- Email: [email protected]
Pregnancy enhances cocaine's actions on the heart and within the peripheral circulation
Pregnancy enhances cocaine's actions on the heart and within the peripheral circulation James R. Woods, Jr., MD, Kimberly
J. Scott, BS, MT, ASCP, and Mark A. Plessinger, NIS
Rochester,New York OBJECTIVE: Our purpose was to determine whether cocaine's enhanced cardiovascular actions in pregnancy are cardiac alone or involve the peripheral vascular system. STUDY DESIGN: Six pregnant and five nonpregnant ewes chronically instrumented for heart rate, blood pressure, cardiac output, and systemic vascular resistance were given cocaine at 1.0 and 2.0 mg/kg and monitored for 60 minutes. Blood samples for cocaine levels were taken at 5,15,30, and 60 minutes. RESULTS: Cocaine initially (first 60 seconds) produced increased heart rate, decreased cardiac output, decreased stroke volume, and increased cardiac oxygen consumption, which were greater in pregnant than nonpregnant ewes. After 1 minute recovery of cardiac responses was accompanied by increased systemic vascular resistance, which was greater at each dose in pregnant than nonpregnant ewes. Cocaine levels at 5 minutes for pregnant ewes were eightfold to tenfold higher than for nonpregnant ewes. CONCLUSION: Cocaine produces cardiovascular alterations that are dose and time related but, in each case, enhanced in pregnant ewes. Cocaine metabolism may contribute to this pregnancy-related GYNECOL 1994;170:1027-35.) phenomenon. (AmJ OBSTET
Key words: Cocaine, heart, adult, pregnancy, nonpregnancy, cardiovascular,vascular
Our understanding of cocaine's actions on adult cardiovascular function during pregnancy remains incomplete. In part this may be attributed to the fact that primary attention has been focused on cardiovascular and neurologic effects of prenatal cocaine exposure on the developing fetus. That the adult female would incur an enhanced cardiovascular risk from cocaine use while pregnant has recently emerged from studies of the sheep and rat. Indeed, these data indicate that pregnancy enhances the cardiovascular response to cocaine and that increased production of steroid hormones by the placenta may contribute to this enhanced drug action. in the nonpregnant adult cocaine produces an inblood in pressure and heart rate by altering the crease in which the body regulates catecholamines. manner Normally, a nerve impulse releases norepinephrine from presynaptic sites into the synaptic cleft to stimulate a, -postsynaptic receptors as a means of propagat-
From the Division of Maternal-Fetal Medicine, University of Rochester School of Medicine and Dentist7N. Supported in part by National Institute on Drug Abuse grant No. 00415. Presented by in-vitatwn at the Twelfth Annual Meeting of the American Gynecologicaland Obstetrical Societ%Carlsbad, California, September9-11,1993, Reprint requests:James R. Woods, Jr., MD, Strong Memorial Hospital, 601 Elmwood Ave., Box 8668, Rochester, NY 146428668. Copyright (V 1994 by Mosbv-Year Book, Inc. 0002-9378194 $3.60 + 0' 616153990
ing signals. ' Norepinephrine is then removed from the in by the mechanism an uptake synaptic cleft, primarily for in subsequent vesicles and stored area, presynaptic release. Additionally, norepinephrine stimulates %-receptors in the presynaptic area to reduce norepinephfeedback by mechameans of a negative rine release nism. ' By these two mechanisms the concentration of is in regulated and the cleft synaptic norepinephrine Cocontrolled. stimulation of postsynaptic receptors As blocks a consequence, -' this pump. uptake caine in at or cleft the synaptic remain may norepinephrine for in increased a concentrations the nerve terminal Additionally, norepinephrine time. of period prolonged may spill into the vascular system to produce vasoconin resisby receptors activating (x,-adrenergic striction 0-adrenergic by and ctstimulating tance arterioles and heart. in the receptors An enhanced cardiovascular response to cocaine during pregnancy was first observed in sheep. In that study in increase twofold greater pregnant ewes exhibited a blood pressure to the same dose per kilogram of cocaine than was observed in nonpregnant oophorectomized larger doses of ' different to A response pattern ewes. cocaine was also observed. At 5 mg/kg pregnant ewes exhibited seizure activity, cardiac arrhythmias, opisthotonos, respiratory distress, and death, whereas nonpreghypertension and tachycardia. only nant ewes exhibited To produce seizure activity in nonpregnant ewes, co5 15 doses mg/kg were needed. of caine To determine whether elevated hormones of preg-
1027
1028 Woods, Scott, and Messinger
nancy predispose the cardiovascular system to enhanced cocainetoxicity, nonpregnant eweswere administered intravenous cocaine before and then daily during 3 days of progesterone treatment at 10 mg/kg intramuscularly to produce blood progesterone levels comparable to those of late gestation.' In this series norepinephrine was also given intravenously each day to determine whether changesin peripheral adrenergic receptor sensitivity could account for differences in cocaine toxicity noted in pregnant and nonpregnant sheep. The results indicated that during progesterone treatment cocaine produced a twofold greater increase in blood pressure than before progesterone treatment. In contrast, blood pressure responses to norepinephrine were similar before and during progesterone treatment. These data suggested that adrenergic receptor sensitivity to exogenous catecholamines was not involved in this enhanced cocaine effect. To date, these studies have raised serious concerns that pregnancy,through some as yet undefined mechanism, increasesmaternal cardiovascularrisks of cocaine. The purpose of this study was to determine whether increased hypertensive responses to cocaine in pregnancy were attributable to enhanced drug action on the heart itself as measured by heart rate, cardiac output, and myocardial oxygen consumption or a result of changes in peripheral vascular resistance. A second question waswhether alterations in cocaine metabolism could explain these differences in drug action. Methods Instrumentation. Six near-term ewes with singleton fetuses at 120 to 126 days' gestation (term 145 days) and five nonpregnant eweswere selectedfor the study. For surgical instrumentation each ewewasadministered balanced general anesthesia(isoflurane and halothane). Catheterization of the ewe involved placement of polyvinyl catheters in the femoral artery for blood pressure and heart rate monitoring and in the femoral vein for blood sampling and drug administration. Each ewewas then placed on its right side with left upper leg extended, and a thoracotomy incision was made at the fourth intercostal space.The left lung wasretracted, the pericardial sac was incised, and the pulmonary artery flow dissected for artery a pulmonary placement of was probe (Dienco, Los Angeles) for cardiac output measurements. Before pulmonary artery dissection 0.4 mg of atropine was administered to prevent arrhythmias secondaryto vagal stimulation during pulmonary artery dissection. Once the pulmonary artery flow probe was secured,a chest tube was placed for pleural fluid evacuation, the ribs were reapproximated with 2-0 wire, and the pulmonary artery flow probe terminals were brought through a subcutaneoustunnel to a small bag in the ewe's left shoulder. The muscles and skin were then closed in three layers to create a complete seal.
April 1994 Am j Obstet Gynecol
Additional instrumentation of pregnant ewes inThe fetal fetal to viability. assure catheterization volved ewe was repositioned in a supine position, and the abdomen was entered under sterile conditions through incision incision. (2-inch) A was uterine small a vertical then made, allowing exposure of the fetal neck, and a for in the artery carotid polyvinyl catheter was placed blood gas determinations. After closure of the fetal neck incision and uterine incision the fetal catheter and the matemal femoral artery and vein catheters were brought through a subcutaneous tunnel to a bag secured to the ewe's left flank. Oophorectomy was performed in nonpregnant ewes by repositioning them in a supine position. Through a small abdominal incision the nonpregnant uterus was identified and the ovaries clamped and removed and the pedicles ligated. As with the pregnant ewes, femoral vascular catheters were brought subcutaneously to a left the ewe's side. After surgery all animals on pouch day 2 tube removal on postoperative underwent chest and were permitted 5 days recovery from surgery before experiments were begun. Cardiovascular measurements. For blood pressure and heart rate determinations the femoral artery catheter was attached to a pressure transducer (Micron MP-15D) connected to a pressure amplifier (Sensormedic 9853C). Heart rate was recorded by a cardiotachometer (Sensormedic 9857), connected in series with the pressure transducer. Cardiac oxygen consumption was calculated as the product of heart rate x systolic blood pressure. This calculation was carried out to determine whether differences in cardiac oxygen consumption measurements between pregnant and nonpregnant ewes existed during cocaine exposure. For cardiac output determinations pulmonary artery blood flow was measured by a square-wave electromagnetic flow meter (Dienco RF-2100) and medium gain flow (Sensormedic 9853C). Electromagnetic amplifier probes were calibrated with saline solution before each flows linear meathe and of surgery were over range sured. Flow meters were equipped with electronic zeros. The error of electronic flow measurements was < 101/c. Blood pressure, heart rate, and cardiac output were recorded continuously on an eight-channel recorder (Sensormedic R-616). Systemic vascular resistance was recorded as mean arterial pressure -- cardiac output. Cocaine administration. Before cocaine administration each animal was monitored for I hour to establish Cobaseline heart blood pressure. rate and a steady intrave2.0 1.0 mgfkg caine was administered at and nously to each animal. Only one dose was administered each day. Cocaine hydrochloride (Sigma) was adminisby a followed 15 tered as a5 in] volume over seconds, 3 ml saline flush. Blood samples for cocaine analysis five from nonpregthree and pregnant were obtained
Woods, Scott, and Plessinger 1029
Volume 170, Number 4 Am j Obstet Gynecol
and 60 nant ewes. Samples were drawn at 5,15,30, minutes, placed in gray-top tubes containing sodium fluoride and sodium oxalate, and spun, and the serum was frozen at -20' C for later analysis. Cocaine analysis. Blood samples were analyzed for cocaine as previously described. ' Briefly, cocaine was analyzed by a solid-phase extraction technique with 130 mg clean-screen DAU columns (Worldwide Monitoring). Each sample was derivatized with hexfluor 0-2anhydride, evapopropanol and penta-fluoropropionic in dryness toluthe to and residue reconstituted rated ene. The samples were quantified for cocaine with a Hewlett-Packard 5890A gas chromatograph with a nitrogen-phosphorus detector, on the basis of a standard cocaine curve of 0.05,0.25,0.5,1.0,2.0,5.0,10.0,20.0, 50.0, and 100.0 ng/gl. Mepivacaine was used as an internal standard. The recovery yield of cocaine was 100%. Statistical significance consisted of two-way analysis of variance (factors group and time) and post hoc Newman-Keuls test for difference of mean values between groups and magnitude of change from baseline between groups. Statistical significance was set at
120 115
i
106
100 96
90
es
so 75 70
140130120** 110'
p<0.05.
