Materials and Methods

Animals—Eight neonatal foals (5 males, 3 females) between 1 and 5 days of age and weighing 38.5 to 59.0 kg at the beginning of the study were used. All foals had an uneventful birth, and physical examination findings as well as CBC determination and serum biochemical analysis were used before inclusion in the study to confirm their health status. Adequate transfer of passive immunity was confirmed prior to initiation of the study by measurement of plasma IgG concentration with a commercial immunoassay.^a The Institutional Animal Care and Use Committee of the University of Florida approved the study.

Anesthesia and instrumentation—Initial instrumentation was performed the day before or the day of the experiment with manual restraint. Instrumentation included a 7-F, 30-cm, double lumens catheter^b placed in the jugular vein into the right atrium for measurement of CVP and collection of central venous blood, and a 14-F nasogastric tonometry catheter^c placed into the stomach for measurement of PgCO₂.^d Radiography was used to verify correct placement of the jugular and nasogastric catheters. Anesthesia was induced by IV administration of diazepam^e (0.05 mg/kg) followed by ketamine^f (2.0 mg/kg), and the foals were positioned in left lateral recumbency. An orotracheal tube was placed for administration of inhalant anesthesia. Isoflurane^g in O₂ was administered by use of a rebreathing circuit system with an O₂flow of 2 to 3 L/min and an end-tidal isoflurane concentration of 2.0% to 2.5%. For monitoring, an infrared gas analyzer^h was calibrated before each experiment with the recommended standardized calibration gas mixture.ⁱ For each foal, end-tidal isoflurane concentration was kept constant throughout each study. End-tidal CO₂ concentrations were monitored^h and maintained between 40 and 45 mm Hg for the first 30 minutes of anesthesia by means of mechanical ventilation at a rate of 8 to 10 breaths/min and a tidal volume of 10 to 15 mL/kg. Subsequent end-tidal CO₂ concentrations during the experiment were recorded without adjusting ventilator settings to avoid interference with PgCO₂ and PaCO₂ during the administration of vasoactive drugs. Measurements of SAP, DAP, and MAP were obtained from a 20-gauge, 4.78-cm catheter placed in the right metatarsal artery and an electronic pressure transducer positioned and zeroed at the level of the sternal manubrium. Arterial blood pressures, ECG, and heart rate were monitored^h continuously and displayed throughout the experiment. A balanced electrolyte solution with 5% dextrose was administered IV at a rate of 5 mL/kg/h during anesthesia to prevent hypoglycemia and maintain normovolemia.^j Body temperature was monitored electronically and maintained between 37.5° and 39°C by use of a blanket.

Measurements of CO were determined by LiDCO. A LiDCO computer^k was used as previously described.⁴ The lithium chloride sensor^l was attached to the side port of a 3-way valve connected to the arterial catheter. Extension tubing attached the sensor to a blood collection bag, and blood passed through a peristaltic pump that produced a blood flow of 4 mL/min across the sensor. The hemoglobin concentration and serum sodium concentration required by the LiDCO computer were determined by use of a blood gas analyzer,^m immediately prior to obtaining CO measurement. The dose of lithium chlorideⁿ (0.16 to 0.28 mmol) was parked into an extension set attached to the jugular vein catheter and flushed with 12 mL of heparinized saline (0.9% NaCl) solution 8 seconds after starting the injection phase on the LiDCO computer. Measurements of CO were obtained in duplicate to verify < 20% variation between the 2 measurements, and the mean was used for data analysis.

Experimental design—Each foal received 3 vasoactive drugs by use of an orthogonal factorial design. Each drug was administered at 2 rates (LIR and HIR). At least 24 hours elapsed between administrations of each drug. Dobutamine^o was administered at rates of 4 and 8 μg/kg/min, norepinephrine^p at 0.3 and 1.0 μg/kg/min, and vasopressin^q at 0.3 and 1.0 mU/kg/min. Drugs were always administered at the LIR first. All solutions were prepared by dilution in a 5% dextrose solution to a concentration that resulted in the same volume of infusion (dobutamine, 2 mg/mL; norepinephrine, 0.2 mg/mL; and, vasopressin, 0.2 U/mL) and were administered with a syringe infusion pump.^r

