Thoracic Epidural Anesthesia with Ropivacaine Does Not Compromise the Tolerance of Acute Normovolemic Anemia in Pigs

THORACIC epidural anesthesia (TEA) has become an integral part of the anesthetic management of major surgery (e.g., abdominal, thoracic, or urologic surgery). The initial treatment of surgical blood loss usually consists of the infusion of acellular fluids to maintain normovolemia. The consequence is a dilution of the red cell mass remaining in the vascular system (acute normovolemic anemia). In this context, the term “anemia tolerance” refers to (1) the patient’s physiologic potential to tolerate even profound stages of acute anemia, and (2) the anesthesiologist’s intention to accept low hemoglobin concentrations to avoid premature and unnecessary transfusions of allogeneic blood products.¹ 

As a result of the organism’s tolerance to anemia, tissue oxygenation is maintained, although oxygen transport capacity decreases progressively with dilutional anemia. At physiologic hemoglobin concentrations, oxygen supply to the tissues (Do₂) exceeds total body oxygen demand—reflected by oxygen consumption (Vo₂) under rest conditions—by the factor three to four.²  Therefore, oxygen supply to the tissues is sufficient to meet demand over a wide range of hemoglobin concentrations. As a consequence, Vo₂ remains constant, even when Do₂ decreases with falling hemoglobin concentrations (supply independency of Vo₂, see fig. 1). Acute anemia is essentially compensated by increases in cardiac output and arteriovenous oxygen extraction,³  the latter occurring to a variable extent within different organs.⁴  When these compensatory mechanisms are exhausted (i.e., when cardiac output and oxygen extraction cannot be further increased), the amount of oxygen delivered to the tissues becomes insufficient to meet demand and Vo₂ starts to decline (supply dependency of Vo₂).^5,6  The hemoglobin concentration corresponding with the sudden decrease in Vo₂ is referred to as the “critical” hemoglobin concentration (Hb_crit) and reflects the individual limit of an organism’s tolerance to anemia.⁷ 

Aside from analgesic effects, sympathetic block produced by TEA has been demonstrated to optimize myocardial and splanchnic oxygen balances and is therefore purported to be cardioprotective, particularly in high-risk patients.⁸  On the contrary, thoracic sympathetic nerve block may result in bradycardia, vasodilation, and hypotension and, thus, in reduced cardiovascular performance. Hence, we hypothesized that (1) the sympathetic nerve block elicited by TEA might compromise hemodynamic compensation and, therefore, the tolerance to acute anemia, and (2) Hb_crit would be met at higher hemoglobin levels than in individuals not subjected to TEA.

After approval by the local governmental review board (Department for veterinary affairs/V54, Regional Council Darmstadt, Germany), experiments were carried out in 18 healthy farm-bred pigs of either sex (body weight 23.8 ± 1.9 kg). All animals received care in compliance with the Guide for the Use of Laboratory Animals.

Twelve hours prior to the experiments, animals were denied food but had free access to water. After intramuscular premedication with ketamine (10 mg/kg) and midazolam (1 mg/kg), anesthesia was induced by intravenous injection of propofol (3 mg/kg) and fentanyl (30 µg/kg) and maintained by continuous infusion of propofol (160 µg·kg⁻¹·min⁻¹), fentanyl (0.8 µg·kg⁻¹·min⁻¹), and midazolam (10 µg·kg⁻¹·min⁻¹). To facilitate intubation and mechanical ventilation, muscular paralysis was achieved with pancuronium bromide (bolus injection 200 µg/kg followed by continuous infusion of 130 µg·kg⁻¹·min⁻¹). Estimated fluid losses were replaced with Ringer’s solution (3 ml·kg⁻¹·h⁻¹). The trade names of all anesthetic drugs applied are listed in the appendix.

Animals were orotracheally intubated and ventilated with a Servo 900B ventilator (Siemens-Elema, Solna, Sweden; Fio₂ 0.21, 14 cycles/min, positive end-expiratory pressure 5 cm H₂O). Tidal volume was individually adjusted to provide arterial normocapnia and was then maintained throughout the entire protocol.

