- Email: [email protected]
Neonatal effects of beta-adrenergic drugs
391
Neonatal effects of beta-adrenergic drugs G.G.M. Essed, Department of Obstetrics and Gynaecology, St. Annadal Hospital, State University of Limburg, N-Maastricht, The NetherIan& Since Rucker [l] Boume and Burn [2] reported 55 yr ago upon inhibition of uterine activity by intravenous administration of adrenaline, much progress has been made in the tocolytic field. Today relaxation of the uterus can be attained effectively with utero-selective adrenergic drugs, and these agents are increasingly used for indications other than the prevention of preterm labor. Unfortunately, these drugs have a great many side-effects, especially in the dosage levels used for tocolysis. Most of the side effects are not unexpected as we look at the effects of adrenergic recepter stimulation in several organs. The fact that so many different organs are involved in the actions of beta-mimetic agonists, makes these drugs potentially dangerous to the unborn child, since transplacental transfer has been proven for some drugs in this category and is likely for all of them. Several authors have reported upon shorter and long-term influence from these drugs on fetus and newborn. Efforts have been made to measure the influence upon the fetal heart. Studies of the magnitude of acceleration of fetal heart rates have not been in agreement but a mild elevation of fetal heart rate seems to be a common side-effect of maternal beta-mimetic medication. Hofmann et al. [3] reported myocardial necrosis following fenoterol exposure in an in vitro study of myocardial strips, and there have been a growing number of case reports and retrospective investigations suggesting the possibility of fetal myocardial damage by these drugs. Neither Spelger [4] nor Plieth [5] were able to demonstrate pathological cardiac findings (CK-MB, electro- and echocardiograms) in newborns following tocolysis with fenoterol or isoxsuprine and verapamil, which were related to the preceding medication. Freysz and colleagues [6] performed a long-term evaluation of infants who were exposed to ritodrine while in utero. No statistically significant differences were found in psychomotor development (Denver Developmental test), clinical examination, electrocardiograms, and other investigated parameters between the exposed neonates and control infants. A negative influence on the body weight gain until the fifth day of life was reported by Stormi [7] in the newborn who was exposed to ritodrine while in utero, compared to control neonates. Favourable side-effects were mentioned as well: recent research suggests an accelerated fetal lung maturation after intrauterine exposure to beta-agonists. The research-group of Juan Esteban-Altarriba plays a leading role in this field. Beta-mimetic drugs are also believed to exert an influence on carbohydrate metabolism by a stimulation of adenylcyclase in liver, muscle and probably pancreatic cells. The net effect of these actions will be a transient elevation in blood glucose concentration, in spite of an increased insulin secretion. For a proper understanding of the possible effects on the fetus we have to realize that glucose crosses the placenta easily, beta-mimetics probably, while insulin does not. The maternal carbohydrate intolerance which is induced by these drugs and the
398
fact that beta-mimetics do cross the placental barrier give support to the suggestion that long-term beta-mimetic medication, as used in the prevention of premature labor, could result in fetal hyperinsulinism. Epstein et al. [8] studied umbilical cord blood insulin levels and neonatal blood glucose values in 12 infants whose mothers had received beta-mimetic therapy (fenoterol and terbutaline). In 5 of the 6 cases where the interval from termination of beta-mimetic therapy to delivery was 2 days or less, the infants exhibited sustained hypoglycemia. This was not observed in any case where the interval exceeded 2 days. The mean cord blood insulin levels in the 6 infants with a short termination-delivery time was 50% greater than in the larger interval group. Umbilical cord glucose levels were not significantly different. However, no comparison was made with untreated neonates. In retrospect, Brazy and Pupkin [9] found a significantly higher incidence of hypoglycemia (as well as