Morphine Stimulates Adrenocorticotropin and Cortisol Release in the Late-Term Ovine Fetus

Abstract

Opiates are widely used as obstetrical analgesics during pregnancy and, as such, their interactions with the fetal endocrine system may have important consequences. In this study, the effects of morphine administration to fetal sheep in utero on fetal plasma immunoreactive (ir)-ACTH and ir-cortisol were examined. At the lowest dose administered (0.6 mg/h, i.v.) morphine reduced, although not significantly, plasma ir-cortisol levels. A dose-dependent stimulation of cortisol release was observed with higher doses of morphine. Doses of 2.5 and 5.0 mg/h morphine resulted in a significant increase in ir-cortisol with a change from control levels equal to 9.6± 1.1 ng/mL (p = 0.03) and 17.6 ± 5.1 ng/mL(p = 0.03), respectively. This increase in plasma ir-cortisol was associated with a significant increase in ir-ACTH (111.8 ± 23.2 pg/mL versus 42.8 ± 5.1 pg/mL; p = 0.02) that was naloxone-reversible. These effects of morphine were observed in fetal lambs only >125 d of gestation, suggesting a maturation of functional opioid receptors in the ovine fetal hypothalamic-pituitary-adrenal axis after this time.

Main

Morphine and other opiates are widely used as obstetrical analgesics during pregnancy, and their lipid solubility enables them to pass through the placenta and readily enter fetal circulation⁽¹⁾. Opioid drugs are known to have significant effects on the HPA axis⁽²⁾. As such, their interactions with the fetal HPA axis may have important consequences. Fetal adrenal cortisol plays an important role in the onset of parturition as well as the maturation of several fetal organ systems, including the lung and kidney, that are essential for extrauterine survival⁽³⁾. The stimulatory effects of morphine on the HPA axis have been observed as early as postnatal d 5 in the rat⁽⁴⁾. The effects of morphine on the developing fetal HPA axis in utero have not been previously investigated. A stimulatory action of opiates on the fetal HPA axis is suggested by reports of increased fetal cortisol levels after the administration of opioid peptides to fetal sheep^{(5, 6)}. The aim of this study was to determine the effects of morphine on the fetal HPA axis to better understand the impact of opiate administration during pregnancy.

The effects of opioids on the HPA axis is greatly species-dependent. Acute morphine administration stimulates the adult HPA axis in many mammalian species, including the rat, mouse, guinea pig, and cat⁽²⁾. The response of the HPA axis to opioids in humans is the opposite of the response in these animal species. Opioid agonists have been reported to inhibit, whereas opioid antagonists stimulate, the release of ACTH and cortisol in humans^{(2, 7)}. The reason behind this difference in HPA response to opioids between humans and most experimental animal species is not known. Interestingly, available evidence indicates that the HPA response to opioids in sheep is similar to that in humans. Morphine has been reported to decrease plasma ACTH and cortisol^{(8, 9)}, whereas naloxone has been shown to enhance cortisol release in sheep^{(10, 11)}. Thus the sheep appears to be a suitable model to investigate the response of the HPA axis to opioid agents.

To this end, we have investigated the effects of morphine on fetal plasma ACTH and cortisol levels in the unanesthetized ovine fetus. Furthermore, we have examined the age dependency and naloxone sensitivity of this endocrine action of morphine. The responsivity of the fetal pituitary and adrenal very as a function of gestational age, and it is well established that the fetal adrenal does not respond to ACTH until after 125 d⁽⁶⁾. We have therefore compared the effects of morphine on fetal plasma ACTH and cortisol levels in fetal lambs before and after 125 d of gestation. To determine the direct effects of morphine on the fetus and minimize possible confounding factors with maternal morphine administration, we have chosen to administer morphine directly to the fetus.

METHODS

Guidelines approved by the Institution for the Care and Use of Animals at Cornell University Medical College were followed for all surgical procedures and experimental protocols. Surgery was performed on 31 pregnant ewes of mixed Western breed. Under lidocaine anesthesia supplemented with pentobarbital, the fetus was surgically instrumented with chronic indwelling polyvinyl catheters into the distal aorta for blood sampling and inferior vena cava for drug infusion. These surgical procedures have been previously described in detail⁽¹²⁾.

A minimum of 5 d after surgery and after a 2-h control period, an i.v. infusion of morphine sulfate was administered to the fetus at a dose of 0.6, 1.2, 2.5, or 5.0 mg/h for a duration of 1 h. In five studies saline vehicle was given in the same manner as morphine. In those studies in which the effect of the opioid antagonist, naloxone, on the morphine response was investigated, naloxone (6 mg/h) was given as an i.v. infusion beginning 1 h before, during, and 1 h after morphine administration. These studies used an incomplete randomized cross-over design where each fetus was allowed a minimum of 2 d between studies and did not receive the same drug combination in successive studies. Each fetus received a total of one to three drug treatments.