100 90.
Results Baseline measurements for heart rate, blood pressure, cardiac output and systemicvascular resistancefor pregnant and nonpregnant ewesdemonstrate that during pregnancy, heart rate is slightly increased, mean is lower, cardiac output is similar, and arterial pressure systemicvascular resistance is reduced (Figs. I and 2). These characteristicsare consistentwith normal cardiovascular adaptation of pregnancy. Initial response to cocaine. During the first 60 seconds after cocaine injection differing responses in pregnant and nonpregnant eweswere noted. For nonpregnant ewescocaine at 1.0 mg/kg produced an initial rise in blood pressure accompanied by a reflex bradycardia, no change in cardiac output, and a small increasein systemicvascular resistance.In contrast, pregnant ewesgiven 1.0 mg/kg cocaine exhibited tachycardia, a rise in blood pressure, a fall in cardiac output, and an increase in systemicvascular resistance. At 2.0 mg/kg cocaine produced more dramatic differences in cardiovascular response between the two groups of animals. Nonpregnant ewesexhibited little or in blood in heart pressure, no change rate, a rise no change in cardiac output, and a rise in systemicvascular resistance. In contrast, pregnant ewesgiven 2.0 mg/kg cocaine showed marked tachycardia and a rise in blood pressure, a fall in cardiac output, and a rise in systemic vascular resistance.The tachycardia accompanied by a fall in cardiac output strongly suggeststhat at 2.0 mg/kg cocaine produces an initial decreasein stroke volume. Indeed, although a pattern of change in stroke volume
110
80 - wr
111iuu1;
I
r1
I
II"I"III
9
0
2
* 21 19 17 15 13 11 9
'tt1I11'II
7-
BASELtj't'Wý UNE RECTION
ifiH H-ivivo
. 1.1
lpm
6 10 16 30 60 MINUTES SEC=S TIMEAFTER19JECTION
Fig. 1. Cardiovascular responses of pregnant and nonpregValues 1.0 intravenous mg/kg. cocaine, to are nant ewes in differences Significant Asterisk, SEM. mean values t mean between the two groups; arrowhead,significant differences in baseline from between the two groups. magnitude of change
Woods, Scott, and Plessinger
1030
140135V 130. 126120lis. 110. W 106
April 1994 Am j Obstet (; ý necol
NONPFE< PFEGNAf
100 5.
0 90. :076 70
160
I
ISO 140 130 120
100-
go 80 -
VT
9
8
at 1.0 mg/kg was lessclear, at 2.0 mg/kg stroke volume at 30 secondsfell in pregnant ewesby 16%,whereas in nonpregnant ewesstroke volume fell only by 6%. To evaluateoxygen consumption by the heart during the initial 60 secondsof cocaine exposure, the product of heart rate X systolic blood pressure was computed. As can be seen in Fig. 3, baseline cardiac oxygen consumptions by pregnant and nonpregnant eweswere comparable. However, during initial cocaine exposure an increase in oxygen consumption by the heart was dose dependent and was greater in pregnant than in nonpregnant ewes. Subsequent response to cocaine. By I minute cocaine-induced changesin heart rate and cardiac output in pregnant and nonpregnant ewes had returned toward baseline, whereas the hypertensive response persisted in both groups. The increase in blood pressure from baseline in response to cocaine at 1.0 and 2.0 mg/kg, however, was greater in pregnant ewes compared with nonpregnant ewes.Furthermore, the maximum increases in systemic vascular resistance over baseline at cocaine doses of 1.0 and 2.0 mg/kg in the pregnant ewes (maximum 35% and 48%, respectively) were greater than those for nonpregnant ewes (maximum 22% and 35%, respectively). In Table I cocaine levels obtained from both groups are shown. Not unexpectedly, the highest levels occurred at 5 minutes. Cocaine levels at 5 minutes for pregnant ewes at 1.0 or 2.0 mg/kg were significantly higher than were nonpregnant levels.Although cocaine levels declined rapidly in both groups, the values in pregnant ewes were consistently greater during the 60-minute study.
Comment 2
23 2121 fig cc
17' 17 Is 13 11 1, 99 71
-"rF BASEILINE
Mnwrrl"
T"
20 40ý6 0 81700 120 6 10 1'6 30 60 MINUTES SECONDS IRIECTION TIME AFTER IWECTION
Fig. 2. Cardiovascular responsesof pregnant and nonpregnant ewes to intravenous cocaine, 2.0 mg/kg. Values are mean t SEM. Astensh,Significant differences in mean values between the two groups; arrowhead, significant differences in magnitude of change from baseline between the two groups.
The results of our study indicate that intravenous cocaine produces alterations in cardiac ftinction and systemic vascular resistance that are dose and time dependent. Moreover, the magnitude of change in pregnant ewes is greater than that observed in nonpregnant ewes. Cocaine levels meausured in the two groups of animals provide a likely explanation for the differences in cardiovascular responses noted in pregnant and nonpregnant ewes.At 5 minutes cocaine levels after either 1.0 or 2.0 mg/kg were considerably higher in pregnant ewesthan in nonpregnant ewes.Of note, cocaine levels at 5 minutes after 1.0 mWkg are similar to levels reported by Moore et al." in pregnant sheep. Cocaine is metabolized primarily by esterasesand by cytochromic P450enzymes. Studies in pregnant rats have indicated that there is a reduction in liver cytochrome P450 enzymesand monooxygenaselevels.' Such a reduction in cocaine metabolism might explain the elevated cocaine levels in pregnant ewescompared with nonpregnant levels.
Woods, Scott, and Plessinger 1031
Volume 170, Number 4 Am j Obstet Gynecol
17000 ----o-
16000 E E cc x x ClCo 0 F-
NONPREGNANT
PREGNANT
15000 14000 13000 12000 11000 10000 9000
28000 26000
NONPREGNANT
24000
PREGNANT
E E 22000 a: 20000 m x 18000 M co 16000
0
14000
*H
12000
10000 + BASELINE
V-1.111-1
.........
III
ITI
20 40 60 80100120 5 10 15 30 60 MINUTES SE-CrNM INJECTION TIMP AF=TF=PINAFCTION
Fig. 3. Comparison in pregnant and nonpregnant ewes of cardiac oxygen consumption (Nystolic blood pressure[BPJ X Heart rate [HRJ) at 1.0 and 2.0 mg/kg cocaine.Values are mean ± SEM. 7'6P, Responseto intravenous cocaine 1.0 mg/kg; bottom,responseto intravenous cocaine 2.0 mg/kgTable I. Plasma cocaine levels in micrograms per milliliter mg/kg to pregnant and nonpregnant ewes Cocainedose
Pregnancystatus
1.0 mL-/kg
Nonpregnant Pregnant Nonpregnant Pregnant
2.0 mglkg
after cocaine administration 15 min
30 min
60 min
24 :t 7* 135 ± 29 54 ±6 818 t 559
31 t9 106 40 77 26 119 42
17 t4 48 20 18 5 110 48
5 min 74 t 756 ± 215 ± 1975 ±
29* 28 20* 312
Values represent mean ± SEM. *Differences in mean values between pregnant and nonpregnant
Cardiovascular responsesto cocaine in pregnant and nonpregnant ewes differed on the basis of time after exposure. After cocaine injection the most pronounced changes in the first 60 secondswere cardiac in nature. After cocaine administration at both doses tachycardia accompanied by decreasesin cardiac output were observed in pregnant ewes, whereas nonpregnant ewes,
of 1.0 and 2.0
groups.
