- Email: [email protected]
Carbohydrate intolerance in patients receiving oral tocolytics
Carbohydrate intolerance in patients receiving oral tocolytics Jeffrey L. Angel, MD, William F. O'Brien, MD, Robert A. Knuppel, MD, MPH, Walter J. Morales, MD, and Cynthia]. Sims, MD
Tampa, Florida This prospective study was designed to evaluate the effects on glucose metabolism of terbutaline used as an oral tocolytic agent. Eighty-six patients were studied when admitted for preterm labor from 24 to 35 weeks' gestation. After intravenous tocolysis, these patients were maintained on 5 mg of terbutaline every 4 or 6 hours. An oral 50 gm, 1-hour glucose challenge test was done 48 hours after terbutaline dosing began. All abnormal glucose chaHenge test results (;;.135 mg/dl) were followed by a standard 100 gm oral glucose tolerance test. Sixty-three percent (54 of 86) of the terbutaline group had an abnormal 1-hour screening result, which was significantly different than the 26.7% (23 of 86) observed in the control group (p < 0.001 ). The mean fasting blood sugar and 1-hour postchallenge values were significantly higher in the study than in the control group (p < 0.0001 ). Ten of 86 in the treated group (11.6%) and 2 of 86 in the control group (2.3%) with abnormal results met the criteria for gestational diabetes. These numbers achieve statistical significance at p < 0.05. This study shows a significant effect of oral terbutaline therapy on glucose tolerance during pregnancy. Patients receiving oral terbutalinetherapy for suppression of preterm labor should undergo screening for gestational diabetes. (AM J OssTET GYNECOL 1988;159:762-6.)
Key words: Terbutaline, tocolysis, glucose intolerance, diabetes Intravenous administration of 13-mimetics may lead to significant alterations in glucose metabolism with resultant maternal hyperglycemia. I. 2 These effects may be transient and related to diminished responsiveness in 13-adrenergic mechanisms affecting plasma glucose concentrations. I. 3 Long-term use of oral 13-mimetics is not uncommon for suppression of preterm labor. Significant alterations in glucose homeostasis with chronic ritodrine therapy have not been found in prior reports. •-6 The purpose of this study was to determine if terbutaline used as an oral tocolytic agent changes carbohydrate tolerance significantly. Subjects and methods The study population was derived from those patients admitted to Tampa General Hospital for preterm labor at 24 to 35 weeks' gestation from July 1, 1985 through October 1, 1987. Eighty-six patients were studied prospectively under the following protocol. Ritodrine sulfate (up to 350 f..Lg/min), magnesium sulfate (2 to 4 gm/hr), or a combination of the two were used for intravenous tocolysis. After successful intravenous tocolysis, patients began therapy with 5 mg of terbuFrom the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of South Florida College of Medicine. Presented by invitation at the Fiftieth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Palm Beach, Florida, january 10-13, 1988. Reprint requests: Jeffrey L. Angel, MD, 1 Davis Blvd., Suite 310, Tampa, FL 33606.
762
taline orally every 4 or 6 hours. The dosage interval was at the discretion of the attending physician. An oral 50 gm 1-hour glucose challenge test with the inclusion of a fasting blood sugar level determination was done at least 48 hours after the start of oral terbutaline dosing. Two 12 mg doses of betamethasone administered intramuscularly 24 hours apart were offered to patients with gestations at 27 to 34 weeks on admission. If glucocorticoids were used for enhancement of fetal lung maturity, the glucose challenge test was performed 96 hours after the last steroid dose. The control group consisted of 86 patients randomly selected from the university obstetric service who were routinely screened for gestational diabetes mellitus between 27 and 32 weeks' gestation. These control patients were matched to the study patients for age, weight, and parity. A !-hour serum glucose level of ;;;,J35 mg/dl was considered to be abnormal and was followed by a standard 100 gm oral glucose tolerance test. 7 The diagnosis of gestational diabetes required the presence of at least two of three abnormal glucose values after the 100 gm challenge. Previously reported oral glucose tolerance test criteria were used, with values of 190, 165, and 145 mg/dl plasma glucose considered the upper limits of normal for 1-, 2-, and 3hour values, respectively." Patients with 1-hour postchallenge values ;;;,200 mg/dl did not undergo formal 3-hour oral glucose tolerance tests and were classified as having gestational diabetes. 9 Patients with established diabetes and patients with a known predisposition to diabetes mellitus were excluded from study.