After induction of anesthesia, animals were placed in the prone position for insertion of an epidural catheter. The fifth and the tenth thoracic vertebrae and the eighth interspace (Th8/9) were identified using fluoroscopy, and the borders of the dermatomes Th5 and Th12 were marked. The eighth interspace was surgically dissected under sterile conditions (resection of the interspinal ligaments and the spinal process of the eighth vertebra). The ligamentum flavum was visually identified, an 18-G Tuohy needle (Braun, Melsungen, Germany) was carefully advanced into the epidural space, and a 20-G epidural catheter was inserted for 5 cm. The catheter was fixed in this position before the incision was closed.

Subsequently, animals were placed in the supine position, and a five-lead electrocardiogram (II, V5) was placed. A central venous catheter (Arrow, Reading, PA) was inserted into the cranial vena cava, and a Swan-Ganz catheter (Baxter, Irvine, CA) was floated into a pulmonary arterial branch for monitoring of pulmonary arterial pressures and withdrawal of mixed venous blood samples. A 6-F introducer sheath was inserted into the right femoral artery and vein, respectively. A thermodilution catheter (Pulsion, Munich, Germany) was placed into the left femoral artery for continuous measurement of arterial blood pressure and cardiac output. Ambient temperature was kept constant at 23°C, and body core temperature was monitored during the entire protocol.

After completion of instrumentation, a 60-min stabilization period was allowed to elapse to achieve stable baseline conditions. Upon recording the first data set (“Baseline”), animals were randomized to receive an epidural bolus injection of either 5ml ropivacaine 0.2% or normal saline (ROP group, n = 9 vs. SAL group, n = 9). The investigator remained blinded until the completion of the epidural injection.

The intended spread of sympathetic block was Th5-10. If necessary, additional 1-ml boluses of ropivacaine were injected to establish this degree of sympathetic block; animals of the SAL group did not receive further epidural bolus injections. For maintenance of the TEA level, a continuous epidural infusion of ropivacaine (5 ml/h) was initiated (animals of the SAL group received the corresponding volume of normal saline) and the second data set (“TEA”) was recorded.

As a measure of sympathetic block, skin temperature was assessed before and 30 min after epidural injection by infrared thermometry, a method that has been described elsewhere.⁹  Using a handheld infrared thermometer (Micro-Epsilon Messtechnik, Ortenburg, Germany), skin temperature was measured along the right and left anterior axillary lines in the dermatomes Th5–Th10 and on two reference points on the surface of the lower abdomen below Th12. Within each dermatome, a difference in temperature of at least 0.5°C compared to the reference points was considered to indicate a sympathetic block.

Subsequently, a hemodilution protocol was initiated with replacement of whole blood for hydroxyethyl starch (6% HES 130/0.42) in a 1:1 ratio. For precise synchronization of blood withdrawal with the HES infusion rate (1 mL·kg⁻¹·min⁻¹), a bidirectional precision pump (Harvard Apparatus, Holliston, MA) was used. Target parameter was the animal’s individual Hb_crit, which was assessed in an automated and investigator-independent manner (see the next section, “Assessment of Hb_crit”).

When Hb_crit was reached, skin temperature was again scanned to verify the final level of the sympathetic block. Thereafter, the final data set was collected (“Hb_crit”) and animals were sacrificed by intracardiac injection of potassium chloride.

Hb_crit correlates with critical limitation of Do₂ and marks the onset of total body oxygen supply dependency (see fig. 1). Using specific software (Delta-Crit-System),¹⁰  the corresponding Vo₂ decrease was detected by analysis of the Vo₂ values obtained every 60 s. Vo₂ was assessed with a metabolic monitor (Delta-Trac II, Datex-Engstrom, Helsinki, Finland). During the stabilization period, Delta-Crit-System incorporated the Vo₂ values into a regression analysis and calculated the mean and SD for this baseline value. During the subsequent hemodilution period, Vo₂ values were compared with the mean value baseline calculated by Delta-Crit-System. When three subsequent Vo₂ values fell below a predefined level (three times the SD of the baseline regression line), a significant Vo₂ decrease was signaled by visual and acoustic computer alerts and recorded.

Intravascular blood volume was determined at baseline using “whole-blood” dilution of the indocyanin green.¹¹  Measurements of hemodynamic and oxygen-derived parameters were performed at “Baseline,” “TEA,” and at “Hb_crit.”