hypocalcemia, hypotension and ileus) in premature neonates whose mothers were treated with isoxsuprine within 48 h of delivery (n = 43) in comparison with a reference population of 107 untreated infants. The present investigation, was undertaken to assess the possible influence of prenatal exposure to beta-adrenergics upon the neonatal glucose levels in the first 30 min after delivery. Material and methods From January 1976 until January 1979, plasma glucose levels were measured in the first 30 min of extrauterine life in neonates from mothers treated with betamimetic drugs (fenoterol and/or ritodrine) for the prevention of premature birth. In general, these tocolytic agents were administered intravenously for at least 2 days at the minimum dose necessary to stop labor successfully (dose i.v. 67-700 pg/min ritodrine, 1.0-9.3 pg/min fenoterol). The dosage was then decreased; and if uterine contractions did not return, oral treatment with 5 mg fenoterol or 10 mg ritodrine 6 to 12 times daily, was substituted. In general, medication was continued until the end of the 37th wk of gestation. Ritodrine was given to 83 patients; and fenoterol, to 43 patients. Fifteen women were treated with both beta-agonists. For every patient in the treatment group who gave birth in 1976 and 1977, the next subsequent patient who delivered with the same duration of amenorrhea (f4 days) was taken as a control. Mean f SD and range of gestational age at delivery were 257 $- 19 days (203-294) for the treated neonates and 259 f 15 days (206-290) for the control infants. Only patients who delivered vaginally were included in the study. Children of patients with diabetes mellitus, gestational diabetes, severe dysmaturity (below 2.3th percentile of birthweight for gestational age) and congenital malformations were not included or removed afterwards, because of the influence that these factors may have upon the serum glucose level. For the same reason neonates with pH values of the umbilical artery blood under 7.10 and neonates whose body temperature, measured exactly at 30 min after delivery, had dropped below 35.0°C, were excluded. After exclusion of patients according to the mentioned criteria, 141 treated patients and 110 control babies were left for comparison. The parametes were tested according to Wilcoxon.
399
Results After exclusion of twin pregnancies, the birthweights of the treated and control babies were similar: mean + SD and range being 2728 + 614 g (1300-4220) and 2837 f 696 g (1386-4450), respectively. No significant differences were found in glucose values of the control neonates with respect to the delivery-sample (d-s) interval. This was true for both premature infants and the term neonates. A strong influence could be measured according to the number of days treatment had been stopped before a premature baby was born. Neonates born before the 37th wk of gestation had lower plasma glucose levels when treatment was continued to within 72 h of delivery (2.6 + 0.6 mmol/l) than with an interval of 3 or more days (3.0 f 0.8 mmol/l) (P < 0.05). There was a statistically significant difference between the plasma glucose values of premature and term infants in the control group (P < 0.05). This difference was more pronounced (P -C0.0001) when treatment was given, probably as a result of the fact that far more premature (70%) than term neonates (19%) were born with a treatment-delivery interval of less than 48 hours. Significantly lower plasma glucose concentrations were measured in babies from treated mothers than were found in the control group: 2.9 k 0.8 versus 3.3 + 0.8 mmol/l (P < 0.001). After subdivision, these differences were evident in prematurely born infants (2.7 f 0.7 versus 3.1 + 0.8 mmol/l; P-C 0.001). whereas the glucose values of the treated infants born after the 37th wk of amenorrhea did not differ significantly from control neonates of the same gestational age (Table I). As mentioned above, it must be stressed that the mothers of only 19% of the term neonates were still being treated 48 h before delivery. In conclusion, we strongly advices monitoring of glucose values to be started very early (within 30 min) in these infants in order to recognize and correct promptly neonatal hypoglycemia. This applies even more when gestational age is lower and the treatment-delivery interval is shorter. References 1. Rucker MP.