little dose, higher change exhibited cocaine even at the in heart rate or cardiac output. Ilese changes noted in pregnant ewes are consistent with a decrease in first 60 during the secondsafter cocaine stroke volume exposure. Our data also indicate that, although baseline measurements of cardiac oxygen consumption (systolic
1032
Woods, Scott, and Plessinger
blood pressure x heart rate) for pregnant and nonpregnant ewes are comparable, cocaine initially produces an increase in cardiac oxygen consumption that is greater in pregnant than nonpregnant ewes.Cocaine has been shown to increase myocardial oxygen consumption in unconscious dogs." Under normal conditions, such as exercise, in which myocardial oxygen consumption is increased, an increase in coronary artery blood flow and a decrease in coronary artery vascular resistance occur, reflecting vasodilatation. However, in that study cocaine blocked the coronary artery response. When conscious dogs were administered increasing doses of cocaine, a dose-dependent increase in myocardial oxygen consumption occurred but was not accompanied by a reduction in coronary artery resistance.These authors concluded that cocaine increases myocardial oxygen demand but does not allow vasodilation of coronary arteries to occur during these cardiac changes. In our study no evaluation of coronary artery blood flow was made. Nonetheless, cocaine-induced increases in myocardial oxygen consumption were clearly greater in pregnant than in nonpregnant ewes,were dose dependent, and if associated with a lack of coronary artery vasodilation, may explain the decreased stroke volume noted during initial cocaine exposure. By 60 seconds cocaine actions on cardiac function decreaseas cocaine'seffect to elevate peripheral vascular resistancerapidly emerges. in a previous study we witnesseda greater blood pressure responseto cocaine in pregnant than in nonpregnant ewes.' This enhanced hypertensive response to cocaine was then reproduced in nonpregnant ewesduring progesterone administration. ' Becauseexogenously administered norepinephrine failed to produce these differences, we concluded that cocaine's enhanced actions were not a result of increaseda-adrenergic receptor sensitivity.This conclusion may have been a simplistic interpretation of our data. In the current study cocaine at both doses produced a marked elevation in systemic vascular resistance, which was greater in pregnant than in nonpregnant ewes.The baseline systemicvascular resistancein pregnant ewesis lower than that in nonpregnant ewes. A comparison of maximum percent change from baseline after administration of cocaine at 2.0 mg/kg reveals a 48% increase in systemicvascular resistancein pregnant ewes,whereasnonpregnant ewesexhibited only a 35% increase. These data indicate that the peripheral circulation in pregnancy is more responsiveto cocaine's vasoconstrictiveactions than in the nonpregnant state. Norepinephrine, when given intravenously, does not mimic the effect of enclogenouslyreleased norepinephrine within the synaptic cleft. The effect of exogenous norepinephrine will depend on blood concentrations at oti-adrenergic receptors within blood vessels.The ac-
April 1994 Am j Obstet Gynecol
tions of endogenously released norepinephrine within the synaptic cleft are dependent on central nervous system stimulation and locally within the synaptic cleft by negative feedback to 412adrenergic receptors. In the dog heart inhibition of norepinephrine uptake by desipramine does not lead to elevated norepinephrine concentrations in the synaptic cleft when sympathetic tone is low.' As such, negative feedback by presynaptic: is sufficient to control the synaptic norepi4X2-receptors nephrine concentrations. During increased sympathetic stimulation, however, synaptic norepinephrine concentrations increase during neuronal uptake blockade and cannot be adequately regulated by a2 adrenergic receptor feedback. Pregnancy is associated with increased in " increase As tone. the greater sympathetic such, in to observed cocaine systemic vascular resistance inthan reflect nonpregnant could pregnant ewes creased central stimulation associatedwith pregnancy, leading to a greater increasein norepinephrine concentrations in the synapsethan occurred in nonpregnant ewes.This also suggeststhat synaptic norepinephrine in nonpregnant ewes is adequately regulated by a2receptor feedback.This concept is supported in part by the observation of Moore et al.' that cocaine, when given to pregnant ewes, produced elevated norepinephrine and epinephrine concentrations in the adult circulation, presumably as a reflection of blockade of synaptic norepinephrine uptake and spillover into the circulation. Studies of cocaine's actions in the isolated heart provide some insight into its pharmacologic actions on cardiac function. It should be noted, however, that in this preparation cocaine'seffect on the central nervous system is absent and only local tissfie actions are recorded. When isolated ventricular papillary muscles of the ferret were paced while being exposed to lower cocaine concentrations, cocaine progressively increased cardiac contractility, as indicated by increasing peak tension and rate of relaxation. " Moreover, these changes correlated with increasesin the amplitude of the intracellular Ca* ' transient. That these calciuminduced alterations were catecholamine induced was demonstrated by the fact that propranolol or supramaximum field pulses to exhaust catecholaminerelease attenuated these events. At higher concentrations cocaine produced negative inotropy in papillary muscles,asjudged by a decreasein peak tension and a decreasein measured availability of intracellular Ca* *. This cocaine effect was attributed to its ability to block the sodium channel, thereby interfering with movement of sodium into the sarcoplasm. In the isolated nonpregnant rat heart cocaine appears to produce a similar pattern of contractility response to that of the ferret. " In this preparation, in which paced left ventricular papillary muscles were
Volume 170, Number 4 Am j Obstet Gynecot
exposed to progressively increasing doses of cocaine, positive inotropy resulted from exposure to lower cocaine doses, whereas negative inotropy occurred at higher cocaine concentrations. Pregnancy appears to enhance the cardiodepressantactions of cocaine on the isolated rat heart. In that sameseriesan altered pattern of contractility wasobservedwhen the papillary muscles of the pregnant rats were exposed to the same increasing doses of cocaine. In this group only negative inotropy was seen. Moreover, when nonpregnant rats were administered progesterone intramuscularly to attain late gestational progesterone blood levels, cocaine produced only a negative inotropic response, similar to that of pregnant rats. In a follow-up study the effect of progesterone treatment on cocaine cardiotoxicity was shown to be reversible by blockade of progesterone with mifepristone (RU 486). " These data obtained from the isolated heart suggestthat factors in addition to cocaine concentrations alone influence cardiac dynamics. -fbe adverseimpact of pregnancy on cardiotoxicity of local anesthetics related to cocaine has also been reported. When pregnant and nonpregnant ewes were bupivacaine administered as a bolus injection, bupivacaine produced cardiovascular collapse at significantly lower doses in pregnant ewes than in nonpregnant ewes."' This pattern of response is similar to that we have observed when pregnant and nonpregnant ewes were administered larger cocaine dosesthan were used in the current study.' Progesterone exposure of the rabbit heart also appears to enhance the cardiotoxic actions of certain caines. Moller et al. " measured the rate of depolarization in Purkinje fibers and ventricular musclesfrom rabbits treated with progesterone or vehicle. In oophorectomized rabbits pretreated with progesterone, 30 mg/kg/day for 4 days, bupivacaine produced a greater conduction block, as measured by a decreased rate of depolarization than that which occurred in vehicletreated rabbits. In that study 2 days of progesterone treatment were not sufficient to alter the effect of bupivacaine on rate of cardiac depolarization in the rabbit, a phenomenon that did occur after 4 daysof progesterone (Moller RA. Personal communication). This differential response, based on length of progesterone exposure, suggeststhat chronic but not acute progesterone exposure may alter cocaine cardiotoxicity. The need to better understand how progesterone or pregnancy and cocaine interact is obvious. However, many questions remain. If cocaine has central and peripheral actions, are observationsmade from isolated organ responsesvalid for explaining cocaine'seffects in the intact animal? Likewise, do data from anesthetized animals apply in the unstressed and unanesthetized state? Do different species metabolize cocaine differently? And finally, can data obtained from nonprimates
Woods, Scott, and Plessinger 1033
extrapolate to the clinical setting? Cocaine is an illicit drug whose abuse in the clinical setting is widespread among women of reproductive age. The study of cocaine's actions on the adult heart and cardiovascular system in pregnancy is likely to emerge as an important aspect of the biology of addiction medicine. REFERENCES 1. Lefkowitz Rj, Caron MG, Stiles GL. Mechanisms of membrane-receptor regulation. N Engl J Med 1984; 310: 1570-9. 2. Yamaguchi N, DeCharnplain J, Nadeau RA. Regulation of norepinephrine release from cardiac sympathetic fibers in the dog by presynaptic alpha and beta receptors. Circ Res 1977; 41: 108-17. 3. Wilkerson RD. Cardiovascular toxicity of cocaine. In: Clouet D, Asghar K, Brown R, eds. Mechanisms of cocaine abuse and toxicity. Rockville, Maryland: US Department of Health and Human Services, Alcohol, Drug Abuse, and Mental Health Administration, 1988: 304-24; National Institute on Drug Abuse Monograph 88.
4. Woods JR, Plessinger
MA. Pregnancy increases cardiovasto cocaine. Am J OBSTET GYNECOL 1990; 162:
cular toxicity 529-33.
5. Woods JR, Plessinger MA, Scott K, Miller RK. Prenatal cocaine exposure to the fetus: a sheep model for cardiovascular evaluation. Ann NY Acad Sci 1989; 562: 267-79. 6. Plessinger
increases cardioMA, Woods JR. Progesterone to cocaine in nonpregnant vascular toxicity ewes. A-m J OBsTET GYNECOL 1990; 163: 1659-64.
7.
R. HemodyJ, Miller L, Key TC, Resnik ewe namic effects of intravenous on the pregnant cocaine and fetus. Am J OBsTET GYNECOL 1986; 155: 883-8.
8.
Dean
ME,
drugs
during
Moore
Sorg
TR,
Stock
BH.
Hepatic in
pregnancy
of metabolism Metab Dispos
microsomal the rat. Drug
1975; 3: 325-31. 9.
Cousineau
D,
cardiac
interstitial inhibition
10.
uptake Assali
11.
PeiTeault
NS,
Goresky
CA,
CL.
Rose
norepinephrine in dogs. Girc
basal
Decreased
release after neuronal Res 1986; 58: 859-66.