Terbutaline-associated glucose intolerance
Volume 159 Number 3
Table I. Demographic data Variable
Age (yr) Race White Black Hispanic Parity EGA (wk) Weight (lb)
763
Table II. Abnormal 1-hour values after the glucose challenge test
Study group (n = 85)
Control group (n = 91)
p Value*
23.6 :±: 5.4
25.4 :±: 4.5
0.02
46 27 15 1.2 :±: 1.0 30.7 :±: 2.5 146.2 :±: 20.3
51 38 28 1.3 :±: 1.1 29.9 :±: 4.2 147.4 :±: 20.3
NS NS NS NS NS NS
Data are the mean :±: SD. EGA, Estimated gestational age when glucose challenge test done. *Student unpaired t test.
The statistical analysis of the data was done by x2 analysis, Student t test, or Fisher exact test where appropriate. A p value of <0.05 was considered significant.
Results Glucose challenge test results from 86 study patients and 86 control patients were available for analysis. Results from 37 and 17 oral glucose tolerance tests were analyzed from the study and control groups, respectively. Demographic comparisons of both groups are noted with respect to age, race, gestational age at time of the glucose challenge test, parity, and weight. With the exception of maternal age, there were no significant differences noted for these variables between the two groups (Table 1). Abnormal 1-hour values after the glucose challenge test were noted in 54 (62.8%) of the study group, which was statistically different from the 23 (26.7%) observed in the control group (p < 0.001; Table II). In addition, 10 (11.6%) in the study group versus 2 (2.3%) in the control group met criteria for the diagnosis of gestational diabetes. Mean fasting blood sugar levels and !-hour postchallenge glucose values were significantly higher in the study group (Table III). As noted in Table IV, once glucose intolerance was identified by the screening test, the percentage of abnormal oral glucose tolerance test results was not significantly different between the two groups. The postchallenge plasma glucose values for the oral glucose tolerance tests are listed in Table V, with no statistical difference between the study and control groups. The incidence of abnormal glucose challenge tests when every-4-hour versus every-6-hour terbutaline doses were compared was not statistically different. There was no statistical difference noted for abnormal glucose challenge test results between the patients who had received steroids and those who had not (Table VI).
Study group (n = 86)
Abnormal GCT* Gestational diabetes*
No.j
%
54 10
62.8 11.6
Control group (n = 86) No. j
23 2
%
p Value
26.7 2.3
<0.001 t <0.04§
*Abnormal glucose challenge test (GCT) = hour postchallenge value ;:;,:J35 mg/dl plasma glucose. tx2 analysis. *Gestational diabetes = at least two postchallenge values abnormal or 1 hour postchallenge ::;,:200 mg/dl (two patients) without an oral glucose tolerance test. §Fisher exact test.
Plasma glucose (mg!dl)
p Value*
Fasting 93.7 :±: 19.7 78.6 :±: 7.8 <0.0001 1 hr post challenge 153.0 :±: 35.9 116.5 :±: 27.3 <0.0001 Data are the mean :±: SD. *Student t test.
Comment The metabolic consequences of oral(3-mimetic therapy in man are not well studied. Intravenous 13-mimetic therapy may lead to significant, albeit transient, hyperglycemia in the pregnant patient.'· 3 During the third trimester of pregnancy, there is an elevation of basal insulin concentration by as much as 50% to 80%.' 0 · " Glucose intolerance in the last half of an otherwise normal pregnancy has been well documented.' 2 · 13 The combined effects of a postreceptor insulin defect and insulin antagonism by placental hormones result in modest insulin resistance and mobilization of hepatic stores of glycogen with an increase in hepatic glucose production and stress on normal glucose production in the fed state.' 4 · 15 Thus there is appropriate concern with the use of (3-mimetic therapy, which through stimulation of the adenylate cyclase system increases cyclic adenosine monophosphate production and hepatic glycogenolysis, with resultant hyperglycemia. 16 A current review of the English literature provides evidence that oral ritodrine and oral salbutamol do not significantly alter glucose tolerance during pregnancy.'· 6 · ' 7 · '" In 1978, Hastwell and Lamberti 7 investigated six patients treated with oral salbutamol and found no conformity in glucose values. Four of these patients were in the early second trimester and none underwent glucose challenges. In 1976 Blouin eta!.' reported no evidence
764 Angel et al.
September !988 Am .J Obstet Gynecol
Table IV. Glucose tolerance test results Oral glucose tolerance test
Study group No.