The action of adrenaline on the pregnant human uterus. S Med J 1925; 18: 412. 2. Bourne A, Burn JH. The dosage and action of pituitary extract and of the ergot alkaloids on the uterus in labor, with a note on the action of adrenalin. J Obstet Gynaecol Br Emp 1927; 34: 249-272. TABLE
I
Mean f SD of plasma after delivery
Treated
Control
patients
patients
glucose
levels (mmol/l)
of treated
and control
neonates
measured
All patients
Patients
< 37 wk
Patients
2.9*0.8 n = 141
27:07.P<~‘~~3.siO.8 . . n = 71
n = 70
P < 0.001
P
P50.14
3.3 rt0.8 ?I=110
3.1 zbO.8 n = 49
3.4*0.9 n=61 P i 0.05-----J
within
> 37 wk
30 min
3. Hofmann W, Schleich A, Schroeter D, Weidinger H, Wiest W. Der Einflusz von Beta-sympathomimetika und sog. Ca.++ - antagonist&her Hemmstoffe auf den menschlichen Herzmuskel in vitro. Virchows Arch (Path Anat) 1977; 373: 85-95. 4. Spelger G, Wiest W, Schlicker H, Kachel W, Puls F. Kardiale Parameter beim Neugeborenen nach Tokolyse. Klin Paediatr 1980; 192: 319-324. 5. Plieth M, Ache-Ebelt H, Fitzner R, Bein G, Karkut G. Untersuchung der Herzfunktion nach Langxeittokolyse mit Fenoterol und Isoxsuprin bei Neugeborenen. Z Geburtshilfe Perinatol 1980; 184: 275-282. 6. Freysz H, Willard D, Lehr A, Messer J, Boog G. A long term evaluation of infants who received a beta mimetic drug while in utero. J Perinat Med 1977; 5: 94-99. 7. Stormi M, Gigli C, Nordio S, Mandruazato GP. Effects of the treatment with ritodrine on the fetus and newborn baby. In: Bompiani A, Cosmi EV, Fischetti B, Gasparri F, Romanini C, eds. Recent Advances on Beta-mimetic Drugs in Obstetrics. Societa Editrice Universo 1977. 8. Epstein MF, Nicholls E, Stubblefield PG. Neonatal hypoglycemia after beta-sympathomimetic tocolytic therapy. J Pediatr 1979; 94: 449-453. 9. Brazy JE, Pupkin MJ. Effects of maternal isoxsuprine administration on preterm infants. J Pediatr 1979; 94: 444-448.
Clinical study of indomethacin for prevention of prematurity C. Sureau and P. Piovani, Unique 75674 Paris Cedex 14, France
Universitaire Baudelocque, 123 Bd. de Port-Royal,
Introduction
Prostaglandin inhibitors have been demonstrated to be effective in blocking uterine activity [l-3], but this effectiveness involves possible side-effects in the mother and the child, since prostaglandin inhibitors pass easily through the placenta. As far as the mother is concerned, the contra-indications for indomethacin such as digestive-tract disturbances are well known. Less clear is the situation concerning the fetus: experimental studies have shown possible noxious effects on the pulmonary circulation, leading to closure of the ductus arteriosus in utero. Some other deleterious actions have been described for the kidney, the digestive tract and the brain [4]. The results of these studies call for caution when using indomethacin. Nevertheless, definite contra-indications for the use of betamimetics [5] as well as intolerance or failure of this treatment have made it legitimate to try to evaluate the usefulness and consequences of the use of indomethacin. Clinical studies The first study conducted in the Baudelocque Hospital was done during the years
1978-1979 and the results were reported in 1980 [6]. This study comprised 84 patients, 82% of them with immediate tocolysis, a mean duration of treatment of 23 days (O-86), a mean total dose of 3742 mg (125-15,000), 35% recurrence of premature labour, 65% delivery later than 37 wk, and an average newborn weight of 2670 g (1240-3550), good maternal tolerance, but 3 cases of neonatal persistance of fetal circulation, with high pulmonary arterial pressure. One of these newborns died in the neonatal period. In these 3 cases the delivery occurred during the indomethatin treatment. A second study was conducted in 1980- 1982. Indomethacin was used only in cases