CK III, et al. Autonomic B, Brinkman in pregin vivo and in vitro control of pelvic circulation: nant and nonpregnant sheep. Am J OBSTET GYNECOL 198 1; 141: 873-84.
effects ferret
Nuwayhid
CL,
Hague
of cocaine ventricular
on
NL,
intracellular
Morgan
ýJ,
Ransil
JP.
'
Ca
The
responsiveness 1990; Br J Pharmacol
myocardium.
of 10 1:
679-85. 12.
Sharma RK,
A,
Plessinger
Woods
JR.
cocaine: role of 1992; 113: 30-5. 13.
Sharma
A,
MA,
Pregnancy
MA,
terone mifepristone antagonist Proc Soc diotoxicity of cocaine. 279-87. HO,
Pedersen
RK,
Miller
H,
Woods
(RU
486)
Exp,
Biol
Finster
Med
15.
in pregnant toxicity and nonpregnant ogy 1985; 63: 134-9. S, Fox J, Johnson M, Covino Moller RA, Datta the cardiac liclocaine. and
JR.
Proges-
1993;
car202:
M, et al. Bupivacaine ewes. Anesthesiol-
Morishima
on
Miller
decreases
14.
of progesterone of bupivacaine
CS,
cardiotoxicity of App] Pharmacol
enhances Toxicol
progesterone.
Plessinger
Liang
DM,
Sherer
electrophysiologic Anesthesiology
BG.
Effects
actions 1992; 76:
604-8.
Discussion DP- Guy M. HA"ERT, JIL, Charlottesville, Virginia. For the past 6 or so years Woods et al. have enlightened us in a progressive manner on the cardiotoxic effects of In cocaine in the ovine experimental preparation.
1034
April 1994 Am j Obstet Gynecol
Woods, Scott, and Plessinger
1987, using a protocol similar to that presented today, they' reported that intravenous administration of cocaine to pregnant ewes produced dose-dependent increasesin maternal mean arterial blood pressure. They also observeda marked reduction in uterine blood flow, which they attributed to a dramatic increasein vasoconstriction of the uterine arteries. In 1988 they' reported that the response of maternal mean arterial blood pressure was maximum by I minute and returned to baseline levels by 30 minutes after injection. In 1990 they' expanded the measurement parameters to include systolic and diastolic blood pressure and performed similar studies in nonpregnant oophorectomized ewes.The observation that pregnant ewesdemonstrated significantly greater absolute increases in blood pressureresponsecompared with those observed in the castrated subjects led Woods et al. to conclude that pregnancy increasesthe cardiotoxicity of cocaine. This leads to my first question. Is the response to cocaine of the oophorectomized and the nonpregnant ewe with intact ovaries the same? In this study Dr. Woods has confirmed his observations that the magnitude of cardiac response to cocaine is greater in pregnant than in nonpregnant, oophorectomized animals. He has reconfirmed his earlier reports that cocaine produces increasesin systolic, diastolic, and mean arterial blood pressures that are dose and time related. He also reiterated the findings reported in 19882that 1.0 mg/kg dosages of cocaine produce a mild bradycardia but 2.0 mg/kg produces a transient tachycardia. However, he does not present statistics to indicate that these observed trends in heart rate are significantly different from baseline values.Consequently,my secondquestion is what is the basis for the conclusion that cocaine produces a doseand time-related increase in heart rate? The observationsthat cardiac oxygen consumption is increased in both pregnant and nonpregnant animals are consistentwith the oxygen consumption values at I and 2 minutes after cocaine injection, which can be readily calculated from data Woods and Plessinger' published in 1990. The figures presented today have extremely wide bars representing the SEM. My third question is, becausein actuality each animal could serve as its own control, why did you choose nonparametric tests to measure statistical significance? The new information presented today is data related to systemic vascular resistance and the measure of cardiac output. These data are of interest in that at dosages of 1.0 and 2.0 mg/kg cocaine produced a maximum response in systemicvascular resistancethat was 35% and 78% over baseline, respectively. Earlier Woods et al. ' reported that the maximum increase in uterine vascular resistance to similar cocaine dosages was 98% and 168%. Does this indicate that the uterine vasculaturehas an enhanced response to cocaine compared with the systemic circulation? This question is asked more in speculation and as possible grounds for future investigation than in anticipation of a definitive answerat this time.
Throughout his presentations Dr. Woods has reported that the cardotoxic effects of cocaine are greater in pregnant than in nonpregnant, oophorectornized preparations. However,there is evidenceto suggestthat any differences in response may be more related to drug dosagesand plasma concentrations than to alterations in response mechanisms. First, cocaine was administered on a milligram per kilogram of body weight basis. In an earlier publication' Dr. Woods reported that the mean weight of pregnant animals was 15% greater than the nonpregnant, oophorectomized ones. Becausedosage is based on body weight and administered over 15 secondsthrough a catheter placed in the femoral vein, the amount of cocaine to which the heart of the pregnant animal would be exposed on first pass would be greater. Second, the plasma cocaine concentrations in pregnant eweswere consistently greater than those measured in the nonpregnant, oophorectomized animals. This great disparity, plus the large SENIs,leads me to my next three questions. (1) What is the specificity and reproducibility of your plasma cocaine measurements?(2) Do we have any idea of what proportion of the plasma levels is active and what proportion, if any, is inactive? and (3) Have you performed pharmacokinetic studies to see if there is a difference in the half-life of cocainebetweenpregnant and nonpregnant, oophorectomized sheep? REFERENCES
1. Woods JR, Plessinger MA, Clark KE. Effect of cocaine on uterine blood flow and fetal oxygenation. JAMA 1987; 257: 957-61. 2. Woods JR, Plessinger MA, Scott K, Miller RK. Prenatal cocaine exposure to the fetus: a sheep model for cardiovascular evaluation. Ann NY Acad Sci 1988; 562: 267-79. 3. Woods JR, Plessinger MA. Pregnancy increases cardiovascular toxicity to cocaine. Am j ORSrET GYNECOL 1990; 162; 529-33.
D& WOODS(Closing). The question was whether the response to cocaine in nonpregnant, oophorectomized ewes was the same as that in those nonpregnant ewes that were not oophorectomized. We have not repeated the studies in nonpregnant ewes, leaving the ovaries intact. The purpose of oophorectomy, of course, was that we were trying to eliminate the possibility of hormonal flux. We preferred to have a clean system where we could address specifically the issue of progesterone and pregnancy in an intact model. Your question regarding the wide variability of cardiovascular measurements during initial cocaine exposure is an appropriate one. There's no question that during the first 60 seconds after cocaine administration there are a number of events that are occurring, and they are occurring very rapidly. When cardiovascular measurements were obtained at 5,15,30 and 60 minutes, at least four measurements were made over a 30-second period and then averaged. During the initial exposure we obtained measurements every 10 seconds and therefore could only take one measurement at any time point. During this acute
Woods, Scott, and Plessinger 1035
Volume 170, Number 4 Am j Obstet Gynecol
exposure period the animals are often quite agitated; this in part contributes to our observations of the wider variability. Nonetheless, we now know that many of the acute, and probably life-threatening, cardiac events to cocaine appear within the first 60 seconds of exposure. I might add that our study was based on the premise that cocaine-induced enhanced cardiac performance would explain the elevated hypertensive response seen in pregnant and progesterone-treated, nonpregnant ewes. We feel that we have effectively shown to our satisfaction that the augmented hypertensive response to cocaine in pregnant ewes is not driven by increased cardiac output secondary to tachycardia but, in fact, represents a primary event in the peripheral circulation. Dr. Harbert asked whether the augmented response to cocaine in pregnant ewes was not simply based on higher cocaine concentrations in the pregnant ewe's circulation. Cocaine was given to pregnant and nonpregnant ewes in similar concentrations on the basis of total body weight. If the total body weight of the pregnant ewe included fetus, amniotic fluid, and placenta, we would be giving a higher dose of cocaine to the pregnant ewes than to the nonpregnant ewes. That issue inspired us to administer progesterone to very nonpregnant animals. We believed that if we could demonstrate the same augmented cocaine effect with nonpregnant progesterone-treated ewes as we did in the pregnant ewes, then it would be very unlikely that
cocainedistribution in the adult circulation was responsible for these differences. It is possible, however, that metabolism may play a significant role in altering cocaine actions in pregnant ewes. We have not compared elimination half-life measurements for cocaine in pregnant ewesversusnonpregnant ewes.Others have examined this factor but not during direct comparison of half-life in the pregnant and nonpregnant states. Cocaine's elimination half-life is dose dependent, and in the nonpregnant human it is 40 to 82 minutes,' whereasin the pregnant guinea pig it is 20 to 67 rninuteS2and in the pregnant sheep 4.4 to 5 minutes.' It is possible that conversion of cocaine to some of the more active metabolites may be important and may need to be further evaluated. Dr. Harbert asked about the reproduceability in our cocaine assays.We have not split samplesto run simultaneous measurementsto determine that. However,our minimal level of detection for cocaine is 5 ng/ml.
REFERENCES
1. Barnett G, Hawk R, Resnik R. Cocainepharmacokineticsin humans.j Ethnopharmacol 1981;3:353-66. 2. Sanberg JA, Olsen GD. Cocaine pharmacokinetics in the pregnant guinea pig. Pharmacol Exp Ther 1991;258:477.1 81. 3. De Vane CL, Burchfield Dj, Abrams RM, et al. Disposition of cocaine in pregnant sheep. Dev Pharmacol Ther 1991; 16:123-9.