Cardiovascular catheters were connected with pressure transducers and linked with a multichannel recorder (Hugo-Sachs, March-Hugstetten, Germany). All measurements were recorded with a personal computer. Arterial and mixed venous blood samples were withdrawn for blood gas analyses and assessment of hemoglobin concentration (GEM-3000 and 682 CO-Oxymeter; Instrumentation Laboratory, Lexington, MA). Derivative parameters were calculated as described in the appendix.

Statistical analysis was performed with the SAS 9.3 software package (SAS Institute, Cary, NC). All data are presented as mean ± SD. The sample size was estimated according to our experience with previous studies using a similar experimental protocol. Distribution of data was assessed with the Shapiro-Wilk test. In the case of normal distribution, time effect on the different variables as well as differences between groups at the investigated time points were tested by a two-way repeated-measures ANOVA. Post hoc analysis of differences detected with ANOVA was performed with the Student–Newman–Keuls test. In the case of non-normal distribution, time effect on the parameters as well as between-group differences were tested by ANOVA on ranks. Post hoc analysis of differences detected with ANOVA on ranks was performed with the Tukey test. For all parameters, statistical significance was accepted at P value less than 0.05.

There were no significant differences between the groups regarding body weight and all investigated parameters observed at baseline (see table 1).

As a measure of sympathetic block, skin temperature was measured along the left and right anterior axillary lines within the dermatomes Th5–Th10 (see fig. 2). While the epidural injection of normal saline was void of any effects (see fig. 2, B and D), TEA with 0.2% ropivacaine elicited a sympathetic block of the segments Th5–Th10, which was reflected by an increase in skin temperature (see fig. 2, A and C).

As a consequence of sympathetic block, systemic vascular resistance index decreased by 34% (P = 0.06), which was accompanied by decreases in mean arterial pressure (MAP) by 17% (P = 0.07) and coronary perfusion pressure by 19% (P = 0.056) and a moderate increase of cardiac index (P = 0.32). Moreover, mixed venous oxygen saturation (Svo₂) increased by 33% (P = 0.006), while oxygen extraction ratio (O₂-ER) decreased by 45% (P = 0.002).

The induction of critical normovolemic anemia required the exchange of 106 ± 40% of circulating blood volume in the ROP group and 99 ± 31% in the SAL group. In both groups, Hb_crit was met at 2.5 ± 0.6 g/dl (difference of means: −0.05, 95% confidence interval: −0.64, 0.53), so that no differences between the groups could be observed (P = 0.86). Hemoglobin concentrations are displayed at baseline, TEA, and Hb_crit in figure 3.

During hemodilution, arterial oxygen content decreased proportionally with falling hemoglobin concentrations (P < 0.001 vs. TEA in either group). To compensate for acute anemia, cardiac index (CI) and O₂-ER increased significantly in both groups. The increase of CI (P < 0.001 vs. TEA in either group) was achieved by increases in heart rate and stroke volume. Systemic vascular resistance index decreased by 61% and 56%, respectively (P < 0.001 vs. TEA in either group), as MAP decreased by 24% in the ROP group (P = 0.012) and by 11% in the SAL group (P = 0.33). Coronary perfusion pressure decreased by an average of 41% in the ROP group (P < 0.001) and by 24% in the SAL group (P = 0.039). At Hb_crit, animals subjected to TEA were more hypotensive than controls, as indicated by lower values of mean arterial and pulmonary arterial pressures (P = 0.011 and P = 0.009 vs. SAL group, respectively).

Despite the compensatory increase of CI, oxygen delivery (Do₂) decreased significantly (ROP group: P = 0.002, SAL group: P = 0.005) during progressive anemia. Hence, O₂-ER increased by 40% in the ROP group (P = 0.03 vs. TEA) and by 27% in the SAL group (P = 0.037 vs. TEA). While base excess decreased in both groups (P = 0.018 and P = 0.017, respectively), lactate concentration was not yet elevated at Hb_crit in either group.

The main finding of the present study is that sympathetic block (Th5–Th10) induced by TEA with 0.2% ropivacaine has no effect on the limit of tolerance to anemia. This was reflected by identical values of Hb_crit in animals subjected to TEA compared to those that received epidural saline.