Normal Abnormal
_!
Total
37
I
29
Control group
%
No.
78.4 21.6
15 ~ 17
I
%
p Value
88.2 ll.8
NS* NSt
All oral glucose tolerance tests followed an abnormal glucose challenge test. *)( 2 analysis. t Fisher exact test.
Table V. Glucose tolerance test plasma glucose values (mg/dl) Time (hr)
0 I
2 3
Study group (n = 37)
90.0 159.4 147.7 124.4
± ± ± ±
15.5 29.8 40.3 32.2
Control group (n = 17)
83.9 162.5 136.4 107.4
± ± ± ±
6.8 32.9 31.9 24.6
p Value*
NS NS NS NS
Data are the mean ± SD. *Student t test.
of a decline in carbohydrate tolerance in eight pregnant patients who underwent glucose tolerance testing in the third trimester while taking oral ritodrine. These patients were receiving 40 mg/day of ritodrine and were tested with a 50 gm oral glucose tolerance test. Schreyer et al., 6 in 1981, tested 10 patients taking 80 mg/day of oral ritodrine and showed that this did not influence glucose levels. These patients were tested with a 25 gm intravenous glucose tolerance test. In 1981, Wager et al. 18 reported the effects of oral salbutamol on glucose tolerance in 10 patients who did not have diabetes during the third trimester of pregnancy. These patients received 4 mg of salbutamol orally, followed by an intravenous glucose tolerance test. Glucose values at 60, 90, and 120 minutes were significantly increased over basal levels in comparison with a control group who did not receive salbutamol. It is hazardous to extrapolate data from any of these four studies to present obstetrics and tocolytic therapy, in which higher drug dosages and more frequent dosing intervals are used. In addition, there is a large variation in the glucose challenge testing schemes used in each study. More recently, Main et al! reported on 42 patients receiving 120 mg/day of oral ritodrine during the third trimester of pregnancy. Patients were tested with a standard glucose challenge test. An oral glucose tolerance test was obtained if the 1-hour glucose level ws ;;..135 mg I dl from the glucose challenge test. The patients receiving oral ritodrine had a 21 % incidence of abnormal 1-hour screening values as compared with 20% in their control population. In addition, none of
Table VI. Effect of dose of terbutaline and use of corticosteroids on the incidence of abnormal glucose challenge test results Abnormal glucose challenge test No.
Terbutaline 30 mg/day 20 mg/day Steroids Yes No
I
%
5
48
55.6 63.2
23 30
43.4 56.6
p Value
NS NS
the study population had an abnormal oral glucose tolerance test result. In 1987, Main et al. 19 reported on the effects of chronic terbutaline tocolytic therapy. An increase in positive screening results (63%) as well as abnormal oral glucose tolerance tests after abnormal screenings (52.6%) were noted. The results of this prospective study with oral terbutaline contrast to an extent those reported by Main et al. 19 Consistent with their report, 62.8% of our patients treated with oral terbutaline had an abnormal 1hour glucose challenge test value, which was significantly different from both the 26.7% observed in our control population and the 20.8% positive rate observed in low-risk patients ;;..25 years old in the original study of O'Sullivan et al. 20 Statistically significant elevations in fasting blood sugar levels and 1-hour postchallenge glucose values when compared with the control group provide sound evidence for augmentation of glucose intolerance by oral terbutaline used in the third trimester of pregnancy. The incidence of gestational diabetes as demonstrated by an abnormal oral glucose tolerance test result was 11.6% (10 of 86) in the terbutaline group versus 2.3% (2 of 86) in the control group. These numbers achieve statistical significance and differ sharply from the 33.1% incidence of gestational diabetes reported in the terbutaline group by Main et a!. 19 In the present study, once glucose intolerance was revealed by an abnormal glucose challenge test, whether endogenous or related to 13-mimetic therapy, there was no statistical difference in the number of abnormal oral glucose tolerance test results (11.8% and 21.6%, respectively). In addition, comparison of postchallenge oral glucose tolerance test glucose values revealed no difference between the control and terbutaline groups. These percentages are in accordance with the 19% abnormal oral glucose tolerance test results reported by O'Sullivan et al! 0 in 1973. Ritodrine and terbutaline are 132 -specific agonists that, in addition to inhibition of uterine contractility, will with intravenous administration produce maternal hyperglycemia. The pharmacophysiologic mechanisms
Volume 159 Number 3
to explain the glucose intolerance provoked by oral