J. Scott, BS, MT, ASCP, and Mark A. Plessinger, NIS
Rochester,New York OBJECTIVE: Our purpose was to determine whether cocaine's enhanced cardiovascular actions in pregnancy are cardiac alone or involve the peripheral vascular system. STUDY DESIGN: Six pregnant and five nonpregnant ewes chronically instrumented for heart rate, blood pressure, cardiac output, and systemic vascular resistance were given cocaine at 1.0 and 2.0 mg/kg and monitored for 60 minutes. Blood samples for cocaine levels were taken at 5,15,30, and 60 minutes. RESULTS: Cocaine initially (first 60 seconds) produced increased heart rate, decreased cardiac output, decreased stroke volume, and increased cardiac oxygen consumption, which were greater in pregnant than nonpregnant ewes. After 1 minute recovery of cardiac responses was accompanied by increased systemic vascular resistance, which was greater at each dose in pregnant than nonpregnant ewes. Cocaine levels at 5 minutes for pregnant ewes were eightfold to tenfold higher than for nonpregnant ewes. CONCLUSION: Cocaine produces cardiovascular alterations that are dose and time related but, in each case, enhanced in pregnant ewes. Cocaine metabolism may contribute to this pregnancy-related GYNECOL 1994;170:1027-35.) phenomenon. (AmJ OBSTET
Key words: Cocaine, heart, adult, pregnancy, nonpregnancy, cardiovascular,vascular
Our understanding of cocaine's actions on adult cardiovascular function during pregnancy remains incomplete. In part this may be attributed to the fact that primary attention has been focused on cardiovascular and neurologic effects of prenatal cocaine exposure on the developing fetus. That the adult female would incur an enhanced cardiovascular risk from cocaine use while pregnant has recently emerged from studies of the sheep and rat. Indeed, these data indicate that pregnancy enhances the cardiovascular response to cocaine and that increased production of steroid hormones by the placenta may contribute to this enhanced drug action. in the nonpregnant adult cocaine produces an inblood in pressure and heart rate by altering the crease in which the body regulates catecholamines. manner Normally, a nerve impulse releases norepinephrine from presynaptic sites into the synaptic cleft to stimulate a, -postsynaptic receptors as a means of propagat-
From the Division of Maternal-Fetal Medicine, University of Rochester School of Medicine and Dentist7N. Supported in part by National Institute on Drug Abuse grant No. 00415. Presented by in-vitatwn at the Twelfth Annual Meeting of the American Gynecologicaland Obstetrical Societ%Carlsbad, California, September9-11,1993, Reprint requests:James R. Woods, Jr., MD, Strong Memorial Hospital, 601 Elmwood Ave., Box 8668, Rochester, NY 146428668. Copyright (V 1994 by Mosbv-Year Book, Inc. 0002-9378194 $3.60 + 0' 616153990
ing signals. ' Norepinephrine is then removed from the in by the mechanism an uptake synaptic cleft, primarily for in subsequent vesicles and stored area, presynaptic release. Additionally, norepinephrine stimulates %-receptors in the presynaptic area to reduce norepinephfeedback by mechameans of a negative rine release nism. ' By these two mechanisms the concentration of is in regulated and the cleft synaptic norepinephrine Cocontrolled. stimulation of postsynaptic receptors As blocks a consequence, -' this pump. uptake caine in at or cleft the synaptic remain may norepinephrine for in increased a concentrations the nerve terminal Additionally, norepinephrine time. of period prolonged may spill into the vascular system to produce vasoconin resisby receptors activating (x,-adrenergic striction 0-adrenergic by and ctstimulating tance arterioles and heart. in the receptors An enhanced cardiovascular response to cocaine during pregnancy was first observed in sheep. In that study in increase twofold greater pregnant ewes exhibited a blood pressure to the same dose per kilogram of cocaine than was observed in nonpregnant oophorectomized larger doses of ' different to A response pattern ewes. cocaine was also observed. At 5 mg/kg pregnant ewes exhibited seizure activity, cardiac arrhythmias, opisthotonos, respiratory distress, and death, whereas nonpreghypertension and tachycardia. only nant ewes exhibited To produce seizure activity in nonpregnant ewes, co5 15 doses mg/kg were needed. of caine To determine whether elevated hormones of preg-
1027
1028 Woods, Scott, and Messinger
nancy predispose the cardiovascular system to enhanced cocainetoxicity, nonpregnant eweswere administered intravenous cocaine before and then daily during 3 days of progesterone treatment at 10 mg/kg intramuscularly to produce blood progesterone levels comparable to those of late gestation.' In this series norepinephrine was also given intravenously each day to determine whether changesin peripheral adrenergic receptor sensitivity could account for differences in cocaine toxicity noted in pregnant and nonpregnant sheep. The results indicated that during progesterone treatment cocaine produced a twofold greater increase in blood pressure than before progesterone treatment. In contrast, blood pressure responses to norepinephrine were similar before and during progesterone treatment. These data suggested that adrenergic receptor sensitivity to exogenous catecholamines was not involved in this enhanced cocaine effect. To date, these studies have raised serious concerns that pregnancy,through some as yet undefined mechanism, increasesmaternal cardiovascularrisks of cocaine. The purpose of this study was to determine whether increased hypertensive responses to cocaine in pregnancy were attributable to enhanced drug action on the heart itself as measured by heart rate, cardiac output, and myocardial oxygen consumption or a result of changes in peripheral vascular resistance. A second question waswhether alterations in cocaine metabolism could explain these differences in drug action. Methods Instrumentation. Six near-term ewes with singleton fetuses at 120 to 126 days' gestation (term 145 days) and five nonpregnant eweswere selectedfor the study. For surgical instrumentation each ewewasadministered balanced general anesthesia(isoflurane and halothane). Catheterization of the ewe involved placement of polyvinyl catheters in the femoral artery for blood pressure and heart rate monitoring and in the femoral vein for blood sampling and drug administration. Each ewewas then placed on its right side with left upper leg extended, and a thoracotomy incision was made at the fourth intercostal space.The left lung wasretracted, the pericardial sac was incised, and the pulmonary artery flow dissected for artery a pulmonary placement of was probe (Dienco, Los Angeles) for cardiac output measurements. Before pulmonary artery dissection 0.4 mg of atropine was administered to prevent arrhythmias secondaryto vagal stimulation during pulmonary artery dissection. Once the pulmonary artery flow probe was secured,a chest tube was placed for pleural fluid evacuation, the ribs were reapproximated with 2-0 wire, and the pulmonary artery flow probe terminals were brought through a subcutaneoustunnel to a small bag in the ewe's left shoulder. The muscles and skin were then closed in three layers to create a complete seal.
April 1994 Am j Obstet Gynecol
Additional instrumentation of pregnant ewes inThe fetal fetal to viability. assure catheterization volved ewe was repositioned in a supine position, and the abdomen was entered under sterile conditions through incision incision. (2-inch) A was uterine small a vertical then made, allowing exposure of the fetal neck, and a for in the artery carotid polyvinyl catheter was placed blood gas determinations. After closure of the fetal neck incision and uterine incision the fetal catheter and the matemal femoral artery and vein catheters were brought through a subcutaneous tunnel to a bag secured to the ewe's left flank. Oophorectomy was performed in nonpregnant ewes by repositioning them in a supine position. Through a small abdominal incision the nonpregnant uterus was identified and the ovaries clamped and removed and the pedicles ligated. As with the pregnant ewes, femoral vascular catheters were brought subcutaneously to a left the ewe's side. After surgery all animals on pouch day 2 tube removal on postoperative underwent chest and were permitted 5 days recovery from surgery before experiments were begun. Cardiovascular measurements. For blood pressure and heart rate determinations the femoral artery catheter was attached to a pressure transducer (Micron MP-15D) connected to a pressure amplifier (Sensormedic 9853C). Heart rate was recorded by a cardiotachometer (Sensormedic 9857), connected in series with the pressure transducer. Cardiac oxygen consumption was calculated as the product of heart rate x systolic blood pressure. This calculation was carried out to determine whether differences in cardiac oxygen consumption measurements between pregnant and nonpregnant ewes existed during cocaine exposure. For cardiac output determinations pulmonary artery blood flow was measured by a square-wave electromagnetic flow meter (Dienco RF-2100) and medium gain flow (Sensormedic 9853C). Electromagnetic amplifier probes were calibrated with saline solution before each flows linear meathe and of surgery were over range sured. Flow meters were equipped with electronic zeros. The error of electronic flow measurements was < 101/c. Blood pressure, heart rate, and cardiac output were recorded continuously on an eight-channel recorder (Sensormedic R-616). Systemic vascular resistance was recorded as mean arterial pressure -- cardiac output. Cocaine administration. Before cocaine administration each animal was monitored for I hour to establish Cobaseline heart blood pressure. rate and a steady intrave2.0 1.0 mgfkg caine was administered at and nously to each animal. Only one dose was administered each day. Cocaine hydrochloride (Sigma) was adminisby a followed 15 tered as a5 in] volume over seconds, 3 ml saline flush. Blood samples for cocaine analysis five from nonpregthree and pregnant were obtained
Woods, Scott, and Plessinger 1029
Volume 170, Number 4 Am j Obstet Gynecol
and 60 nant ewes. Samples were drawn at 5,15,30, minutes, placed in gray-top tubes containing sodium fluoride and sodium oxalate, and spun, and the serum was frozen at -20' C for later analysis. Cocaine analysis. Blood samples were analyzed for cocaine as previously described. ' Briefly, cocaine was analyzed by a solid-phase extraction technique with 130 mg clean-screen DAU columns (Worldwide Monitoring). Each sample was derivatized with hexfluor 0-2anhydride, evapopropanol and penta-fluoropropionic in dryness toluthe to and residue reconstituted rated ene. The samples were quantified for cocaine with a Hewlett-Packard 5890A gas chromatograph with a nitrogen-phosphorus detector, on the basis of a standard cocaine curve of 0.05,0.25,0.5,1.0,2.0,5.0,10.0,20.0, 50.0, and 100.0 ng/gl. Mepivacaine was used as an internal standard. The recovery yield of cocaine was 100%. Statistical significance consisted of two-way analysis of variance (factors group and time) and post hoc Newman-Keuls test for difference of mean values between groups and magnitude of change from baseline between groups. Statistical significance was set at
120 115
i
106
100 96
90
es
so 75 70
140130120** 110'
p<0.05.