On each of the 3 days, foals were anesthetized and instrumented while a deep level of anesthesia that produced a steady hypotensive hemodynamic state was reached (approx 15 minutes). Foals were then maintained at this steady hypotensive hemodynamic state for an additional period of 15 minutes prior to baseline data acquisition. Measurements of CO; SAP, DAP, and MAP; heart rate; CVP; arterial hemoglobin saturation and ScvO₂; arterial blood gas tensions and electrolyte concentrations; central venous blood gas tensions and electrolyte concentrations; arterial and central venous lactate concentrations; end-tidal isoflurane and end-tidal CO₂ concentrations; and PgCO₂ were recorded. From these variables, CI, SV, SVI, SVR, arterial O₂ content, CcvO₂, VO₂, DO₂, ERO₂, and ΔCO₂ were calculated (Appendix).

Immediately after baseline measurements, the vasoactive drug was administered at an LIR for a period of 15 minutes and all cardiovascular variables were recorded a second time. The vasoactive drug was then administered at an HIR for 15 minutes, after which all the cardiorespiratory variables were recorded a third time.

Recovery from anesthesia—All instrumentation was removed once the experiment was concluded each day, with the exception of jugular and nasogastric catheters, which were left in place with a protective bandage applied to the neck between treatment days to avoid repetitive placement. Foals were assisted throughout the recovery, returned to the stall once fully awake, and observed for normal nursing and behavior.

Data analysis—Data are expressed as mean ± SD values. A 2-way ANOVA for repeated measurements was used to determine the effects of drug type (dobutamine, norepinephrine, vasopressin), dose (baseline, LIR, HIR), and the interaction between drug type and dose on each measured and calculated parameter. Variables that did not meet the assumptions of the ANOVA were rank-transformed prior to analysis. When appropriate, multiple pairwise comparisons were done by use of the Holm-Sidak test. A value of P < 0.05 was considered significant.

Results

Six foals completed the study. Three of the 8 foals went into cardiac arrest just after baseline cardiovascular parameters were collected and while drug administration at the LIR was started. One foal assigned to the dobutamine group developed ventricular asystole immediately after obtaining baseline measurements and required resuscitation with drug intervention. Two foals in the norepinephrine group arrested < 1 minute after the start of drug administration at the LIR. One foal was successfully resuscitated by only providing cardiac massage for < 5 seconds, and the other foal required resuscitation with drug intervention. All foals were successfully resuscitated, but only the norepinephrine-group foal that did not require drug intervention was kept in the study. All data from the other 2 foals were excluded. All 8 foals recovered uneventfully and had normal behavior.

No significant differences were found in any of the measured or calculated parameters between groups at baseline. The effects of drug type (norepinephrine, dobutamine, and vasopressin), dose (baseline, LIR, and HIR), and, when applicable, their interactions were determined (Tables 1 and 2). Significant dose-dependent increases in CI and SVI were apparent during norepinephrine and dobutamine administration but not during vasopressin administration. The CI and SVI were significantly higher during dobutamine and norepinephrine infusion than during vasopressin infusion at both rates. The SVR increased during norepinephrine and vasopressin administration at both infusion rates. The SVR was significantly higher during vasopressin and norepinephrine administration at the HIR than during dobutamine administration. As a result, MAP was increased in a dose-dependent manner by norepinephrine and vasopressin, but the increase in MAP caused by dobutamine was not significantly different between the 2 infusion rates. During drug administration at the LIR, MAP was significantly higher during norepinephrine and dobutamine administration than during vasopressin administration. At the HIR, MAP was significantly higher during norepinephrine administration than during either vasopressin or dobutamine administration. Heart rate was only significantly increased by dobutamine.

Measurements of PgCO₂ were not completed in 1 foal receiving dobutamine because of equipment failure. The PgCO₂ significantly increased in all groups after the initial hypotensive period (baseline). The increase in PgCO₂ was also associated with a significant increase in PaCO₂ in all 3 treatments. No significant difference in ΔCO₂ was found in foals treated with dobutamine or norepinephrine, but ΔCO₂ significantly increased during vasopressin infusion. Arterial and central venous lactate concentrations increased significantly after the initial hypotensive period, and pH decreased progressively despite improvement of cardiovascular parameters.