terbutaline in this study as compared with previous reports• with ritodrine are at present co~ectural. Caritis et al. 2 ' provide data that supports intravenous terbutaline as a more potent stimulator of hyperglycemia than ritodrine. The dosage of terbutaline we used is not considered excessive and is currently used in many centers throughout the United States; it is considered to be the potency equivalent of 120 mg of ritodrine per day. 22 When compared with a control population, there was a 2.4-fold increase in the number of positive screening results and five times the frequency of gestational diabetes in our study population. Patients receiving oral terbutaline should undergo screening for gestational diabetes mellitus. An oral glucose tolerance test should be obtained in patients with abnormal screens. Two deficiencies in our study exist. First, preterbutaline glucose challenge testing would have been helpful in identifying those patients with a predisposition for glucose intolerance. Second, glucose tolerance testing after discontinuation of oral tocolytic therapy to evaluate for resolution of the carbohydrate intolerance would have allowed for the study patients to serve as their own controls. One may want to consider the apparent difference in the effects of terbutaline and ritodrine on carbohydrate tolerance in pregnancy and the proposed diabetogenicity of terbutaline when choosing a specific oral tocolytic agent. One proposal would be to start with terbutaline; if an abnormal glucose tolerance test result is found, then switch to oral ritodrine. A cost-effective analysis comparing oral ritodrine and oral terbutaline was not within the realm of this study, but would be of great interest for future investigation. REFERENCES 1. Bassett JM, Burks AH, Levine DH, et al. Maternal and fetal metabolic effects of prolonged ritodrine infusion. Obstet Gynecol 1985;66:755. 2. Caritis SN, Lin LS, Toig G, Wong LK. Pharmacodynamics of ritodrine in pregnant women during preterm labor. AMJ 0BSTET GYNECOL 1983;147:752. 3. Young DC, Toofanian A, Leveno KJ. Potassium and glucose concentrations without treatment during ritodrine tocolysis. AMj 0BSTET GYNECOL 1983;145:105. 4. Main DM, Main EK, Strong SE, Gabbe SG. The effect of oral ritodrine therapy on glucose tolerance in pregnancy. AMj 0BSTET GYNECOL 1985;152:1031. 5. Blouin D, Murray MAF, Beard RW. The effect of oral ritodrine on maternal and fetal carbohydrate metabolism. Br J Obstet Gynaecol1976;83:7ll. 6. Schreyer P, Caspi E, Snir E, et al. Metabolic effects of intramuscular and oral administration of ritodrine in pregnancy. Obstet Gynecol 1981;57:730. 7. Carpenter MW, Causton DR. Criteria for screening tests for gestational diabetes. AM j 0BSTET GYNECOL 1982; 144:768. 8. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28:1039.
Terbutaline-associated glucose intolerance
765
9. Marquette GP, Klein VR, Repke JT, Niebyl JR. Costeffective criteria for glucose screening. Obstet Gynecol 1985;66:181. 10. Phelps RL, Metzger BE, Freinkel N. Carbohydrate metabolism in pregnancy XVIII. Diurnal profiles of plasma glucose, insulin, free fatty acids, triglycerides, cholesterol and individual amino acids in late normal pregnancy. AM j 0BSTET GYNECOL 1981;140:730. 11. Kuhl C. Glucose metabolism during and after pregnancy in normal and gestational diabetic woman: influence of normal pregnancy on serum glucose and insulin concentration during basal fasting conditions and after a challenge with glucose. Acta Endocrinol 1975;79:709. 12. Picard C, Ooms HA, Balasse E, et al. Effect of normal pregnancy on glucose assimilation, insulin and nonesterified fatty acids levels. Diabetologia 1968;4:16. 13. Knapp RH, Montes, Childs M, et al. Metabolic adjustments in normal and diabetic pregnancies. Clin Obstet Gynecol 1981;24:21. 14. Pvavilai G, Drobny EC, Domont LA, et al. Insulin receptors and insulin resistance in human pregnancy: evidence for a post receptor defect in insulin action. J Clin Endocrinol Metab 1982;54:24 7. 15. Hollingsworth DR. Maternal metabolism in normal pregnancy and pregnancy complicated by diabetes mellitus. Clin Obstet Gynecol 1985;28:457. 16. Roberts JM. Current understanding of pharmacologic mechanisms in the prevention of preterm birth. Clin Obstet Gynecol 1984;27:592. 17. Hastwell G, Lambert BE. The effect of oral salbutamol on serum potassium and blood sugar. Br ] Obstet Gynaecol 1978;85:767. 18. Wager J, Fredholm BB, Lunell NO, et al. Metabolic and circulatory effects of oral salbutamol in the third trimester of pregnancy in diabetic and non-diabetic women. Br ] Obstet Gynaecol 1981 ;88:352. 19. Main EK, Main DM, Gabbe SG. Chronic oral terbutaline tocolytic therapy is associated with maternal glucose intolerance. AM J 0BSTET GYNECOL 1987; 157:644. 20. O'Sullivan JB, Charles D, Mahan CM, et al. Screening criteria for high· risk gestational diabetic patients. AM J 0BSTET GYNECOL 1973; 116:895. 21. Caritis SN, Toig G, Reddinger LA, et al. A double blind study comparing ritodrine and terbutaline in the treatment of preterm labor. AM J 0BSTET GYNECOL 1984; 150:7. 22. Creasy RK, Resnik R. Maternal-fetal medicine. Principles and practice. Philadelphia: WB Saunders, 1984:428.
