivery, whereas dobutamine increased the Qp/Qs ratio and decreased oxygen delivery. Dopamine had no statistically significant effects on either the Qp/Qs ratio or oxygen delivery. Systemic venous oxygen saturation and the AVO 2 difference were useful in determining if an inotropic agent increased or decreased oxygen delivery. These findings suggest that inotropic agents may have substantial and opposite effects on oxygen delivery in the human infant with the hypoplastic left heart syndrome. Further examination of this hypothesis is necessary. Measuring SvO 2 and AVO 2 differences may be a useful way to determine drug effects in the clinical setting. This study was supported in part by grants from the Alliant Community Trust Fund, the Louisville Institute for Heart and Lung Disease-Jewish Hospital and the American Heart Association, Kentucky Affiliate. Doctor Riordan is a Cardiothoracic Research Fellow, funded by Alliant Community Trust Fund. References 1. Norwood WI Jr. Hypoplastic left heart syndrome. In: Sabiston DC, ed. Textbook of surgery. Philadelphia: Saunders, 1991:1930-5. 2. Starnes VA, Griffen ML, Pitlick D, Ivens K, Shumway NE. Current approach to hypoplastic left heart syndrome: palliation, transplantation, or both? J Thorac Cardiovasc Surg 1992;104:189-93. 3. Meliones JN, Snider AR, Bove EL, Rosenthal AR, Rosen DA. Longitudinal results after first stage palliation for hypoplastic left heart syndrome. Circulation 1990;82(Suppl 4):151-6. 4. Randsbaek F, Riordan CJ, Montgomery WD, Storey JH, Santamore WP, Austin EH. An animal model of the univentricular heart and single ventricle physiology. J Invest Surg (in press). 5. Riordan CJ, Randsbaek F, Storey JH, Montgomery WD, Santamore WP, Austin EH. Respiratory interventions in an animal model of the univentricular heart: effects on oxygen delivery and means of monitoring outcomes. Surg Forum 1995;46:274-6. 6. Steinberg C, Notterman DA. Pharmacokinetics of cardiovascular drugs in children. Clin Pharmacokinet 1994;27:345-67. 7. Stephenson LW, Edmunds LH, Raphaely R, Morrison DF, Hoffmann WS, Rabis LJ. Effects of nitroprusside and dopamine on the pulmonary artery vasculature in children after cardiac surgery. Circulation 1979;60(Suppl 1):104-10. 8. Fyler DC. Report of the New England Regional Infant Cardiac Program. Pediatrics 1980;65(Suppl):377-461.

90 RIORDAN ET AL Ann Thorac Surg INOTROPES IN HYPOPLASTIC LEFT HEART SYNDROME 1996;62:83-90 9. Weldner PW, Meyer TC, Gleason MM, et al. The Norwood operation and subsequent Fontan operation in infants with complex congenital heart disease. J Thorac Cardiovasc Surg 1995;109:654-62, 10. Jobes DR, Nicolson SC, Steven JM, Miller M, Jacobs ML, Norwood WI. Carbon dioxide prevents pulmonary overcirculation in hypoplastic left heart syndrome. Ann Thorac Surg 1992;54:150-1. 11. Morray JP, Lynn AM, Mansfield PB. Effect of ph and pco 2 on pulmonary and systemic hemodynamics after surgery in children with congenital heart disease. J Pediatr 1988;113: 474-9. 12. Mora GA, Pizarro C, Jacobs ML, Norwood WI. Experimental model of single ventricle: influence of carbon dioxide on pulmonary vascular dynamics. Circulation 1994;90(Suppl 2):43-6. 13. Reddy VM, Liddicoat JR, Fireman FR, Chang R, Klein JR, Hanley FL. Fetal model of single ventricle physiology: hemodynamic effects of nitric oxide, oxygen, and carbon dioxide in early postnatal life [Abstract]. Circulation 1994; 90(Suppl 1):312A. 14. Driscoll DJ, Gillette PL, Duff DF, et al. Hemodynamic effects of dobutamine in children. Am J Cardiol 1979;43:581-5. 15. Zaritsky A, Chernow B. Use of catecholamines in pediatrics. J Pediatr 1984;105:341-50. 