Blood concentrations of hemoglobin increased with the administration of norepinephrine and dobutamine but decreased during vasopressin administration. The ScvO₂ also increased with norepinephrine and dobutamine administration. The increase was dose dependent for dobutamine, whereas no change in ScvO₂ occurred with vasopressin administration. As a consequence, DO₂ increased with norepinephrine and dobutamine administration but not with vasopressin. Conversely, VO₂ did not change with norepinephrine and vasopressin administration and significantly decreased with dobutamine. The ERO₂ decreased significantly with dobutamine administration in a dosedependent manner, whereas it decreased significantly during administration of norepinephrine at both infusions rates, and did not change with vasopressin administration. No significant effects of drug type, dose, or their interactions were found on CVP, endtidal isoflurane concentration, PaO₂, and arterial hemoglobin saturation.

Discussion

Different cardiovascular parameters and indicators of tissue oxygenation can be used to evaluate the effects of vasoactive drugs during hypotension. In our study, vasopressor drugs (norepinephrine and vasopressin) were effective in increasing MAP through increases in SVR, inotropic drugs (norepinephrine and dobutamine) were effective in increasing CI through increases in SVI, and the chronotropic drug (dobutamine) was effective in increasing CI through increases in heart rate.

Dobutamine increased CI by 100% during administration at the LIR and 210% during the HIR. The increase in SVI during dobutamine administration at the LIR and HIR was 64% and 100%, respectively, and the increase in heart rate during dobutamine administration at the LIR and HIR was 17% and 54%, respectively. Because CI is the product of SVI and heart rate, it is interesting that the contribution of SVI and heart rate does not account for the entire magnitude of increase of CI. Conversely, norepinephrine increased SVI by 50% and 125% during administration at the LIR and HIR, respectively, which accounts for the increase of CI of 57% and 126% during norepinephrine administration at the LIR and HIR, respectively.

Anesthetic and vasoactive drugs can alter arterial blood pressure. In our study, isoflurane was a potent vasodilator that caused profound hypotension and cardiovascular suppression at end-tidal isoflurane concentrations of 2.0% to 2.5%. These concentrations represent approximately 2.4 to 3.0 times the minimum alveolar concentration for foals in this age range.^s The SVR in foals from 1 to 10 days of age is approximately 600 to 1,000 dynes·s/cm⁵,¹⁷ whereas measurements of SVR in the foals of our study were < 500 dynes·s/cm⁵ in the presence of isoflurane. The inhalant-induced hypotensive state in the foals of our study resulted in extremely low blood pressures that were probably associated with the cardiac arrest episodes in 3 of the foals that occurred after completing baseline data collection. Resuscitation was successful in all 3 foals, and no shortterm or long-term deleterious effects were observed.

Drugs with vasoconstrictor activity can increase MAP without improving CI or even while decreasing CI, as a result of the negative effects of an increase in SVR and cardiac work. Arterial blood pressure is more frequently monitored in the anesthetized and awake patient because it is more readily measured than CI. However, CI more accurately provides an assessment of overall cardiovascular function. Results of studies^4,5 in anesthetized dogs and foals reveal a poor correlation between CI and arterial pressure. Therefore, assessing cardiovascular status and instituting pharmacologic support on the basis of arterial blood pressure alone may not be entirely appropriate.

Dobutamine is used extensively to support cardiovascular function in critically ill foals and in hypotensive states caused by the depressive effects of inhalation anesthesia.^13,18 Dobutamine better preserves blood flow to muscles than other vasoactive drugs like dopamine, dopexamine, and phenylephrine.¹⁹ In vasodilatory shock, the use of vasoconstrictors offsets the decreased SVR. Norepinephrine, through its α-adrenergic effects, is effective in increasing MAP in septic foals that are refractory to dobutamine treatment.¹³ However, its effects on CI and DO₂ had not been evaluated in hypotensive foals.