Discussion DR. JoHN H. BAKER, Williamsburg, Virginia (Official Guest). Oral terbutaline appears to be unique among the oralj3-mimetic agents used for tocolysis in its ability to cause carbohydrate intolerance in pregnant patients. Whether this is due to a relative lack of 132 -receptor agonist selectivity on the part of terbutaline or to some other tissue receptor variability remains to be determined. In screening the general obstetric population, we expect to find a 15% to 25% incidence of abnormal 1-hour glucose screening results (glucose challenge test or O'Sullivan screen). There will be a 15% to 20% incidence of abnormal 3-hour glucose tolerance test results in the group with abnormal screening tests, giving us a 2% to 5% incidence of gestational diabetes in the tested patients. This contrasts dramatically and significantly with the 62.4% abnormal 1-hour results and 9.4% abnormal 3-hour glucose tolerance test results in the study of Dr. Angel and his coworkers, whose pa-
766
Angel et al.
ttents were tested after 48 hours of oral terbutaline therapy according to a standard dosage schedule. The data presented by Angel et al. are clear, well stated, and essentially confirm the findings of a similar study presented to the Society of Perinatal Obstetricians in February 1987. 1 Brown and Tejani! reporting on 23 patients receiving oral terbutaline therapy for ~3 weeks after initial successful tocolysis with ethanol, described no significant adverse maternal or fetal effects, but blood sugar status was not mentioned. Obstetric care providers are highly sensitized now to the prevention of preterm birth, both through prenatal risk assessment and aggressive management of preterm labor. Oral tocolytic agents, particularly terbutaline, are being used quite liberally in offices and clinics everywhere. Do the findings of Angel et al. pose a serious challenge to our abiding philosophic dictum of "first do no harm?" My personal bias as a private practitioner is that we can manage maternal hyperglycemia far easier than we can the complications of preterm labor and delivery. No discussion would be complete without a few questions for the presenter. Do you think that glucose screening after 48 hours of oral terbutaline, as compared with Main et al., 1 who screened after 7 or more days, would account for the difference in gestational diabetes in the two study groups (9.4% versus 33%)? Do you have any comments about the 16 patients in the study group for whom 3-hour oral glucose tolerance test results are not known? Were there any patients in the study group who were screened for gestational diabetes before receiving oral terbutaline, thereby becoming their own controls? Was terbutaline stopped when gestational diabetes was diagnosed in the study group? Also, was there any difference between the patients with gestational diabetes in the two groups in terms of the need for insulin therapy? Was there any difference in adverse maternal or perinatal outcomes between the two groups? Finally, do you have any recommendations regarding chronic oral tocolytic therapy in patients with diabetes? REFERENCES I. Main EK, Main DM, Gabbe SG. Chronic oral terbutaline
tocolytic therapy is associated with maternal glucose intolerance. AM J 0BSTET GYNECOL 1987;157:644. 2. Brown SM, Tejani NA. Terbutaline sulfate in the prevention of recurrence of premature labor. Obstet Gynecol 1981;57:22.
