Abstract
Two anaesthetic protocols were compared using pregnant sheep. In both groups of animals, anaesthesia was induced using an intravenous (i.v.) injection of diazepam and ketamine. The ewes were then intubated for positive pressure ventilation using 0.8 L/min of nitrous oxide and 2 L/min oxygen with 1.1–1.8% halothane. If the ewe showed any signs of awakening, one of two protocols was followed. First, the halothane concentration was increased to 2–3% until the ewe was completely anaesthetized. Second, the halothane concentration was not altered, but the ewe was given doses of i.v. diazepam (0.1 mg/kg) and ketamine (1 mg/kg) until again completely anaesthetized. At the completion of surgery, maternal recovery was rapid and similar between the two groups. However, five days after surgery, the fetal arterial Po2 and oxygen content of the fetuses receiving additional halothane (1.9 ± 0.2 kPa and 4.4 ± 1.0 mL/100 mL) were statistically significantly depressed when compared with the fetuses receiving additional diazepam and ketamine (2.9 ± 0.1 kPa and 7.0 ± 0.5 mL/100 mL). These results led us to conclude that certain anaesthetic protocols, in spite of good maternal recovery, can lead to deleterious effects upon the fetus that persist for at least five days after surgery.
Much of our information regarding fetal physiology has been obtained from experiments in pregnant sheep (Anderson et al. 2005, Faber et al. 2005, Giraud et al. 2005, Yang et al. 2005). For many of these experiments, an integral part of the protocol involves anaesthetizing the ewe and catheterizing the fetus. Experiments are begun after allowing time for recovery from surgery, generally three to five days. Physiologically meaningful data are dependent upon good recovery of both the ewe and the fetus. Here, we report our serendipitous observation that a relatively minor change in the protocol for anaesthetizing pregnant sheep resulted in a deterioration of the fetus that persists for at least five days after surgery.
Materials and methods
Anaesthesia
Methods of anaesthesia and surgery protocols on pregnant ewes were approved by the Institutional Animal Care and Use Committee (IACUC). Surgeries were performed on 15 pregnant ewes of mixed Western breeds (Ovis ovis) at about 120 days gestation (term ∼147 days). For 24 h prior to surgery, the ewes were denied food but provided with water ad libitum. Immediately prior to the induction of anaesthesia, each ewe (weighing approximately 60–70 kg) received an intramuscular injection of 7.5 mg of atropine (Neogen Vet, Lexington, KY, USA). Anaesthesia was induced using an initial i.v. injection of diazepam (10 mg) (Abbott Laboratories, Abbott Park, IL, USA) together with ketamine (400 mg) (Fort Dodge Animal Health, Overland Park, KS, USA). Rarely an additional dose, up to one-half the original dose, was titrated as needed to allow intubation. The ewe was then intubated for positive pressure inhalation anaesthesia (Ohmeda Modulus II Plus anaesthesia machine; GE Healthcare, Bedford, CT, USA) using 0.8 L/min of nitrous oxide and 2 L/min of oxygen with 1.1–1.8% halothane (Halocarbon Laboratories, River Edge, NJ, USA). An i.v. infusion of lactated Ringer's solution (B/Braun, Bethlehem, PA, USA) was begun (∼8 mL/min) into the jugular vein of the ewe. In preparation for the surgery, the ewe was placed in dorsal recumbency on a heating pad positioned on a cradle resting upon the surgical table. After the abdomen of the ewe was shorn and cleansed (Betadine surgical scrub, Povidone iodine 7.5%), the ewe was transported into the surgical suite. One hour before the anticipated completion of surgery, the lactated Ringer's solution was stopped and a 2.5% dextrose solution (B/Braun) was begun at the same rate and continued until the completion of surgery.
Surgery
Maternal pulse rate (Nonin pulseoximeter; Nonin, Plymouth, MN, USA), end tidal Pco2 (Criticare Systems Inc, Waukesha, WI, USA), respiratory rate (Ohmeda) and arterial oxygen saturation (Nonin pulseoximeter) were monitored every 15 min for the duration of surgery and all remained within the normal range. Using sterile surgical techniques, a midline incision was made on the ewe. A purse string suture was placed in the uterus over the fetus. An incision was made in the uterus and the fetal membranes were attached to the uterus using interrupted sutures. Either the head (for the unilateral catheterization of a carotid artery and jugular vein
The different treatments of the two groups of animals began during surgery when the ewe showed any signs of awakening, as judged by an increase in heart rate, blink reflex or an increase in jaw tone. In the first group of six ewes (seven fetuses), the percentage of halothane in the inspired gas was increased to 2–3% until the ewe was completely anaesthetized. At this point, the halothane concentration was returned to its original level. In the second group of nine ewes (13 fetuses), the level of halothane was left constant. Instead, the ewe was slowly given i.v. diazepam (∼0.1 mg/kg) and ketamine (∼1 mg/kg) until the ewe was completely anaesthetized.
At the completion of the surgery, anaesthesia was stopped and both groups were treated identically from here on. Once the ewe was spontaneously breathing, the endotracheal tube was removed. The ewe was returned to her pen and was generally eating within 15 min. All ewes received routine postoperative pain medication (0.6 mg buprenorphine; Bedford Laboratories, Bedford, OH, USA), twice a day for two days. Adequacy of analgesia was verified by observing that ewes had a normal appetite, relaxed posture and were bright, alert and responsive.
Experiment
After five days recovery from surgery, the ewe was taken to the laboratory and confined in a stanchion where she had free access to food and water and was able to stand or lie down. Fetal arterial, venous and amniotic fluid pressures were measured using pressure transducers (Abbott Transpac; Abbott Laboratories) and a Macintosh-based recording system. Intravascular pressures were reported with respect to amniotic fluid pressure. Heart rates were determined from the arterial pressure recording. A fetal arterial blood sample was collected for immediate analysis (Instrumentation Laboratories IL306 for pH, Pco2 and Po2 and IL482 for oxygen content; Instrumental Laboratories, Lexington, MA, USA). All of these data were collected before the initiation of any other experimentation.
