Effects of clonidine on breathing movements and electrocortical activity in the fetal lamb Owen S. Bamford, PhD, Robert L. Hawkins, BS, and Carlos E. Blanco, MD Baltimore, Maryland Cionidine is a recommended antihypertensive for use during pregnancy, although little is known of its fetal effects. This study examines the effects of clonidine on breathing and sleep-state cycling in fetal lambs. Cionidine was infused into a fetal lateral ventricle for up to 24 hours at 128 to 135 days' gestation. Control infusions of artificial cerebrospinal fluid had no effect. Cionidine infusion signiflCBrltiy reduced the incidence and episode duration of fetal breathing for the duration of the infusion period. Cycling of electrocortical activity became irregular and rapid, and the incidence of high-voltage electrocortical activity (equivalent to quiet sleep) was reduced. Fetal heart rate decreased but arterial pressure was unaffected. After infusion the breathing incidence and episode duration both increased significantly compared with control, with continuous high-amplitude breathing for several hours, whereas the incidence of high-voltage electrocortical activity remained low. Because lung development is promoted by fetal breathing, long-term use of clonidine during pregnancy could slow lung development by reducing fetal breathing actiVity. (AM J OesTET GVNECOl 1990;163:661-8.)
Key words: Clonidine, fetal breathing, electrocortical activity, lung development
The a.-adrenergic agonist clonidine is used extensively as an antihypertensive agent and has been recommended for use during pregnancy. 1 Although it has no teratogenic effects or acute toxicity, the possibility of more subtle developmental effects on the fetus has not been studied. In adults clonidine reduces arterial pressure by acting at inhibitory presynaptic a.-adrenergic receptors on noradrenergic neurones. This causes hyperpolarization and inhibition of firing2-4 and leads to a reduction in sympathetic outflow and a decrease in arterial pressure. Other effects include sedation and hypothermia. Although these effects of clonidine are well known, the fetal effects of clonidine have attracted little attention. Intravascular injections of 10 j.l.g clonidine caused an immediate arrest offetal breathing,' although clonidine had no consistent effect on breathing postnatally.6 The a.-adrenergic antagonist idazoxan blocked the actions of clonidine, thus confirming its a.-adrenergic action, although when given alone idazoxan had little effect on fetal breathing; presumably the a.-adrenergic receptors were not tonically active, and thus their blockFrom the Department of Pediatrics, University ofMarylnnd Hospital. Supported by research funds from the School of Medicine, University of Marylnnd. Received for publication December 2, 1989; revised March 26, 1990; accepted April 4, 1990. Reprint requests: Owen Bamford, PhD, Division of Neonatology, Department of Pediatrics, University of Marylnnd Hospital, 22 S. Greene St., Baltimore, MD 21201 .
611121411
ade made no difference under normal conditions. Clonidine injections also caused an immediate onset of high-voltage electrocortical activity. Breathing in the fetal lamb is normally suppressed in high-voltage electrocortical activity, but apnea and the shift to highvoltage were shown to be independently mediated since both still occurred after complete brainstem transection. This evidence suggested that in the fetus noradrenergic facilitatory tone is required to maintain breathing, perhaps via a 1 receptors, and could be blocked by activating a.-adrenergic receptors either exogenously with clonidine or by an endogenous neurotransmitter, probably noradrenaline. However, we examined only short-term acute effects of systemic injections of these substances and the responses seen could have been nonspecific effects of a transient high plasma concentrations. To address this possibility and to study the effects of maintained high central concentrations of clonidine we infused clonidine into chronically prepared fetal lambs in utero for periods of up to 24 hours. Because prolonged intravascular infusions of diffusible drugs such as clonidine are likely to produce unpredictable indirect effects on the fetus via changes in the maternal circulation, the agents were administered via an indwelling cannula to a fetal lateral cerebral ventricle for periods up to 24 hours while breathing, nuchal electromyogram, electrocortical activity, arterial pressure, heart rate, and arterial blood gases were recorded.