September 1988 Am J Obstet Gynecol
DR. ANGEL (Closing). Our protocol required glucose tolerance testing after at least 48 hours of tocolytic dosing. Some of the patients were tested at 3 and 4 days. Due to the logistics of the testing protocol, in that patients went home and came back to the clinic, many were tested after a 7-day period. I do not believe that there is an augmentation of glucose intolerance that has a dose- or a time-related frequency, such that a week would result in a more common or a more frequent incidence of gestational diabetes. In fact, we did retest some women later, and the time lag did not make any difference. However, Main's group did look at the effects on glucose challenge test results 7 days after the start of terbutaline dosing. I think what should be done are serial tests, but I have found very few patients who will agree to serial 3-hour glucose tolerance testing. We did not test any patients in our study group before they went into preterm labor. Main's group did, as a method partially related to the retrospective nature of their study. These patients were in the high-risk clinic and they were tested per protocol; then, if they had preterm labor, they were tested again. We did not have that setup. However, in those patients with gestational diabetes while taking terbutaline who achieved 36 weeks' gestation and progressed to where we could discontinue the tocolytic drugs, we did so and retested them. The numbers were not sufficient to add up to something we could present. There were three patients, one of whom was insulin dependent, and she no longer required insulin therapy. The two other patients who controlled their disease by diet no longer required dietary therapy. These were the controls that we had. The diabetes was not harder to control in the terbutaline group versus the control group. In the patients who had gestational diabetes, the numbers were small but the perinatal outcomes, primarily birth weight and cesarean section rates, were the same. As to patients with overt diabetes and the agents for tocolysis, I do not have the answer. We proceed with ritodrine or terbutaline and watch the blood sugar levels closely, readjusting insulin therapy as necessary. Dr. Morrison has some data supporting magnesium gluconate as an oral tocolytic, and I think we need some more information on that, as is the case with the magnesium oxide data from Dr. Petrie. My closing statement would be that I am not sure where to go as far as testing terbutaline versus ritodrine. If we use terbutaline, we test all of our patients for gestational diabetes; if a positive glucose tolerance test results, they are treated as if the diabetes were endogenous.
Tampa, Florida This prospective study was designed to evaluate the effects on glucose metabolism of terbutaline used as an oral tocolytic agent. Eighty-six patients were studied when admitted for preterm labor from 24 to 35 weeks' gestation. After intravenous tocolysis, these patients were maintained on 5 mg of terbutaline every 4 or 6 hours. An oral 50 gm, 1-hour glucose challenge test was done 48 hours after terbutaline dosing began. All abnormal glucose chaHenge test results (;;.135 mg/dl) were followed by a standard 100 gm oral glucose tolerance test. Sixty-three percent (54 of 86) of the terbutaline group had an abnormal 1-hour screening result, which was significantly different than the 26.7% (23 of 86) observed in the control group (p < 0.001 ). The mean fasting blood sugar and 1-hour postchallenge values were significantly higher in the study than in the control group (p < 0.0001 ). Ten of 86 in the treated group (11.6%) and 2 of 86 in the control group (2.3%) with abnormal results met the criteria for gestational diabetes. These numbers achieve statistical significance at p < 0.05. This study shows a significant effect of oral terbutaline therapy on glucose tolerance during pregnancy. Patients receiving oral terbutalinetherapy for suppression of preterm labor should undergo screening for gestational diabetes. (AM J OssTET GYNECOL 1988;159:762-6.)
Key words: Terbutaline, tocolysis, glucose intolerance, diabetes Intravenous administration of 13-mimetics may lead to significant alterations in glucose metabolism with resultant maternal hyperglycemia. I. 2 These effects may be transient and related to diminished responsiveness in 13-adrenergic mechanisms affecting plasma glucose concentrations. I. 3 Long-term use of oral 13-mimetics is not uncommon for suppression of preterm labor. Significant alterations in glucose homeostasis with chronic ritodrine therapy have not been found in prior reports. •-6 The purpose of this study was to determine if terbutaline used as an oral tocolytic agent changes carbohydrate tolerance significantly. Subjects and methods The study population was derived from those patients admitted to Tampa General Hospital for preterm labor at 24 to 35 weeks' gestation from July 1, 1985 through October 1, 1987. Eighty-six patients were studied prospectively under the following protocol. Ritodrine sulfate (up to 350 f..Lg/min), magnesium sulfate (2 to 4 gm/hr), or a combination of the two were used for intravenous tocolysis. After successful intravenous tocolysis, patients began therapy with 5 mg of terbuFrom the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of South Florida College of Medicine. Presented by invitation at the Fiftieth Annual Meeting of the South Atlantic Association of Obstetricians and Gynecologists, Palm Beach, Florida, january 10-13, 1988. Reprint requests: Jeffrey L. Angel, MD, 1 Davis Blvd., Suite 310, Tampa, FL 33606.