Fetal blood samples (2 mL) were collected before and at the end of the morphine infusion, as well as 1 and 3 h after the end of infusion. The plasma was separated by centrifugation at 1500 rpm for 10 min at 4°C and subsequently frozen at -70°C until assayed for ir-ACTH and ir-cortisol by RIA (INCSTAR, Stillwater, MN).

The data are represented as mean ± SEM. The effects of morphine on change from control in ir-ACTH and ir-cortisol was assessed using a single factor analysis of variance (factor = time) followed by the Dunnett post hoc analysis. The effects of naloxone on the morphine response was analyzed by multifactor analysis of variance (factor = drug, time). Significance was set at p < 0.05.

RESULTS

The dose-response relationship of morphine on fetal plasma ir-cortisol in fetuses >125 d. Basal plasma ir-cortisol levels were highly variable and averaged 16.7 ± 2.6 ng/mL in fetal lambs >125 d of gestation (n = 28). The administration of the saline vehicle had no effect on ir-cortisol. Fig. 1 shows the change in fetal plasma ir-cortisol after four different doses of morphine administered directly to the fetus. With the 0.6 mg/h dose there was a decrease in ir-cortisol from control levels by the end of the morphine infusion, although it did not reach statistical significance because of the large variation in basal levels in the five animals. There was no effect with 1.2 mg/h(n = 4). Plasma ir-cortisol was significantly increased after the administration of 2.5 mg/h (p = 0.03; n = 9) and 5.0 mg/h(p = 0.03; n = 5). Peak plasma ir-cortisol reached 20.4± 2.5 ng/mL and 50.7 ± 10.3 ng/mL, respectively, at 1 h after the morphine infusion.

Figure 1

The change from control values in fetal plasma ir-cortisol after administration of i.v. morphine at a dose of 0.6 (▾), 1.2(•), 2.5 (▪), and 5.0 (▴) mg/h in fetuses >125 d of gestation. The morphine infusion was administered from 0 to 1 h. *p< 0.05.

Age dependency of morphine's effect on ir-ACTH and ir-cortisol. The response to 5.0 mg/h was compared in two age groups, those ≤125 d of gestation and >125 d of gestation (Fig. 2). Different animals were used in the two groups. In the older age group, the increase in plasma ir-cortisol (p = 0.03, n = 5) was associated with a significant increase in ir-ACTH (p = 0.02, n = 5) with a peak level of 111.8 ± 23.2 pg/mL compared with a basal level of 42.8± 5.1 pg/mL. In contrast, there was no significant change in either ir-ACTH or ir-cortisol in the younger age group (n = 5).

Figure 2

The effects of 5.0 mg/h i.v. morphine on the change from control in fetal plasma levels of (A) ir-ACTH and (B) ir-cortisol in fetuses >125 d gestation (▴) and ≤125 d gestation(▵). The morphine infusion was administered from 0 to 1 h. *p< 0.05.

Naloxone sensitivity of the morphine response. The rise in ir-ACTH observed with the administration of 5.0 mg/h morphine in fetuses>125 d was abolished by the administration of naloxone (6.0 mg/h;Fig. 3) with there being a significant difference between treatments (p = 0.04, n = 5). The ir-cortisol response was also significantly blocked by concurrent naloxone administration (p= 0.04; n = 6). There was no effect of naloxone administration alone on fetal plasma levels of ir-ACTH or ir-cortisol, with n = 5.

Figure 3

The change from control values in fetal plasma(A) ir-ACTH and (B) ir-cortisol after administration of 5.0 mg/h morphine (▴), 5.0 mg/kg morphine with concurrent naloxone (6.0 mg/h, i.v., ▿), and naloxone alone (□). The morphine infusion was administered from 0 to 1 h. In studies with morphine and naloxone, naloxone was administered as a continuous infusion 1 h before, during, and 1 h after the morphine infusion. *p < 0.05.

DISCUSSION

The results of the present study show that morphine can exert different effects on the fetal HPA axis as a function of dose. Fetal plasma cortisol levels are inhibited by low doses of morphine and increased by high doses of morphine, such that moderate doses may appear to have no effect. With the highest dose of morphine administered (5 mg/h), fetal plasma levels of ir-cortisol were approximately two to three times that of control levels. This magnitude of stimulation of cortisol release is comparable to that which has been observed with morphine administration to the neonatal rat⁽⁴⁾. Further, the present data suggest that the increase in plasma cortisol appears to result from a release of ACTH from the fetal pituitary, and this action is mediated by specific opioid receptors as it is abolished by naloxone administration.

Similarly mixed effects of morphine on the HPA axis have been reported in the adult human and other animal species. Acute i.v. morphine has been reported to either have no effect or decrease basal plasma cortisol levels in humans^(13–16). The basal cortisol level is also decreased or unchanged after morphine administration in sheep^(8–11). In contrast, morphine stimulates ACTH and cortisol secretion in rats^(17–20). The reason behind these disparities is not known, although the dose of morphine may make a difference, because most studies used a single dose of drug. A decrease in plasma glucocorticoid levels with low doses and an increase with high doses has been reported in the adult rat after morphine administration⁽²¹⁾.