661
662 Bamford, Hawkins, and Blanco
August 1990 Am J Obstet Gynecol
Table I. Effect of lateral ventricle infusion of cerebrospinal fluid on incidence and episode duration of fetal breathing and electrocortical activity Incidence of high-voltage electrocortical activity (min/hr) Incidence of breathing (min/hr) Duration of high-voltage electrical activity episodes (min) Duration of breathing episodes (min)
Control
Infusion
Postinfusion
24.7 ± 3.5
22.4 ± 5.8
25.5 ± 2.1
30.0 ± 4.8 12.1 ± 3.2
29.9 ± 5.5 ILl ± 2.9
26.2 ± 5.1 12.7 ± 2.7
13.8 ± 4.0
14.1 ± 3.2
12.9 ± 3.2
All comparisons between control, infusion, and postinfusion values were not significant by ranked-pairs test.
Material and methods
Experiments were performed on a total of 15 fetal lambs prepared for chronic recording using methods based on those of Boddy et aU All procedures and protocols were approved by the institutional animal care and use committee. Pregnant Dorset cross ewes at between 120 and 125 days' gestation were anesthetized with intravenous Pentothal, intubated, and ventilated with oxygen and nitrous oxide 2: 1 with 2% halothane. Following normal sterile methods, pairs of stainless steel wire recording electrodes were implanted in the fetal diaphragm, nuchal extensor muscle, and parietal cortex. Vinyl catheters (1 mm inner diameter) were placed in a fetal carotid artery and jugular vein and in the trachea. A fourth catheter ended in the amniotic fluid to act as a pressure reference. A lateral ventricle cannula was placed using a method based on that described by Gluckman and Parsons. 8 An acrylic guide tube was anchored to the skull at 7 mm anterior and 7 mm lateral to bregma with screws and cyanoacrylate glue. A saline-solution-filled cannula made from a 21gauge hypodermic needle bent at right angles was introduced through the guide until it entered a lateral ventricle as shown by the fall in the saline solution column. The cannula was then glued in place in the guide. Vinyl tubing (0.25 mm inner diameter x 1 m) was attached to the exposed end of the cannula via an adapter tube and secured with cyanoacrylate glue. Dead space of catheter and cannula combined was about 100 ILl. The uterus was closed in two layers and catheters and electrodes were exteriorized through a flank incision. Recovery from anesthesia was rapid and the ewes usually stood and fed within an hour of skin closure. Three days were allowed to elapse from the time of operation before any experiments were performed. An infusion rate of 33 lLl/hr was used throughout the study. Drugs were made up in artificial cerebrospinal fluid, adjusted if necessary to pH 6.5 to 7 with carbon dioxide, and sterilized by passage through a 0.22 J.Lm filter. Clonidine was made up at 250 J.Lg/ml and infused at 33 lLl/hr (dose rate of 8.3 J.Lg/hr). Preliminary studies indicate that clonidine is stable in solution with no ap-
parent change in potency over several days. In control studies artificial cerebrospinal fluid was infused at the same rate. (See Appendix for composition.) Fetal electrocortical activity, integrated diaphragm and nuchal electromyol~rams, tracheal and arterial pressures, and heart rate were recorded continuously on Grass polygraphs. Data were analyzed from the chart recordings manually. For each preparation steady-state measurements were made of the 10 hours before infusion started ("control"), the infusion period starting 4 hours from the start of infusion and continuing to the end ("infusion"), and the lO-hour period starting with the end of the infusion period ("postinfusion"). Mean heart rates and arterial pressures were measured from the traces twice hourly during control, infusion, and postinfusion periods. In each consecutive 2-hour epoch of these periods, the incidence of fetal breathing and highvoltage electrocortical activity were measured in minutes per hour, and the duration of each episode of breathing and high-voltage electrocortical activity measured. No attempt was made to quantify either low- or intermediate-voltage electrocortical activity. All data were entered into spreadsheets (Lotus 1-2-3) for collation, plotting, and statistical analysis. Blood gas samples were obtained during control, infusion, and postinfusion steady-state conditions. Frequency distributions of some of the data, particularly breathing incidence, were truncated by zero and were therefore clearly nonnormal. Such data cannot be treated by standard parametric statistics and were therefore tested with a nonparametric test (Wilcoxon for unpaired and ranked-pairs test for paired data), and medians are used in these cases instead of means. Results
Artificial cerebrospinal fluid infusions. Recordings from 11 infusions of artificial cerebrospinal fluid in 11 preparations were analyzed. There were no significant differences between values for control, infusion, and postinfusion on breathing or electrocortical activity (Table I). No effects of infusion on heart rate, arterial
Clonidine and fetal breathing
Volume 163 Number 2
CONTROL
CLONIDINE
663
POST-N'USION
Int.