16. Nakazama M, Takahashi Y, Aiba S, Okuda H, Ohta H, Takao A. Acute hemodynamic effects of dopamine, dobutamine, and isoproterenol in congested infants or young children with large ventricular septal defects. Jpn Circ J 1987;51: 1010-5. 17. Abdul-Rasool HM, Chamberlain JH, Swan PC, Ward DS. Cardiorespiratory and metabolic effects of dopamine and dobutamine infusions in dogs. Crit Care Med 1987;15: 1044-9. 18. Toorup GP, Hardjowoop R, Dlinghaus M, et al. Comparative circulatory effects of isoproteronol, dopamine, and dobutamine in conscious lambs with and without aortopulmonary left to right shunts. Circulation 1987;75:1222-8. 19. Driscoll DJ, Gillette PL, Lewis RM, Hartley CJ, Schwartz A. Comparative hemodynamic effects of isoproterenol, dopamine, and dobutamine in the newborn dog. Pediatr Res 1979;13:1006-9. 20. Barnea O, Austin EH, Richman B, Santamore WP. Balancing the circulation: theoretic optimization of pulmonary/systemic flow ratio in hypoplastic left heart syndrome. J Am Coil Cardiol 1994;24:1376-81. 21. Riordan CJ, Randsbaek F, Storey JH, Montgomery WD, Santamore WP, Austin EH. Balancing pulmonary and systemic arterial flows in parallel circulations: the value of monitoring systemic venous oxygen saturations. Cardiol Young (in press). DISCUSSION DR SAFUH ATTAR (Baltimore, MD): Doctor Riordan, have you used this information in your clinical applications? DR RIORDAN: Yes. We have been very impressed with the usefulness of the systemic venous oxygen saturation catheters. We have been placing a 4F oximetric catheter (Abbott Biomedical North Chicago, IL) in our recent series of children with hypoplastic left heart syndrome, and we have found it adds useful information that dramatically changes our clinical results. We have also stopped using dobutamine in these children and have been using epinephrine instead. DR ATTAR: Did you have to alter the dosage pertinent to your studies? DR RIORDAN: Rather than altering specific dosages, we have found that the oximetric catheters allow us to titrate therapy. When using inotropic agents, or adding supplemental nitrogen or carbon dioxide in the ventilator circuit, we adjust dosages to improve SVO2. This has been valuable enough to us that we have started to place the catheters preoperatively to stabilize our patients, and have been able to improve a number of patients substantially this way. DR CONSTANTINE MAVROUDIS (Chicago, IL): This is a very well presented and thoughtful paper. We really need to have more studies like this comparing hemodynamic parameters under varying conditions of hypothermia, drug therapy, and systemic-to-pulmonary artery communications. Systemic-topulmonary artery shunts performed for complex heart disease and pulmonary stenosis/atresia usually have a good outcome. However, when cardiopulmonary bypass is necessary for other intracardiac problems, the shunt may not perform as well due to the increase in pulmonary artery pressure. Have you done any experiments to address this very difficult clinical problem? I enjoyed this presentation very much. I think you have another 10 years of experimenting with this model and I look forward to the results. DR RIORDAN: Thank you, Dr Mavroudis, for the kind comments. We have not performed cardiopulmonary bypass yet in these animals. This was technically not an easy model to develop and was designed not to require cardiopulmonary bypass to make a more reproducible model. Of the two other animal models of a univentricular heart, Dr Norwood's group has one that is constructed on cardiopulmonary bypass. We are planning to use our model for more complex situations in the future and will be adding cardiopulmonary bypass at that point. As you mentioned, there are a number of interventions we wish to do before adding bypass.