Neonatal effects of beta-adrenergic drugs G.G.M. Essed, Department of Obstetrics and Gynaecology, St. Annadal Hospital, State University of Limburg, N-Maastricht, The NetherIan& Since Rucker [l] Boume and Burn [2] reported 55 yr ago upon inhibition of uterine activity by intravenous administration of adrenaline, much progress has been made in the tocolytic field. Today relaxation of the uterus can be attained effectively with utero-selective adrenergic drugs, and these agents are increasingly used for indications other than the prevention of preterm labor. Unfortunately, these drugs have a great many side-effects, especially in the dosage levels used for tocolysis. Most of the side effects are not unexpected as we look at the effects of adrenergic recepter stimulation in several organs. The fact that so many different organs are involved in the actions of beta-mimetic agonists, makes these drugs potentially dangerous to the unborn child, since transplacental transfer has been proven for some drugs in this category and is likely for all of them. Several authors have reported upon shorter and long-term influence from these drugs on fetus and newborn. Efforts have been made to measure the influence upon the fetal heart. Studies of the magnitude of acceleration of fetal heart rates have not been in agreement but a mild elevation of fetal heart rate seems to be a common side-effect of maternal beta-mimetic medication. Hofmann et al. [3] reported myocardial necrosis following fenoterol exposure in an in vitro study of myocardial strips, and there have been a growing number of case reports and retrospective investigations suggesting the possibility of fetal myocardial damage by these drugs. Neither Spelger [4] nor Plieth [5] were able to demonstrate pathological cardiac findings (CK-MB, electro- and echocardiograms) in newborns following tocolysis with fenoterol or isoxsuprine and verapamil, which were related to the preceding medication. Freysz and colleagues [6] performed a long-term evaluation of infants who were exposed to ritodrine while in utero. No statistically significant differences were found in psychomotor development (Denver Developmental test), clinical examination, electrocardiograms, and other investigated parameters between the exposed neonates and control infants. A negative influence on the body weight gain until the fifth day of life was reported by Stormi [7] in the newborn who was exposed to ritodrine while in utero, compared to control neonates. Favourable side-effects were mentioned as well: recent research suggests an accelerated fetal lung maturation after intrauterine exposure to beta-agonists. The research-group of Juan Esteban-Altarriba plays a leading role in this field. Beta-mimetic drugs are also believed to exert an influence on carbohydrate metabolism by a stimulation of adenylcyclase in liver, muscle and probably pancreatic cells. The net effect of these actions will be a transient elevation in blood glucose concentration, in spite of an increased insulin secretion. For a proper understanding of the possible effects on the fetus we have to realize that glucose crosses the placenta easily, beta-mimetics probably, while insulin does not. The maternal carbohydrate intolerance which is induced by these drugs and the
398
fact that beta-mimetics do cross the placental barrier give support to the suggestion that long-term beta-mimetic medication, as used in the prevention of premature labor, could result in fetal hyperinsulinism. Epstein et al. [8] studied umbilical cord blood insulin levels and neonatal blood glucose values in 12 infants whose mothers had received beta-mimetic therapy (fenoterol and terbutaline). In 5 of the 6 cases where the interval from termination of beta-mimetic therapy to delivery was 2 days or less, the infants exhibited sustained hypoglycemia. This was not observed in any case where the interval exceeded 2 days. The mean cord blood insulin levels in the 6 infants with a short termination-delivery time was 50% greater than in the larger interval group. Umbilical cord glucose levels were not significantly different. However, no comparison was made with untreated neonates. In retrospect, Brazy and Pupkin [9] found a significantly higher incidence of hypoglycemia (as well as hypocalcemia, hypotension and ileus) in premature neonates whose mothers were treated with isoxsuprine within 48 h of delivery (n = 43) in comparison with a reference population of 107 untreated infants. The present investigation, was