Statistical methods
The data collected on the first experimental day were used to compare the two groups of ewes using Student's t-test for unpaired data (GraphPad Software; GraphPad Software Inc, La Jolla CA, USA). In the case of twin pregnancies, the data from each of the twins were averaged, so as to give each ewe equal weight for the statistical analyses. All results are reported as mean ± SEM. Differences were judged to be statistically significant if P < 0.05.
Necropsy
At the completion of the experiments, the ewes and fetuses were euthanized using an i.v. injection of a commercial euthanasia solution containing sodium pentobarbital (390 mg/mL) and sodium phenytoin (50 mg/mL) (Euthasol; Virbac Animal Health Inc, Fort Worth, TX, USA) to the ewe as approved by the IACUC. This solution will cross the placenta and euthanize the fetus. Postmortem examinations were performed, and good surgical recovery and catheter placement were confirmed in all animals. There was no evidence of infection in any ewe or fetus.
Results
The group receiving supplemental halothane during surgery consisted of six ewes carrying seven fetuses. The group receiving supplemental ketamine and diazepam during surgery consisted of nine ewes carrying 13 fetuses. There were no differences in the two groups with respect to gestational age at surgery (120 ± 1 day with increased gaseous anaesthetic vs. 119 ± 1 day with increased i.v. anaesthetic), duration of anaesthesia (235 ± 30 min with increased gaseous anaesthetic vs. 263 ± 13 min with increased i.v. anaesthetic), duration of surgery (175 ± 31 min with increased gaseous anaesthetic vs. 210 ± 12 min with increased i.v. anaesthetic) or gestational age on the day of the experiment (125 ± 1 day with increased gaseous anaesthetic vs. 124 ± 1 day with increased i.v. anaesthetic).
Arterial blood pressure, venous blood pressure and heart rate were compared and were not statistically significantly different from each other (Table 1). However, this was not the case when the arterial blood gases were compared. These results are shown in Table 1 and Figure 1.
Comparison of fetal haemodynamics and arterial blood gas values collected after five days of recovery from surgery
*Data reported for only five fetuses due to catheter failure
NS = No statistically significant differences comparing the anaesthetic protocols using unpaired t-test
Discussion
Surgery during pregnancy has become more common both for humans (Mazze & Kallen 1989, Schwarz & Galinkinb 2003, Littleford 2004) and animals (Anderson et al. 2005, Faber et al. 2005, Webster et al. 2005). Without access to the fetus, it is easy to judge a surgery a success if the mother shows good recovery. The results presented here demonstrate that good maternal recovery is not necessarily proof of good fetal recovery. From our results, it is apparent that what could be viewed as a minor change in anaesthetic protocol (with no apparent effect upon maternal wellbeing) has long-lasting effects on the fetus. This observation takes on greater significance in the face of the fact that, in human surgeries, good fetal recovery does not always occur after maternal anaesthesia (Mazze & Kallen 1989).
Pregnant sheep are one of the most commonly used large animals for fetal physiology research. This is in part because, during the last third of gestation, the fetus is relatively large. As a result, the surgical preparation of the fetus can be quite varied, ranging from placement of a few catheters to instrumentation with multiple flow sensors. Essential to the success of any experiment is having a fetus that has recovered from surgery and is in good physiological condition. Here, we compare two anaesthetic protocols for pregnant sheep. While both protocols provide adequate maternal and fetal anaesthesia, only one results in a fetus which could be considered physiologically normal based upon the parameters we examined.
Webster et al. (2005) have recently described an anaesthetic protocol that resulted in both good maternal and fetal recovery. These investigators induced anaesthesia with 10–15 mg/kg i.v. thiopentone sodium and then transferred ewes to the surgical suite on halothane with spontaneous ventilation. They then infused the opioid remifentanil at 0.5 µg/kg/min, titrated up to 2 µg/kg/min, with 0.5–1.5% halothane (similar to the amount described here) and mechanical ventilation for the remainder of the surgery. Following recovery, they note that fetal Po2 values varied from 2.0 to 3.0 kPa, a range that overlaps both groups described in this study (halothane-supplemented ranged from 1.1–2.3 kPA, while ketamine-supplemented ranged from 2.4–3.5 kPA).
Five days of recovery after minor fetal surgery is generally viewed as adequate for both the fetus and the ewe. As can be seen from these results, many of the cardiovascular parameters routinely examined as indicators of fetal wellbeing fall in the normal range for both groups (Anderson et al. 2005, Faber et al. 2005, Giraud et al. 2005, Yang et al. 2005). However, the decrease in the partial pressure of oxygen and in the haemoglobin-oxygen content of the arterial blood suggests that there has been some disruption in the normal oxygen delivery to the fetus. A decrease in fetal oxygenation results when there is a disturbance to the fetal oxygen delivery system. Possible explanations include a decrease in maternal placental blood flow, a decrease in fetal placental blood flow or an increase in the placental diffusion resistance for oxygen. It is not possible to identify the cause of the decrease in oxygen delivery from these results. However, these results demonstrate that, even in the face of a good maternal recovery and normal fetal haemodynamic measurements, there may be subtle, persistent fetal problems that are probably attributable to choice of anaesthetic agents.
Footnotes
Acknowledgements
The authors wish to thank Robert Webber, Loni Socha and Ed Hedge for their assistance with these experiments. This research was supported by NICHD:5PO1-HD34430, HD37376 and HL45043.