Hb_crit represents the ultimate limit of tolerance to anemia at the level of the whole organism.^5,12  In previous experimental studies, we could demonstrate that this critical limitation of oxygen supply was associated with 100% mortality, if no treatment (e.g., elevation of Fio₂, transfusion of red blood cells, or infusion of artificial oxygen carriers) was initiated at Hb_crit.^13–15  Hb_crit has been chosen as the primary endpoint in a variety of studies investigating the impact of interventions on the limit of tolerance to acute anemia. As reflected by reduced values of Hb_crit (listed in table 2), tolerance to anemia was found to be increased by (1) hypothermia (moderate reduction of body core temperature reduces total body oxygen demand),¹⁶  (2) hyperoxic ventilation (excellent bioavailability of oxygen physically dissolved in the plasma),^13,17–19  (3) infusion of norepinephrine (stabilization of coronary perfusion pressure during hemodilution)²⁰  or infusion of artificial oxygen carriers (maintenance of arterial oxygen content despite reduced hematocrit),^15,21,22  (4) deep neuromuscular block (lowering skeletal muscular oxygen demand),²³  and (5) use of colloidal rather than crystalloid fluids for volume replacement (maintenance of microvascular perfusion due to reduced extravasation rate).²⁴  In contrast, tolerance to anemia was reduced by (1) hypovolemia (reduction of microcirculatory oxygen supply), (2) profound anesthesia (pharmacologic reduction of cardiac output),²⁵  and (3) coronary artery disease (reduced coronary flow reserve).²⁶ 

TEA today has become an integral part of perioperative care during and after major surgery, as it provides an effective analgesia.²⁷  TEA-related effects beyond analgesia (i.e., sympathetic block) have been demonstrated to improve clinical outcome in patients undergoing cardiac and noncardiac surgery.⁸  Some meta-analyses^28,29  suggest a reduction of cardiopulmonary morbidity and mortality after major surgery performed with TEA. Moreover, TEA might improve intestinal perfusion and tissue oxygenation,^30–32  gut motility,³³  and anastomotic patency.³⁴  In experimental studies, hepatic³⁵  and intestinal³⁶  perfusion remained unaltered despite significant hypotension elicited by TEA.

However, experimental data regarding cardiovascular performance in pigs subjected to TEA are equivocal: While the hemodynamic response to cross-clamping of the thoracic aorta was not affected by sympathetic block,³⁷  the inotropic reserve of the right ventricle was compromised by TEA.³⁸ 

We hypothesized that TEA might impair the hemodynamic compensation to acute normovolemic anemia resulting in a compromised tolerance to anemia. If so, this should lead to caution in use of TEA in high–blood-loss surgery to reduce transfusion.

Our data do not confirm this hypothesis. The assumed hemodynamic depression (i.e., hypotension and limited cardiac output) was predominantly observed in experimental studies using higher concentrations of local anesthetics for TEA.^31,35,36,39  Contrary to our expectations, the present data suggest that hemodynamic compensation for acute anemia (i.e., increase of cardiac output) was fully maintained during TEA with 0.2% ropivacaine.

Although the decrease in MAP was moderate after injection of ropivacaine, MAP and, thus, coronary perfusion pressure were significantly lower at Hb_crit, indicating that hypotension was more severe in animals subjected to TEA combined with profound anemia than in those experiencing acute anemia alone.

Our findings obtained in healthy pigs may not apply to patients with cardiovascular comorbidity (e.g., coronary artery disease). In the presence of coronary arterial sclerosis, a decrease in coronary perfusion pressure may result in a critical limitation of myocardial oxygenation even at higher hemoglobin concentrations.

The spread of sympathetic block was assessed by measurement of skin temperature, which is an indirect method to assess sympathetic nerve activity. In contrast to direct techniques (e.g., microneurography), thermometry is a simple and noninvasive method that allows a good chronological and spatial resolution of measurements.⁴⁰  Moreover, skin temperature assessed by infrared thermometry has been validated as a surrogate parameter of sympathetic activity in clinical^9,40  as well as in experimental^38,41  studies.

In the present study, Hb_crit was derived from mean total body oxygen consumption. The Delta-Trac II metabolic monitor provides a reliable method of Vo₂ measurement, which has been validated.^10,42  Although the assessment of Hb_crit at the level of the whole organism has been used to investigate the impact of interventions on tolerance to anemia, this approach neglects the limits of anemia tolerance on the cellular and the organ-specific level. While the present study was not designed to address this point, results of a recent experimental study suggest that in the kidneys and in the intestine signs of tissue hypoxia occur before the critical limit of systemic oxygen supply is reached in anesthetized pigs subjected to acute normovolemic hemodilution.⁴  In contrast, hepatic,³⁵  gastric,³¹  and intestinal³⁶  tissue oxygenation were found unaltered despite significant systemic hypotension elicited by high-dose TEA.