undertaken to assess the possible influence of prenatal exposure to beta-adrenergics upon the neonatal glucose levels in the first 30 min after delivery. Material and methods From January 1976 until January 1979, plasma glucose levels were measured in the first 30 min of extrauterine life in neonates from mothers treated with betamimetic drugs (fenoterol and/or ritodrine) for the prevention of premature birth. In general, these tocolytic agents were administered intravenously for at least 2 days at the minimum dose necessary to stop labor successfully (dose i.v. 67-700 pg/min ritodrine, 1.0-9.3 pg/min fenoterol). The dosage was then decreased; and if uterine contractions did not return, oral treatment with 5 mg fenoterol or 10 mg ritodrine 6 to 12 times daily, was substituted. In general, medication was continued until the end of the 37th wk of gestation. Ritodrine was given to 83 patients; and fenoterol, to 43 patients. Fifteen women were treated with both beta-agonists. For every patient in the treatment group who gave birth in 1976 and 1977, the next subsequent patient who delivered with the same duration of amenorrhea (f4 days) was taken as a control. Mean f SD and range of gestational age at delivery were 257 $- 19 days (203-294) for the treated neonates and 259 f 15 days (206-290) for the control infants. Only patients who delivered vaginally were included in the study. Children of patients with diabetes mellitus, gestational diabetes, severe dysmaturity (below 2.3th percentile of birthweight for gestational age) and congenital malformations were not included or removed afterwards, because of the influence that these factors may have upon the serum glucose level. For the same reason neonates with pH values of the umbilical artery blood under 7.10 and neonates whose body temperature, measured exactly at 30 min after delivery, had dropped below 35.0°C, were excluded. After exclusion of patients according to the mentioned criteria, 141 treated patients and 110 control babies were left for comparison. The parametes were tested according to Wilcoxon.
399
Results After exclusion of twin pregnancies, the birthweights of the treated and control babies were similar: mean + SD and range being 2728 + 614 g (1300-4220) and 2837 f 696 g (1386-4450), respectively. No significant differences were found in glucose values of the control neonates with respect to the delivery-sample (d-s) interval. This was true for both premature infants and the term neonates. A strong influence could be measured according to the number of days treatment had been stopped before a premature baby was born. Neonates born before the 37th wk of gestation had lower plasma glucose levels when treatment was continued to within 72 h of delivery (2.6 + 0.6 mmol/l) than with an interval of 3 or more days (3.0 f 0.8 mmol/l) (P < 0.05). There was a statistically significant difference between the plasma glucose values of premature and term infants in the control group (P < 0.05). This difference was more pronounced (P -C0.0001) when treatment was given, probably as a result of the fact that far more premature (70%) than term neonates (19%) were born with a treatment-delivery interval of less than 48 hours. Significantly lower plasma glucose concentrations were measured in babies from treated mothers than were found in the control group: 2.9 k 0.8 versus 3.3 + 0.8 mmol/l (P < 0.001). After subdivision, these differences were evident in prematurely born infants (2.7 f 0.7 versus 3.1 + 0.8 mmol/l; P-C 0.001). whereas the glucose values of the treated infants born after the 37th wk of amenorrhea did not differ significantly from control neonates of the same gestational age (Table I). As mentioned above, it must be stressed that the mothers of only 19% of the term neonates were still being treated 48 h before delivery. In conclusion, we strongly advices monitoring of glucose values to be started very early (within 30 min) in these infants in order to recognize and correct promptly neonatal hypoglycemia. This applies even more when gestational age is lower and the treatment-delivery interval is shorter. References 1. Rucker MP.