In our study, CcvO₂ was measured from the right atrium instead of measuring CmvO₂ from the pulmonary artery to determine VO₂ to avoid possible morbidity complications associated with pulmonary catheter placement. The CcvO₂ may overestimate CmvO₂ by 5% to 18% in patients in shock; however, the usefulness of CcvO₂ has been reviewed, and results are close between the 2 sites (ie, right atrium and pulmonary artery) across a wide range of hemodynamic conditions.²⁰ Because CcvO₂ is used in the calculation of VO₂ and ERO₂, careful consideration should be given to the interpretation of our data for VO₂ and ERO₂. However, the presence of a pathologically low CcvO₂

(implying an even lower CmvO₂) is more clinically important than whether the values are equal.²⁰

In our study, all drugs increased MAP; however, only dobutamine and norepinephrine improved CI and DO₂. In contrast, only dobutamine decreased VO₂ significantly in a dose-dependent manner. Increases in blood lactate concentration are often associated with tissue hypoxia. Blood lactate concentration increased from before anesthesia values (< 1.0 mmol/L for all foals) during the hypotensive state and continued to increase with all treatments during the LIR, despite improvement of most cardiovascular variables (CI, MAP, or both). At the end of drug administration at the HIR, blood lactate concentrations stabilized but failed to decrease, despite a 2- to 3-fold increase in DO₂ during norepinephrine and dobutamine administration. It is possible that the duration of treatment (15 minutes) was too short to document a decrease in blood lactate concentrations. Combinations of dobutamine and norepinephrine administration better reestablish CI and DO₂ than norepinephrine alone in septic human patients, as a result of the cardiac (β₁-inotropic) and peripheral vascular effects (β₂-vasodilatory) of dobutamine that counteract the α-adrenergic effects of norepinephrine.²¹ Effects of combining dobutamine and norepinephrine were not evaluated in our study.

In severe states of tissue dysoxia (ie, oxygen supply that is inadequate to support oxygen demand) and acidosis, catecholamine-mediated vasoconstriction is impaired but vasoconstriction can still be elicited by vasopressin.^6,22 Peripheral effects of vasopressin are mediated by different vasopressin receptors, V1a, V1b, and V2.²¹ The V1-receptors are found in arterial blood vessels and induce vasoconstriction by an increase in cytoplasmic ionized calcium via the phosphatidylinositol-bisphosphonate cascade.^6,23 Vasopressin increases vagal and decreases sympathetic tone by acting on V1-brain receptors.^24,25 Most arterial beds vasoconstrict in response to vasopressin; however, pulmonary, coronary, and vertebrobasilar beds have a nitric oxide–mediated vasodilation response.^26–30 Vasopressor effects are strongest in the splanchnic, muscular, and cutaneous vasculature.³¹ Effects of vasopressin are marked in hypotensive states and less evident under normotensive conditions,⁶ as the physiologic baroreflex-mediated reduction of CO in the normotensive patient in response to an increase in SVR is reduced in the hypotensive patient.^22,32–34

The control and release of endogenous vasopressin depend on cardiopulmonary afferents being sensitive to heart volume³⁵ and on arterial baroreceptors.³³ Endogenous vasopressin is released during hypotension caused by inhalation anesthesia³⁶ and epidural anesthesia³⁷ and during hemorrhage.³⁵ Release of vasopressin compensates for the dose-dependent hypotensive effects of sevoflurane, thereby avoiding more severe hypotension.³⁶ In a recent study,³⁸ a significant increase in endogenous vasopressin concentrations peaking approximately 10 minutes after induction of hypotension was found in 1-to 2-week-old foals. The modest cardiovascular effects achieved by administration of exogenous vasopressin in our study may have been the result of the fact that endogenous release of vasopressin was already maximized as a result of profound hypotension. Plasma concentrations of vasopressin are inappropriately low in hypotensive human patients with prolonged septic shock.³⁹ If vasopressin depletion also occurs during sepsis in horses, vasopressin may prove to be more effective in restoring cardiovascular function in septic foals than in foals with isoflurane-induced cardiovascular suppression.