100 90.
Results Baseline measurements for heart rate, blood pressure, cardiac output and systemicvascular resistancefor pregnant and nonpregnant ewesdemonstrate that during pregnancy, heart rate is slightly increased, mean is lower, cardiac output is similar, and arterial pressure systemicvascular resistance is reduced (Figs. I and 2). These characteristicsare consistentwith normal cardiovascular adaptation of pregnancy. Initial response to cocaine. During the first 60 seconds after cocaine injection differing responses in pregnant and nonpregnant eweswere noted. For nonpregnant ewescocaine at 1.0 mg/kg produced an initial rise in blood pressure accompanied by a reflex bradycardia, no change in cardiac output, and a small increasein systemicvascular resistance.In contrast, pregnant ewesgiven 1.0 mg/kg cocaine exhibited tachycardia, a rise in blood pressure, a fall in cardiac output, and an increase in systemicvascular resistance. At 2.0 mg/kg cocaine produced more dramatic differences in cardiovascular response between the two groups of animals. Nonpregnant ewesexhibited little or in blood in heart pressure, no change rate, a rise no change in cardiac output, and a rise in systemicvascular resistance. In contrast, pregnant ewesgiven 2.0 mg/kg cocaine showed marked tachycardia and a rise in blood pressure, a fall in cardiac output, and a rise in systemic vascular resistance.The tachycardia accompanied by a fall in cardiac output strongly suggeststhat at 2.0 mg/kg cocaine produces an initial decreasein stroke volume. Indeed, although a pattern of change in stroke volume
110
80 - wr
111iuu1;
I
r1
I
II"I"III
9
0
2
* 21 19 17 15 13 11 9
'tt1I11'II
7-
BASELtj't'Wý UNE RECTION
ifiH H-ivivo
. 1.1
lpm
6 10 16 30 60 MINUTES SEC=S TIMEAFTER19JECTION
Fig. 1. Cardiovascular responses of pregnant and nonpregValues 1.0 intravenous mg/kg. cocaine, to are nant ewes in differences Significant Asterisk, SEM. mean values t mean between the two groups; arrowhead,significant differences in baseline from between the two groups. magnitude of change
Woods, Scott, and Plessinger
1030
140135V 130. 126120lis. 110. W 106
April 1994 Am j Obstet (; ý necol
NONPFE< PFEGNAf
100 5.
0 90. :076 70
160
I
ISO 140 130 120
100-
go 80 -
VT
9
8
at 1.0 mg/kg was lessclear, at 2.0 mg/kg stroke volume at 30 secondsfell in pregnant ewesby 16%,whereas in nonpregnant ewesstroke volume fell only by 6%. To evaluateoxygen consumption by the heart during the initial 60 secondsof cocaine exposure, the product of heart rate X systolic blood pressure was computed. As can be seen in Fig. 3, baseline cardiac oxygen consumptions by pregnant and nonpregnant eweswere comparable. However, during initial cocaine exposure an increase in oxygen consumption by the heart was dose dependent and was greater in pregnant than in nonpregnant ewes. Subsequent response to cocaine. By I minute cocaine-induced changesin heart rate and cardiac output in pregnant and nonpregnant ewes had returned toward baseline, whereas the hypertensive response persisted in both groups. The increase in blood pressure from baseline in response to cocaine at 1.0 and 2.0 mg/kg, however, was greater in pregnant ewes compared with nonpregnant ewes.Furthermore, the maximum increases in systemic vascular resistance over baseline at cocaine doses of 1.0 and 2.0 mg/kg in the pregnant ewes (maximum 35% and 48%, respectively) were greater than those for nonpregnant ewes (maximum 22% and 35%, respectively). In Table I cocaine levels obtained from both groups are shown. Not unexpectedly, the highest levels occurred at 5 minutes. Cocaine levels at 5 minutes for pregnant ewes at 1.0 or 2.0 mg/kg were significantly higher than were nonpregnant levels.Although cocaine levels declined rapidly in both groups, the values in pregnant ewes were consistently greater during the 60-minute study.
Comment 2
23 2121 fig cc
17' 17 Is 13 11 1, 99 71
-"rF BASEILINE
Mnwrrl"
T"
20 40ý6 0 81700 120 6 10 1'6 30 60 MINUTES SECONDS IRIECTION TIME AFTER IWECTION
Fig. 2. Cardiovascular responsesof pregnant and nonpregnant ewes to intravenous cocaine, 2.0 mg/kg. Values are mean t SEM. Astensh,Significant differences in mean values between the two groups; arrowhead, significant differences in magnitude of change from baseline between the two groups.
The results of our study indicate that intravenous cocaine produces alterations in cardiac ftinction and systemic vascular resistance that are dose and time dependent. Moreover, the magnitude of change in pregnant ewes is greater than that observed in nonpregnant ewes. Cocaine levels meausured in the two groups of animals provide a likely explanation for the differences in cardiovascular responses noted in pregnant and nonpregnant ewes.At 5 minutes cocaine levels after either 1.0 or 2.0 mg/kg were considerably higher in pregnant ewesthan in nonpregnant ewes.Of note, cocaine levels at 5 minutes after 1.0 mWkg are similar to levels reported by Moore et al." in pregnant sheep. Cocaine is metabolized primarily by esterasesand by cytochromic P450enzymes. Studies in pregnant rats have indicated that there is a reduction in liver cytochrome P450 enzymesand monooxygenaselevels.' Such a reduction in cocaine metabolism might explain the elevated cocaine levels in pregnant ewescompared with nonpregnant levels.
Woods, Scott, and Plessinger 1031
Volume 170, Number 4 Am j Obstet Gynecol
17000 ----o-
16000 E E cc x x ClCo 0 F-
NONPREGNANT
PREGNANT
15000 14000 13000 12000 11000 10000 9000
28000 26000
NONPREGNANT
24000
PREGNANT
E E 22000 a: 20000 m x 18000 M co 16000
0
14000
*H
12000
10000 + BASELINE
V-1.111-1
.........
III
ITI
20 40 60 80100120 5 10 15 30 60 MINUTES SE-CrNM INJECTION TIMP AF=TF=PINAFCTION
Fig. 3. Comparison in pregnant and nonpregnant ewes of cardiac oxygen consumption (Nystolic blood pressure[BPJ X Heart rate [HRJ) at 1.0 and 2.0 mg/kg cocaine.Values are mean ± SEM. 7'6P, Responseto intravenous cocaine 1.0 mg/kg; bottom,responseto intravenous cocaine 2.0 mg/kgTable I. Plasma cocaine levels in micrograms per milliliter mg/kg to pregnant and nonpregnant ewes Cocainedose
Pregnancystatus
1.0 mL-/kg
Nonpregnant Pregnant Nonpregnant Pregnant
2.0 mglkg
after cocaine administration 15 min
30 min
60 min
24 :t 7* 135 ± 29 54 ±6 818 t 559
31 t9 106 40 77 26 119 42
17 t4 48 20 18 5 110 48
5 min 74 t 756 ± 215 ± 1975 ±
29* 28 20* 312
Values represent mean ± SEM. *Differences in mean values between pregnant and nonpregnant
Cardiovascular responsesto cocaine in pregnant and nonpregnant ewes differed on the basis of time after exposure. After cocaine injection the most pronounced changes in the first 60 secondswere cardiac in nature. After cocaine administration at both doses tachycardia accompanied by decreasesin cardiac output were observed in pregnant ewes, whereas nonpregnant ewes,
of 1.0 and 2.0
groups.