~
ECoG I\lJchal e.m.g.
~t ._._,UI\"_IIiI_ ___~
Diaph'agn
'Ii
__.Jl&..
e.m.g. Tracheal presare
t,'I"'U_~!,\::,,j~\,i.
Key words: Clonidine, fetal breathing, electrocortical activity, lung development
The a.-adrenergic agonist clonidine is used extensively as an antihypertensive agent and has been recommended for use during pregnancy. 1 Although it has no teratogenic effects or acute toxicity, the possibility of more subtle developmental effects on the fetus has not been studied. In adults clonidine reduces arterial pressure by acting at inhibitory presynaptic a.-adrenergic receptors on noradrenergic neurones. This causes hyperpolarization and inhibition of firing2-4 and leads to a reduction in sympathetic outflow and a decrease in arterial pressure. Other effects include sedation and hypothermia. Although these effects of clonidine are well known, the fetal effects of clonidine have attracted little attention. Intravascular injections of 10 j.l.g clonidine caused an immediate arrest offetal breathing,' although clonidine had no consistent effect on breathing postnatally.6 The a.-adrenergic antagonist idazoxan blocked the actions of clonidine, thus confirming its a.-adrenergic action, although when given alone idazoxan had little effect on fetal breathing; presumably the a.-adrenergic receptors were not tonically active, and thus their blockFrom the Department of Pediatrics, University ofMarylnnd Hospital. Supported by research funds from the School of Medicine, University of Marylnnd. Received for publication December 2, 1989; revised March 26, 1990; accepted April 4, 1990. Reprint requests: Owen Bamford, PhD, Division of Neonatology, Department of Pediatrics, University of Marylnnd Hospital, 22 S. Greene St., Baltimore, MD 21201 .
611121411
ade made no difference under normal conditions. Clonidine injections also caused an immediate onset of high-voltage electrocortical activity. Breathing in the fetal lamb is normally suppressed in high-voltage electrocortical activity, but apnea and the shift to highvoltage were shown to be independently mediated since both still occurred after complete brainstem transection. This evidence suggested that in the fetus noradrenergic facilitatory tone is required to maintain breathing, perhaps via a 1 receptors, and could be blocked by activating a.-adrenergic receptors either exogenously with clonidine or by an endogenous neurotransmitter, probably noradrenaline. However, we examined only short-term acute effects of systemic injections of these substances and the responses seen could have been nonspecific effects of a transient high plasma concentrations. To address this possibility and to study the effects of maintained high central concentrations of clonidine we infused clonidine into chronically prepared fetal lambs in utero for periods of up to 24 hours. Because prolonged intravascular infusions of diffusible drugs such as clonidine are likely to produce unpredictable indirect effects on the fetus via changes in the maternal circulation, the agents were administered via an indwelling cannula to a fetal lateral cerebral ventricle for periods up to 24 hours while breathing, nuchal electromyogram, electrocortical activity, arterial pressure, heart rate, and arterial blood gases were recorded.
661
662 Bamford, Hawkins, and Blanco
August 1990 Am J Obstet Gynecol
Table I. Effect of lateral ventricle infusion of cerebrospinal fluid on incidence and episode duration of fetal breathing and electrocortical activity Incidence of high-voltage electrocortical activity (min/hr) Incidence of breathing (min/hr) Duration of high-voltage electrical activity episodes (min) Duration of breathing episodes (min)
Control
Infusion
Postinfusion
24.7 ± 3.5
22.4 ± 5.8
25.5 ± 2.1
30.0 ± 4.8 12.1 ± 3.2
29.9 ± 5.5 ILl ± 2.9
26.2 ± 5.1 12.7 ± 2.7
13.8 ± 4.0
14.1 ± 3.2
12.9 ± 3.2
All comparisons between control, infusion, and postinfusion values were not significant by ranked-pairs test.