762
taline orally every 4 or 6 hours. The dosage interval was at the discretion of the attending physician. An oral 50 gm 1-hour glucose challenge test with the inclusion of a fasting blood sugar level determination was done at least 48 hours after the start of oral terbutaline dosing. Two 12 mg doses of betamethasone administered intramuscularly 24 hours apart were offered to patients with gestations at 27 to 34 weeks on admission. If glucocorticoids were used for enhancement of fetal lung maturity, the glucose challenge test was performed 96 hours after the last steroid dose. The control group consisted of 86 patients randomly selected from the university obstetric service who were routinely screened for gestational diabetes mellitus between 27 and 32 weeks' gestation. These control patients were matched to the study patients for age, weight, and parity. A !-hour serum glucose level of ;;;,J35 mg/dl was considered to be abnormal and was followed by a standard 100 gm oral glucose tolerance test. 7 The diagnosis of gestational diabetes required the presence of at least two of three abnormal glucose values after the 100 gm challenge. Previously reported oral glucose tolerance test criteria were used, with values of 190, 165, and 145 mg/dl plasma glucose considered the upper limits of normal for 1-, 2-, and 3hour values, respectively." Patients with 1-hour postchallenge values ;;;,200 mg/dl did not undergo formal 3-hour oral glucose tolerance tests and were classified as having gestational diabetes. 9 Patients with established diabetes and patients with a known predisposition to diabetes mellitus were excluded from study.
Terbutaline-associated glucose intolerance
Volume 159 Number 3
Table I. Demographic data Variable
Age (yr) Race White Black Hispanic Parity EGA (wk) Weight (lb)
763
Table II. Abnormal 1-hour values after the glucose challenge test
Study group (n = 85)
Control group (n = 91)
p Value*
23.6 :±: 5.4
25.4 :±: 4.5
0.02
46 27 15 1.2 :±: 1.0 30.7 :±: 2.5 146.2 :±: 20.3
51 38 28 1.3 :±: 1.1 29.9 :±: 4.2 147.4 :±: 20.3
NS NS NS NS NS NS
Data are the mean :±: SD. EGA, Estimated gestational age when glucose challenge test done. *Student unpaired t test.
The statistical analysis of the data was done by x2 analysis, Student t test, or Fisher exact test where appropriate. A p value of <0.05 was considered significant.
Results Glucose challenge test results from 86 study patients and 86 control patients were available for analysis. Results from 37 and 17 oral glucose tolerance tests were analyzed from the study and control groups, respectively. Demographic comparisons of both groups are noted with respect to age, race, gestational age at time of the glucose challenge test, parity, and weight. With the exception of maternal age, there were no significant differences noted for these variables between the two groups (Table 1). Abnormal 1-hour values after the glucose challenge test were noted in 54 (62.8%) of the study group, which was statistically different from the 23 (26.7%) observed in the control group (p < 0.001; Table II). In addition, 10 (11.6%) in the study group versus 2 (2.3%) in the control group met criteria for the diagnosis of gestational diabetes. Mean fasting blood sugar levels and !-hour postchallenge glucose values were significantly higher in the study group (Table III). As noted in Table IV, once glucose intolerance was identified by the screening test, the percentage of abnormal oral glucose tolerance test results was not significantly different between the two groups. The postchallenge plasma glucose values for the oral glucose tolerance tests are listed in Table V, with no statistical difference between the study and control groups. The incidence of abnormal glucose challenge tests when every-4-hour versus every-6-hour terbutaline doses were compared was not statistically different. There was no statistical difference noted for abnormal glucose challenge test results between the patients who had received steroids and those who had not (Table VI).
Study group (n = 86)
Abnormal GCT* Gestational diabetes*
No.j
%
54 10
62.8 11.6
Control group (n = 86) No. j
23 2
%
p Value
26.7 2.3
<0.001 t <0.04§
*Abnormal glucose challenge test (GCT) = hour postchallenge value ;:;,:J35 mg/dl plasma glucose. tx2 analysis. *Gestational diabetes = at least two postchallenge values abnormal or 1 hour postchallenge ::;,:200 mg/dl (two patients) without an oral glucose tolerance test. §Fisher exact test.