little dose, higher change exhibited cocaine even at the in heart rate or cardiac output. Ilese changes noted in pregnant ewes are consistent with a decrease in first 60 during the secondsafter cocaine stroke volume exposure. Our data also indicate that, although baseline measurements of cardiac oxygen consumption (systolic
1032
Woods, Scott, and Plessinger
blood pressure x heart rate) for pregnant and nonpregnant ewes are comparable, cocaine initially produces an increase in cardiac oxygen consumption that is greater in pregnant than nonpregnant ewes.Cocaine has been shown to increase myocardial oxygen consumption in unconscious dogs." Under normal conditions, such as exercise, in which myocardial oxygen consumption is increased, an increase in coronary artery blood flow and a decrease in coronary artery vascular resistance occur, reflecting vasodilatation. However, in that study cocaine blocked the coronary artery response. When conscious dogs were administered increasing doses of cocaine, a dose-dependent increase in myocardial oxygen consumption occurred but was not accompanied by a reduction in coronary artery resistance.These authors concluded that cocaine increases myocardial oxygen demand but does not allow vasodilation of coronary arteries to occur during these cardiac changes. In our study no evaluation of coronary artery blood flow was made. Nonetheless, cocaine-induced increases in myocardial oxygen consumption were clearly greater in pregnant than in nonpregnant ewes,were dose dependent, and if associated with a lack of coronary artery vasodilation, may explain the decreased stroke volume noted during initial cocaine exposure. By 60 seconds cocaine actions on cardiac function decreaseas cocaine'seffect to elevate peripheral vascular resistancerapidly emerges. in a previous study we witnesseda greater blood pressure responseto cocaine in pregnant than in nonpregnant ewes.' This enhanced hypertensive response to cocaine was then reproduced in nonpregnant ewesduring progesterone administration. ' Becauseexogenously administered norepinephrine failed to produce these differences, we concluded that cocaine's enhanced actions were not a result of increaseda-adrenergic receptor sensitivity.This conclusion may have been a simplistic interpretation of our data. In the current study cocaine at both doses produced a marked elevation in systemic vascular resistance, which was greater in pregnant than in nonpregnant ewes.The baseline systemicvascular resistancein pregnant ewesis lower than that in nonpregnant ewes. A comparison of maximum percent change from baseline after administration of cocaine at 2.0 mg/kg reveals a 48% increase in systemicvascular resistancein pregnant ewes,whereasnonpregnant ewesexhibited only a 35% increase. These data indicate that the peripheral circulation in pregnancy is more responsiveto cocaine's vasoconstrictiveactions than in the nonpregnant state. Norepinephrine, when given intravenously, does not mimic the effect of enclogenouslyreleased norepinephrine within the synaptic cleft. The effect of exogenous norepinephrine will depend on blood concentrations at oti-adrenergic receptors within blood vessels.The ac-
April 1994 Am j Obstet Gynecol
tions of endogenously released norepinephrine within the synaptic cleft are dependent on central nervous system stimulation and locally within the synaptic cleft by negative feedback to 412adrenergic receptors. In the dog heart inhibition of norepinephrine uptake by desipramine does not lead to elevated norepinephrine concentrations in the synaptic cleft when sympathetic tone is low.' As such, negative feedback by presynaptic: is sufficient to control the synaptic norepi4X2-receptors nephrine concentrations. During increased sympathetic stimulation, however, synaptic norepinephrine concentrations increase during neuronal uptake blockade and cannot be adequately regulated by a2 adrenergic receptor feedback. Pregnancy is associated with increased in " increase As tone. the greater sympathetic such, in to observed cocaine systemic vascular resistance inthan reflect nonpregnant could pregnant ewes creased central stimulation associatedwith pregnancy, leading to a greater increasein norepinephrine concentrations in the synapsethan occurred in nonpregnant ewes.This also suggeststhat synaptic norepinephrine in nonpregnant ewes is adequately regulated by a2receptor feedback.This concept is supported in part by the observation of Moore et al.' that cocaine, when given to pregnant ewes, produced elevated norepinephrine and epinephrine concentrations in the adult circulation, presumably as a reflection of blockade of synaptic norepinephrine uptake and spillover into the circulation. Studies of cocaine's actions in the isolated heart provide some insight into its pharmacologic actions on cardiac function. It should be noted, however, that in this preparation cocaine'seffect on the central nervous system is absent and only local tissfie actions are recorded. When isolated ventricular papillary muscles of the ferret were paced while being exposed to lower cocaine concentrations, cocaine progressively increased cardiac contractility, as indicated by increasing peak tension and rate of relaxation. " Moreover, these changes correlated with increasesin the amplitude of the intracellular Ca* ' transient. That these calciuminduced alterations were catecholamine induced was demonstrated by the fact that propranolol or supramaximum field pulses to exhaust catecholaminerelease attenuated these events. At higher concentrations cocaine produced negative inotropy in papillary muscles,asjudged by a decreasein peak tension and a decreasein measured availability of intracellular Ca* *. This cocaine effect was attributed to its ability to block the sodium channel, thereby interfering with movement of sodium into the sarcoplasm. In the isolated nonpregnant rat heart cocaine appears to produce a similar pattern of contractility response to that of the ferret. " In this preparation, in which paced left ventricular papillary muscles were
Volume 170, Number 4 Am j Obstet Gynecot
exposed to progressively increasing doses of cocaine, positive inotropy resulted from exposure to lower cocaine doses, whereas negative inotropy occurred at higher cocaine concentrations. Pregnancy appears to enhance the cardiodepressantactions of cocaine on the isolated rat heart. In that sameseriesan altered pattern of contractility wasobservedwhen the papillary muscles of the pregnant rats were exposed to the same increasing doses of cocaine. In this group only negative inotropy was seen. Moreover, when nonpregnant rats were administered progesterone intramuscularly to attain late gestational progesterone blood levels, cocaine produced only a negative inotropic response, similar to that of pregnant rats. In a follow-up study the effect of progesterone treatment on cocaine cardiotoxicity was shown to be reversible by blockade of progesterone with mifepristone (RU 486). " These data obtained from the isolated heart suggestthat factors in addition to cocaine concentrations alone influence cardiac dynamics. -fbe adverseimpact of pregnancy on cardiotoxicity of local anesthetics related to cocaine has also been reported. When pregnant and nonpregnant ewes were bupivacaine administered as a bolus injection, bupivacaine produced cardiovascular collapse at significantly lower doses in pregnant ewes than in nonpregnant ewes."' This pattern of response is similar to that we have observed when pregnant and nonpregnant ewes were administered larger cocaine dosesthan were used in the current study.' Progesterone exposure of the rabbit heart also appears to enhance the cardiotoxic actions of certain caines. Moller et al. " measured the rate of depolarization in Purkinje fibers and ventricular musclesfrom rabbits treated with progesterone or vehicle. In oophorectomized rabbits pretreated with progesterone, 30 mg/kg/day for 4 days, bupivacaine produced a greater conduction block, as measured by a decreased rate of depolarization than that which occurred in vehicletreated rabbits. In that study 2 days of progesterone treatment were not sufficient to alter the effect of bupivacaine on rate of cardiac depolarization in the rabbit, a phenomenon that did occur after 4 daysof progesterone (Moller RA. Personal communication). This differential response, based on length of progesterone exposure, suggeststhat chronic but not acute progesterone exposure may alter cocaine cardiotoxicity. The need to better understand how progesterone or pregnancy and cocaine interact is obvious. However, many questions remain. If cocaine has central and peripheral actions, are observationsmade from isolated organ responsesvalid for explaining cocaine'seffects in the intact animal? Likewise, do data from anesthetized animals apply in the unstressed and unanesthetized state? Do different species metabolize cocaine differently? And finally, can data obtained from nonprimates
Woods, Scott, and Plessinger 1033
extrapolate to the clinical setting? Cocaine is an illicit drug whose abuse in the clinical setting is widespread among women of reproductive age. The study of cocaine's actions on the adult heart and cardiovascular system in pregnancy is likely to emerge as an important aspect of the biology of addiction medicine. REFERENCES 1. Lefkowitz Rj, Caron MG, Stiles GL. Mechanisms of membrane-receptor regulation. N Engl J Med 1984; 310: 1570-9. 2. Yamaguchi N, DeCharnplain J, Nadeau RA. Regulation of norepinephrine release from cardiac sympathetic fibers in the dog by presynaptic alpha and beta receptors. Circ Res 1977; 41: 108-17. 3. Wilkerson RD. Cardiovascular toxicity of cocaine. In: Clouet D, Asghar K, Brown R, eds. Mechanisms of cocaine abuse and toxicity. Rockville, Maryland: US Department of Health and Human Services, Alcohol, Drug Abuse, and Mental Health Administration, 1988: 304-24; National Institute on Drug Abuse Monograph 88.
4. Woods JR, Plessinger
MA. Pregnancy increases cardiovasto cocaine. Am J OBSTET GYNECOL 1990; 162:
cular toxicity 529-33.
5. Woods JR, Plessinger MA, Scott K, Miller RK. Prenatal cocaine exposure to the fetus: a sheep model for cardiovascular evaluation. Ann NY Acad Sci 1989; 562: 267-79. 6. Plessinger
increases cardioMA, Woods JR. Progesterone to cocaine in nonpregnant vascular toxicity ewes. A-m J OBsTET GYNECOL 1990; 163: 1659-64.
7.
R. HemodyJ, Miller L, Key TC, Resnik ewe namic effects of intravenous on the pregnant cocaine and fetus. Am J OBsTET GYNECOL 1986; 155: 883-8.
8.
Dean
ME,
drugs
during
Moore
Sorg
TR,
Stock
BH.
Hepatic in
pregnancy
of metabolism Metab Dispos
microsomal the rat. Drug
1975; 3: 325-31. 9.
Cousineau
D,
cardiac
interstitial inhibition
10.
uptake Assali
11.
PeiTeault
NS,
Goresky
CA,
CL.
Rose
norepinephrine in dogs. Girc
basal
Decreased
release after neuronal Res 1986; 58: 859-66.