The action of adrenaline on the pregnant human uterus. S Med J 1925; 18: 412. 2. Bourne A, Burn JH. The dosage and action of pituitary extract and of the ergot alkaloids on the uterus in labor, with a note on the action of adrenalin. J Obstet Gynaecol Br Emp 1927; 34: 249-272. TABLE
I
Mean f SD of plasma after delivery
Treated
Control
patients
patients
glucose
levels (mmol/l)
of treated
and control
neonates
measured
All patients
Patients
< 37 wk
Patients
2.9*0.8 n = 141
27:07.P<~‘~~3.siO.8 . . n = 71
n = 70
P < 0.001
P
P50.14
3.3 rt0.8 ?I=110
3.1 zbO.8 n = 49
3.4*0.9 n=61 P i 0.05-----J
within
> 37 wk
30 min
3. Hofmann W, Schleich A, Schroeter D, Weidinger H, Wiest W. Der Einflusz von Beta-sympathomimetika und sog. Ca.++ - antagonist&her Hemmstoffe auf den menschlichen Herzmuskel in vitro. Virchows Arch (Path Anat) 1977; 373: 85-95. 4. Spelger G, Wiest W, Schlicker H, Kachel W, Puls F. Kardiale Parameter beim Neugeborenen nach Tokolyse. Klin Paediatr 1980; 192: 319-324. 5. Plieth M, Ache-Ebelt H, Fitzner R, Bein G, Karkut G. Untersuchung der Herzfunktion nach Langxeittokolyse mit Fenoterol und Isoxsuprin bei Neugeborenen. Z Geburtshilfe Perinatol 1980; 184: 275-282. 6. Freysz H, Willard D, Lehr A, Messer J, Boog G. A long term evaluation of infants who received a beta mimetic drug while in utero. J Perinat Med 1977; 5: 94-99. 7. Stormi M, Gigli C, Nordio S, Mandruazato GP. Effects of the treatment with ritodrine on the fetus and newborn baby. In: Bompiani A, Cosmi EV, Fischetti B, Gasparri F, Romanini C, eds. Recent Advances on Beta-mimetic Drugs in Obstetrics. Societa Editrice Universo 1977. 8. Epstein MF, Nicholls E, Stubblefield PG. Neonatal hypoglycemia after beta-sympathomimetic tocolytic therapy. J Pediatr 1979; 94: 449-453. 9. Brazy JE, Pupkin MJ. Effects of maternal isoxsuprine administration on preterm infants. J Pediatr 1979; 94: 444-448.
Clinical study of indomethacin for prevention of prematurity C. Sureau and P. Piovani, Unique 75674 Paris Cedex 14, France
Universitaire Baudelocque, 123 Bd. de Port-Royal,
Introduction
Prostaglandin inhibitors have been demonstrated to be effective in blocking uterine activity [l-3], but this effectiveness involves possible side-effects in the mother and the child, since prostaglandin inhibitors pass easily through the placenta. As far as the mother is concerned, the contra-indications for indomethacin such as digestive-tract disturbances are well known. Less clear is the situation concerning the fetus: experimental studies have shown possible noxious effects on the pulmonary circulation, leading to closure of the ductus arteriosus in utero. Some other deleterious actions have been described for the kidney, the digestive tract and the brain [4]. The results of these studies call for caution when using indomethacin. Nevertheless, definite contra-indications for the use of betamimetics [5] as well as intolerance or failure of this treatment have made it legitimate to try to evaluate the usefulness and consequences of the use of indomethacin. Clinical studies The first study conducted in the Baudelocque Hospital was done during the years
1978-1979 and the results were reported in 1980 [6]. This study comprised 84 patients, 82% of them with immediate tocolysis, a mean duration of treatment of 23 days (O-86), a mean total dose of 3742 mg (125-15,000), 35% recurrence of premature labour, 65% delivery later than 37 wk, and an average newborn weight of 2670 g (1240-3550), good maternal tolerance, but 3 cases of neonatal persistance of fetal circulation, with high pulmonary arterial pressure. One of these newborns died in the neonatal period. In these 3 cases the delivery occurred during the indomethatin treatment. A second study was conducted in 1980- 1982. Indomethacin was used only in cases