Clinical data suggest that TEA improves myocardial oxygen balance in patients undergoing coronary bypass surgery, resulting in increased myocardial contractility.^27,43  This finding could have had a positive effect on tolerance to anemia by improving the cardiac response to acute anemia. In the present study, however, the institution of low-dose TEA did not affect CI, but elicited a decrease in O₂-ER and systemic vascular resistance index while Vo₂ remained constant, which may reflect improved tissue perfusion. Theoretically, this finding could also have contributed to an augmentation of tolerance to anemia, for example, due to augmentation of splanchnic tissue oxygenation.

Despite these potentially positive effects of TEA on anemia tolerance, the values of Hb_crit were identical in both groups, indicating that TEA did neither improve nor diminish tolerance to anemia. In particular, the compensatory increases in O₂-ER and CI during progressive anemia were unrestricted despite the presence of a sympathetic block.

In summary, our data indicate that the limit of tolerance to anemia was not compromised by TEA using 0.2% ropivacaine in healthy pigs. In clinical practice, however, blood transfusion should be initiated before this limit is reached, that is, before Hb_crit is reached, particularly since comorbid cardiovascular diseases or critical illness may severely restrict an individual’s tolerance to acute anemia. Yet, a blood transfusion may occasionally be impossible, for example, having to “bridge” an acute blood loss until compatible blood products are available or refusal of blood transfusion for religious reasons. Even in these situations, where extremely low hemoglobin concentrations must be tolerated, essential compensatory mechanisms of acute anemia were maintained over a wide range of hemoglobin values despite sympathetic block.

TEA with ropivacaine does not compromise the tolerance of acute anemia in healthy pigs, as essential compensatory mechanisms were fully preserved despite sympathetic block. Thus, TEA appears to be a reasonable choice even in major blood-loss surgery. However, whether the threshold for transfusion should be altered in those receiving TEA—particularly in the presence of cardiovascular comorbidity—remains to be elucidated.

The authors thank Alf Theisen (D.V.M.), Waldemar Daut, and his team for their excellent animal care and Heinz Winkelmeier for his valuable technical assistance. Dr. Theisen, Mr. Daut, and Mr. Winkelmeier are employed at the Central Research Facility, Goethe University, Frankfurt/Main, Germany. Sincere thanks are given to Hans-Joachim Wilke, D.E.A.A. (Department of Anesthesiology, Intensive Care Medicine and Pain Therapy, University Hospital Frankfurt, Frankfurt/Main, Germany), for language-editing the present manuscript.

Support was provided soley from institutional and/or departmental sources.

The authors declare no competing interests.

For induction and maintenance of anaesthesia, the following drugs were applied:

• Ketamine (Ketavet, Parke-Davis, Berlin, Germany)

• Midazolam (Midazolam, Ratiopharm, Ulm, Germany)

• Propofol (Brevimytal, Braun, Bad Melsungen, Germany)

• Fentanyl (Fentanyl, Janssen, Neuss, Germany)

• Pancuronium (Pancuronium, Curamed, Karlsruhe, Germany)

• Balanced electrolyte solution (Tutofusin, Pharmacia, Erlangen, Germany)

• 6% HES 130/0.42 (Tetraspan, Braun, Melsungen, Germany).

For epidural anesthesia, ropivacaine (Naropin, AstraZeneca, Wedel, Germany) was applied.

Body surface area (BSA) was calculated according to Holt et al.⁴⁴ :

where BW = body weight (in kg) and k = 9 (constant for the species pig).

Cardiac index was calculated as:

where CO = cardiac output.

Stroke volume index was calculated as:

where HR = heart rate.

Systemic and pulmonary vascular resistance indices were calculated as:

and

and where MAP = mean aortic pressure, CVP = central venous pressure, MPAP = mean pulmonary arterial pressure, and LVEDP = left ventricular end-diastolic pressure.

O₂ extraction ratio (O₂-ER) as calculated as follows:

where Cao₂ and Cvo₂ are arterial and mixed venous oxygen content.

Coronary perfusion pressure was calculated as

where DAP = diastolic arterial pressure and PCWP = pulmonary capillary occlusion pressure.