Material and methods
Experiments were performed on a total of 15 fetal lambs prepared for chronic recording using methods based on those of Boddy et aU All procedures and protocols were approved by the institutional animal care and use committee. Pregnant Dorset cross ewes at between 120 and 125 days' gestation were anesthetized with intravenous Pentothal, intubated, and ventilated with oxygen and nitrous oxide 2: 1 with 2% halothane. Following normal sterile methods, pairs of stainless steel wire recording electrodes were implanted in the fetal diaphragm, nuchal extensor muscle, and parietal cortex. Vinyl catheters (1 mm inner diameter) were placed in a fetal carotid artery and jugular vein and in the trachea. A fourth catheter ended in the amniotic fluid to act as a pressure reference. A lateral ventricle cannula was placed using a method based on that described by Gluckman and Parsons. 8 An acrylic guide tube was anchored to the skull at 7 mm anterior and 7 mm lateral to bregma with screws and cyanoacrylate glue. A saline-solution-filled cannula made from a 21gauge hypodermic needle bent at right angles was introduced through the guide until it entered a lateral ventricle as shown by the fall in the saline solution column. The cannula was then glued in place in the guide. Vinyl tubing (0.25 mm inner diameter x 1 m) was attached to the exposed end of the cannula via an adapter tube and secured with cyanoacrylate glue. Dead space of catheter and cannula combined was about 100 ILl. The uterus was closed in two layers and catheters and electrodes were exteriorized through a flank incision. Recovery from anesthesia was rapid and the ewes usually stood and fed within an hour of skin closure. Three days were allowed to elapse from the time of operation before any experiments were performed. An infusion rate of 33 lLl/hr was used throughout the study. Drugs were made up in artificial cerebrospinal fluid, adjusted if necessary to pH 6.5 to 7 with carbon dioxide, and sterilized by passage through a 0.22 J.Lm filter. Clonidine was made up at 250 J.Lg/ml and infused at 33 lLl/hr (dose rate of 8.3 J.Lg/hr). Preliminary studies indicate that clonidine is stable in solution with no ap-
parent change in potency over several days. In control studies artificial cerebrospinal fluid was infused at the same rate. (See Appendix for composition.) Fetal electrocortical activity, integrated diaphragm and nuchal electromyol~rams, tracheal and arterial pressures, and heart rate were recorded continuously on Grass polygraphs. Data were analyzed from the chart recordings manually. For each preparation steady-state measurements were made of the 10 hours before infusion started ("control"), the infusion period starting 4 hours from the start of infusion and continuing to the end ("infusion"), and the lO-hour period starting with the end of the infusion period ("postinfusion"). Mean heart rates and arterial pressures were measured from the traces twice hourly during control, infusion, and postinfusion periods. In each consecutive 2-hour epoch of these periods, the incidence of fetal breathing and highvoltage electrocortical activity were measured in minutes per hour, and the duration of each episode of breathing and high-voltage electrocortical activity measured. No attempt was made to quantify either low- or intermediate-voltage electrocortical activity. All data were entered into spreadsheets (Lotus 1-2-3) for collation, plotting, and statistical analysis. Blood gas samples were obtained during control, infusion, and postinfusion steady-state conditions. Frequency distributions of some of the data, particularly breathing incidence, were truncated by zero and were therefore clearly nonnormal. Such data cannot be treated by standard parametric statistics and were therefore tested with a nonparametric test (Wilcoxon for unpaired and ranked-pairs test for paired data), and medians are used in these cases instead of means. Results
Artificial cerebrospinal fluid infusions. Recordings from 11 infusions of artificial cerebrospinal fluid in 11 preparations were analyzed. There were no significant differences between values for control, infusion, and postinfusion on breathing or electrocortical activity (Table I). No effects of infusion on heart rate, arterial
Clonidine and fetal breathing
Volume 163 Number 2
CONTROL
CLONIDINE
663
POST-N'USION
Int.
~
ECoG I\lJchal e.m.g.
~t ._._,UI\"_IIiI_ ___~
Diaph'agn
'Ii
__.Jl&..
e.m.g. Tracheal presare
t,'I"'U_~!,\::,,j~\,i.