The stimulatory effect of opiates on glucocorticoid release has been shown to be abolished by hypophysectomy^{(17, 22)} and lesions of the median eminence⁽²³⁾, and a similar increase in glucocorticoid release could be observed after microinjections of opiates into the hypothalamus⁽²⁴⁾. This indicates that the action of opiates on adrenal glucocorticoid release is secondary to the secretion of ACTH from the pituitary, most likely via the release of hypothalamic secretagogues. In the rat, morphine has been shown to stimulate an increase in the hypothalamic content and secretion of CRF both in vivo and in vitro^{(18, 25)}. Coadministration of CRF antiserum with morphine to the rat, however, was unable to block the increase in ir-ACTH⁽¹⁹⁾. A site of action at the fetal hypothalamus is supported by ligand binding studies that have revealed [³H]naloxone binding sites in the ovine fetal hypothalamus but not in the anterior pituitary⁽²⁶⁾. Morphine is thought to exert its effects by preferentially binding to theμ, as opposed to δ or κ, subtype of opioid receptor. Evidence suggests that morphine is elicting its effects on the HPA axis via μ receptors as the increase in CRF release from adult rat hypothalami can be blocked by the selective μ antagonist, β-funaltrexamine⁽²⁵⁾. Further, μ binding sites have been demonstrated in the ovine hypothalamus with the use of the relatively μ-selective ligand dihydromorphine⁽²⁷⁾.

The mechanism behind the inhibitory action of morphine on the HPA axis is not understood. The dual response to morphine may be due to the ability of morphine to interact with different subclasses of opioid receptors or to provoke the release of various hypothalamic stimulatory or inhibitory factors. Morphine might have an inhibitory action on one type of hypothalamic secretagogue and a stimulatory action on another type which may be dependent on the dose of drug administered. The inability of CRF antiserum to block morphine's action in the adult rat⁽¹⁹⁾ confirms that morphine may act through more than one ACTH secretagogue. The precise site and mechanism of action of the inhibitory actions of morphine require further investigation.

There is also evidence that the action of morphine may depend on the endocrine state of the subject. Morphine can block the stimulation of the HPA in response to various stressors. Although morphine did not alter basal cortisol levels in sheep, it significantly attenuated the cortisol response to isolation stress⁽¹⁰⁾. Morphine also blocked the cortisol response to surgical stress in humans^{(28, 29)}. This suppression of the normal stress response by morphine appears to be mediated by an inhibition of the pituitary-adrenal response to CRF⁽³⁰⁾. The ability of morphine to reduce the cortisol response to stressors in rats seems paradoxical given the stimulatory action of morphine on the HPA axis under basal conditions⁽³¹⁾. The effects of morphine on the fetal HPA axis under stress conditions have not been examined, and this may have particular relevance when morphine is used for pain relief during labor and delivery.

The action of morphine on the fetal HPA axis was observed in fetuses only>125 d of gestation. These results are in agreement with the study by Brooks and Challis⁽⁶⁾, where the stimulatory effects of FK33-824 were not observed in fetuses at 110-115 d of gestation, but only in fetuses at 125-140 d. Fetal pituitary corticotrophs respond to CRF by releasing ACTH around 97 d of gestation⁽³⁾, and the fetal adrenal is responsive to ACTH from 120 d to term⁽³²⁾. The lack of an ACTH response in fetuses younger than 125 d of gestation is therefore most likely explained by a lack of functional opioid receptors in the hypothalamus before that stage of development. Yang and Challis⁽²⁶⁾ have demonstrated that there are [³H]naloxone binding sites in the ovine fetal hypothalamus at 110-115 d of gestation but the density at this gestational age is much lower than that observed at 125-130 d. Further, Dunlap et al.⁽²⁷⁾ also showed that dihydromorphine binding in the hypothalamus is significantly lower at 97-110 d than at 118-127 d of gestation. Thus, earlier in gestation there may be an inadequate number of receptors present, or alternatively, the second messenger systems may not be functional at this time in development.

Previous studies in our laboratory have shown similar opposing effects of morphine on other fetal physiologic systems as a function of dose. Low doses of morphine tend to result in an excitatory profile in the ovine fetus, with fetal tachycardia, respiratory stimulation, activation of the electrocorticogram, and hypoglycemia^(33–36). Higher doses of morphine lead to respiratory depression, hyperglycemia, and a return of fetal heart rate to control levels^{(35, 36)}. Thus the impact of opiate administration on the fetus is entirely dependent on the extent of fetal drug exposure. Although the release of ir-ACTH and ir-cortisol by morphine in the ovine fetus was much less than what has been reported with hemorrhagic⁽³⁷⁾ or hypertensive⁽³⁸⁾ stress, this degree of perturbation of the fetal HPA, especially under chronic conditions, may affect the ability of the fetus to respond to other stressors, thereby jeopardizing fetal homeostasis.