The DO₂ and VO₂ do not always reflect tissue oxygenation.¹⁴ Similarly, mixed venous O₂ tension and byproducts of tissue metabolism that result from tissue dysoxia (such as blood lactate concentrations) are only crude indices of overall O₂ balance.¹⁴ Tissue dysoxia can result from decreased O₂ availability (hypoxia, anemia, or stagnant hypoxia) or diminished ability to use O₂ (cytopathic hypoxia). The result of tissue dysoxia is an increase in tissue partial pressure of CO₂.^40,41 The splanchnic circulation has a higher critical DO₂ than the whole body and other vital organs; therefore, it is susceptible to decreased tissue perfusion and oxygenation.^41,42

Gastric tonometry is minimally invasive and provides indirect information about gastrointestinal perfusion.¹⁵ Tissue dysoxia as a result of hypoperfusion or malperfusion (stagnant hypoxia) in the splanchnic area is an important cofactor in the pathogenesis of multiple-organ dysfunction syndrome. Patients with septic shock often die of multiple-organ dysfunction syndrome, despite having achieved normal systemic cardiovascular status by administration of fluid resuscitation and vasoactive drugs. Studies investigating the effects of inotropic drugs and vasopressors on gastric mucosal perfusion in humans with sepsis or septic shock have produced conflicting results. A combination of dobutamine and norepinephrine⁴³ or dobutamine alone⁴⁴ improves gastric mucosal blood flow in septic human patients, as assessed by a decrease in DCO₂. In contrast, in another study⁴⁵ dobutamine did not decrease the ΔCO₂. In our study, norepinephrine and dobutamine administration did not result in significant changes in ΔCO₂, despite significant improvement in CI, arterial pressure, and overall DO₂. The duration of treatment may have been too short to document a decrease in ΔCO₂.

The increased ΔCO₂ observed during the HIR in the vasopressin group in our study indicates an increase in CO₂ production as a result of the constant mismatch throughout the experiment between splanchnic DO₂ and VO₂ or the inability of vasopressin to improve overall DO₂. In humans, negative effects of vasopressin on gastrointestinal perfusion do occur. In a study⁹ of human patients in septic shock, continuous infusion of vasopressin was successful in maintaining MAP but resulted in a significant increase in ΔCO₂, compatible with gastrointestinal hypoperfusion. Similarly, vasopressin transiently increases ΔCO₂ when used to counteract hypotension during combined general and epidural anesthesia, suggesting the development of slight splanchnic hypoperfusion.⁴⁶ Pigs with experimentally induced endotoxemia also develop gastrointestinal hypoperfusion during vasopressin infusion.^10,11

In contrast, in human patients with catecholamine-resistant vasodilatory shock, vasopressin combined with norepinephrine results in better cardio vascular function (CI, MAP, and SVI) and gastric mucosal blood flow than norepinephrine alone.⁷ A possible explanation for these conflicting results is that in low doses, vasopressin improves gastrointestinal blood flow by causing vasodilatation of the splanchnic vascular bed.⁴⁷ In our study on foals, vasopressin administration alone resulted in an increase in the ΔCO₂ at the HIR. This effect can probably be attributed to local vasoconstriction, despite the fact that the increase in SVR with vasopressin administration was similar to the increase in SVR with norepinephrine administration. The increased CI that resulted from norepinephrine but not vasopressin administration may account for differences between the effects of these drugs on the splanchnic circulation.

Recent recommendations for vasopressor and inotropic support in septic shock consider norepinephrine or dopamine as the vasopressors of choice in the treatment of septic shock and dobutamine as the drug of choice to increase CO. Epinephrine, phenylephrine, and vasopressin are not recommended as first-line drugs in the treatment of septic shock, but vasopressin is used for treatment of catecholamine-resistant patients.⁴⁸ On the basis of the results of our study, norepinephrine and dobutamine are appropriate choices to treat neonatal foals with isoflurane-induced hypotension because of improvements in MAP, CI, and overall DO₂ and the absence of detrimental effects on ΔCO₂. In contrast, vasopressin failed to increase CI and overall DO₂ and resulted in an increase in ΔCO₂.