CK III, et al. Autonomic B, Brinkman in pregin vivo and in vitro control of pelvic circulation: nant and nonpregnant sheep. Am J OBSTET GYNECOL 198 1; 141: 873-84.
effects ferret
Nuwayhid
CL,
Hague
of cocaine ventricular
on
NL,
intracellular
Morgan
ýJ,
Ransil
JP.
'
Ca
The
responsiveness 1990; Br J Pharmacol
myocardium.
of 10 1:
679-85. 12.
Sharma RK,
A,
Plessinger
Woods
JR.
cocaine: role of 1992; 113: 30-5. 13.
Sharma
A,
MA,
Pregnancy
MA,
terone mifepristone antagonist Proc Soc diotoxicity of cocaine. 279-87. HO,
Pedersen
RK,
Miller
H,
Woods
(RU
486)
Exp,
Biol
Finster
Med
15.
in pregnant toxicity and nonpregnant ogy 1985; 63: 134-9. S, Fox J, Johnson M, Covino Moller RA, Datta the cardiac liclocaine. and
JR.
Proges-
1993;
car202:
M, et al. Bupivacaine ewes. Anesthesiol-
Morishima
on
Miller
decreases
14.
of progesterone of bupivacaine
CS,
cardiotoxicity of App] Pharmacol
enhances Toxicol
progesterone.
Plessinger
Liang
DM,
Sherer
electrophysiologic Anesthesiology
BG.
Effects
actions 1992; 76:
604-8.
Discussion DP- Guy M. HA"ERT, JIL, Charlottesville, Virginia. For the past 6 or so years Woods et al. have enlightened us in a progressive manner on the cardiotoxic effects of In cocaine in the ovine experimental preparation.
1034
April 1994 Am j Obstet Gynecol
Woods, Scott, and Plessinger
1987, using a protocol similar to that presented today, they' reported that intravenous administration of cocaine to pregnant ewes produced dose-dependent increasesin maternal mean arterial blood pressure. They also observeda marked reduction in uterine blood flow, which they attributed to a dramatic increasein vasoconstriction of the uterine arteries. In 1988 they' reported that the response of maternal mean arterial blood pressure was maximum by I minute and returned to baseline levels by 30 minutes after injection. In 1990 they' expanded the measurement parameters to include systolic and diastolic blood pressure and performed similar studies in nonpregnant oophorectomized ewes.The observation that pregnant ewesdemonstrated significantly greater absolute increases in blood pressureresponsecompared with those observed in the castrated subjects led Woods et al. to conclude that pregnancy increasesthe cardiotoxicity of cocaine. This leads to my first question. Is the response to cocaine of the oophorectomized and the nonpregnant ewe with intact ovaries the same? In this study Dr. Woods has confirmed his observations that the magnitude of cardiac response to cocaine is greater in pregnant than in nonpregnant, oophorectomized animals. He has reconfirmed his earlier reports that cocaine produces increasesin systolic, diastolic, and mean arterial blood pressures that are dose and time related. He also reiterated the findings reported in 19882that 1.0 mg/kg dosages of cocaine produce a mild bradycardia but 2.0 mg/kg produces a transient tachycardia. However, he does not present statistics to indicate that these observed trends in heart rate are significantly different from baseline values.Consequently,my secondquestion is what is the basis for the conclusion that cocaine produces a doseand time-related increase in heart rate? The observationsthat cardiac oxygen consumption is increased in both pregnant and nonpregnant animals are consistentwith the oxygen consumption values at I and 2 minutes after cocaine injection, which can be readily calculated from data Woods and Plessinger' published in 1990. The figures presented today have extremely wide bars representing the SEM. My third question is, becausein actuality each animal could serve as its own control, why did you choose nonparametric tests to measure statistical significance? The new information presented today is data related to systemic vascular resistance and the measure of cardiac output. These data are of interest in that at dosages of 1.0 and 2.0 mg/kg cocaine produced a maximum response in systemicvascular resistancethat was 35% and 78% over baseline, respectively. Earlier Woods et al. ' reported that the maximum increase in uterine vascular resistance to similar cocaine dosages was 98% and 168%. Does this indicate that the uterine vasculaturehas an enhanced response to cocaine compared with the systemic circulation? This question is asked more in speculation and as possible grounds for future investigation than in anticipation of a definitive answerat this time.
Throughout his presentations Dr. Woods has reported that the cardotoxic effects of cocaine are greater in pregnant than in nonpregnant, oophorectornized preparations. However,there is evidenceto suggestthat any differences in response may be more related to drug dosagesand plasma concentrations than to alterations in response mechanisms. First, cocaine was administered on a milligram per kilogram of body weight basis. In an earlier publication' Dr. Woods reported that the mean weight of pregnant animals was 15% greater than the nonpregnant, oophorectomized ones. Becausedosage is based on body weight and administered over 15 secondsthrough a catheter placed in the femoral vein, the amount of cocaine to which the heart of the pregnant animal would be exposed on first pass would be greater. Second, the plasma cocaine concentrations in pregnant eweswere consistently greater than those measured in the nonpregnant, oophorectomized animals. This great disparity, plus the large SENIs,leads me to my next three questions. (1) What is the specificity and reproducibility of your plasma cocaine measurements?(2) Do we have any idea of what proportion of the plasma levels is active and what proportion, if any, is inactive? and (3) Have you performed pharmacokinetic studies to see if there is a difference in the half-life of cocainebetweenpregnant and nonpregnant, oophorectomized sheep? REFERENCES
1. Woods JR, Plessinger MA, Clark KE. Effect of cocaine on uterine blood flow and fetal oxygenation. JAMA 1987; 257: 957-61. 2. Woods JR, Plessinger MA, Scott K, Miller RK. Prenatal cocaine exposure to the fetus: a sheep model for cardiovascular evaluation. Ann NY Acad Sci 1988; 562: 267-79. 3. Woods JR, Plessinger MA. Pregnancy increases cardiovascular toxicity to cocaine. Am j ORSrET GYNECOL 1990; 162; 529-33.
D& WOODS(Closing). The question was whether the response to cocaine in nonpregnant, oophorectomized ewes was the same as that in those nonpregnant ewes that were not oophorectomized. We have not repeated the studies in nonpregnant ewes, leaving the ovaries intact. The purpose of oophorectomy, of course, was that we were trying to eliminate the possibility of hormonal flux. We preferred to have a clean system where we could address specifically the issue of progesterone and pregnancy in an intact model. Your question regarding the wide variability of cardiovascular measurements during initial cocaine exposure is an appropriate one. There's no question that during the first 60 seconds after cocaine administration there are a number of events that are occurring, and they are occurring very rapidly. When cardiovascular measurements were obtained at 5,15,30 and 60 minutes, at least four measurements were made over a 30-second period and then averaged. During the initial exposure we obtained measurements every 10 seconds and therefore could only take one measurement at any time point. During this acute
Woods, Scott, and Plessinger 1035
Volume 170, Number 4 Am j Obstet Gynecol
exposure period the animals are often quite agitated; this in part contributes to our observations of the wider variability. Nonetheless, we now know that many of the acute, and probably life-threatening, cardiac events to cocaine appear within the first 60 seconds of exposure. I might add that our study was based on the premise that cocaine-induced enhanced cardiac performance would explain the elevated hypertensive response seen in pregnant and progesterone-treated, nonpregnant ewes. We feel that we have effectively shown to our satisfaction that the augmented hypertensive response to cocaine in pregnant ewes is not driven by increased cardiac output secondary to tachycardia but, in fact, represents a primary event in the peripheral circulation. Dr. Harbert asked whether the augmented response to cocaine in pregnant ewes was not simply based on higher cocaine concentrations in the pregnant ewe's circulation. Cocaine was given to pregnant and nonpregnant ewes in similar concentrations on the basis of total body weight. If the total body weight of the pregnant ewe included fetus, amniotic fluid, and placenta, we would be giving a higher dose of cocaine to the pregnant ewes than to the nonpregnant ewes. That issue inspired us to administer progesterone to very nonpregnant animals. We believed that if we could demonstrate the same augmented cocaine effect with nonpregnant progesterone-treated ewes as we did in the pregnant ewes, then it would be very unlikely that
cocainedistribution in the adult circulation was responsible for these differences. It is possible, however, that metabolism may play a significant role in altering cocaine actions in pregnant ewes. We have not compared elimination half-life measurements for cocaine in pregnant ewesversusnonpregnant ewes.Others have examined this factor but not during direct comparison of half-life in the pregnant and nonpregnant states. Cocaine's elimination half-life is dose dependent, and in the nonpregnant human it is 40 to 82 minutes,' whereasin the pregnant guinea pig it is 20 to 67 rninuteS2and in the pregnant sheep 4.4 to 5 minutes.' It is possible that conversion of cocaine to some of the more active metabolites may be important and may need to be further evaluated. Dr. Harbert asked about the reproduceability in our cocaine assays.We have not split samplesto run simultaneous measurementsto determine that. However,our minimal level of detection for cocaine is 5 ng/ml.
REFERENCES
1. Barnett G, Hawk R, Resnik R. Cocainepharmacokineticsin humans.j Ethnopharmacol 1981;3:353-66. 2. Sanberg JA, Olsen GD. Cocaine pharmacokinetics in the pregnant guinea pig. Pharmacol Exp Ther 1991;258:477.1 81. 3. De Vane CL, Burchfield Dj, Abrams RM, et al. Disposition of cocaine in pregnant sheep. Dev Pharmacol Ther 1991; 16:123-9.