Plasma glucose (mg!dl)
p Value*
Fasting 93.7 :±: 19.7 78.6 :±: 7.8 <0.0001 1 hr post challenge 153.0 :±: 35.9 116.5 :±: 27.3 <0.0001 Data are the mean :±: SD. *Student t test.
Comment The metabolic consequences of oral(3-mimetic therapy in man are not well studied. Intravenous 13-mimetic therapy may lead to significant, albeit transient, hyperglycemia in the pregnant patient.'· 3 During the third trimester of pregnancy, there is an elevation of basal insulin concentration by as much as 50% to 80%.' 0 · " Glucose intolerance in the last half of an otherwise normal pregnancy has been well documented.' 2 · 13 The combined effects of a postreceptor insulin defect and insulin antagonism by placental hormones result in modest insulin resistance and mobilization of hepatic stores of glycogen with an increase in hepatic glucose production and stress on normal glucose production in the fed state.' 4 · 15 Thus there is appropriate concern with the use of (3-mimetic therapy, which through stimulation of the adenylate cyclase system increases cyclic adenosine monophosphate production and hepatic glycogenolysis, with resultant hyperglycemia. 16 A current review of the English literature provides evidence that oral ritodrine and oral salbutamol do not significantly alter glucose tolerance during pregnancy.'· 6 · ' 7 · '" In 1978, Hastwell and Lamberti 7 investigated six patients treated with oral salbutamol and found no conformity in glucose values. Four of these patients were in the early second trimester and none underwent glucose challenges. In 1976 Blouin eta!.' reported no evidence
764 Angel et al.
September !988 Am .J Obstet Gynecol
Table IV. Glucose tolerance test results Oral glucose tolerance test
Study group No.
Normal Abnormal
_!
Total
37
I
29
Control group
%
No.
78.4 21.6
15 ~ 17
I
%
p Value
88.2 ll.8
NS* NSt
All oral glucose tolerance tests followed an abnormal glucose challenge test. *)( 2 analysis. t Fisher exact test.
Table V. Glucose tolerance test plasma glucose values (mg/dl) Time (hr)
0 I
2 3
Study group (n = 37)
90.0 159.4 147.7 124.4
± ± ± ±
15.5 29.8 40.3 32.2
Control group (n = 17)
83.9 162.5 136.4 107.4
± ± ± ±
6.8 32.9 31.9 24.6
p Value*
NS NS NS NS
Data are the mean ± SD. *Student t test.
of a decline in carbohydrate tolerance in eight pregnant patients who underwent glucose tolerance testing in the third trimester while taking oral ritodrine. These patients were receiving 40 mg/day of ritodrine and were tested with a 50 gm oral glucose tolerance test. Schreyer et al., 6 in 1981, tested 10 patients taking 80 mg/day of oral ritodrine and showed that this did not influence glucose levels. These patients were tested with a 25 gm intravenous glucose tolerance test. In 1981, Wager et al. 18 reported the effects of oral salbutamol on glucose tolerance in 10 patients who did not have diabetes during the third trimester of pregnancy. These patients received 4 mg of salbutamol orally, followed by an intravenous glucose tolerance test. Glucose values at 60, 90, and 120 minutes were significantly increased over basal levels in comparison with a control group who did not receive salbutamol. It is hazardous to extrapolate data from any of these four studies to present obstetrics and tocolytic therapy, in which higher drug dosages and more frequent dosing intervals are used. In addition, there is a large variation in the glucose challenge testing schemes used in each study. More recently, Main et al! reported on 42 patients receiving 120 mg/day of oral ritodrine during the third trimester of pregnancy. Patients were tested with a standard glucose challenge test. An oral glucose tolerance test was obtained if the 1-hour glucose level ws ;;..135 mg I dl from the glucose challenge test. The patients receiving oral ritodrine had a 21 % incidence of abnormal 1-hour screening values as compared with 20% in their control population. In addition, none of
Table VI. Effect of dose of terbutaline and use of corticosteroids on the incidence of abnormal glucose challenge test results Abnormal glucose challenge test No.
Terbutaline 30 mg/day 20 mg/day Steroids Yes No
I
%
5
48
55.6 63.2
23 30
43.4 56.6
p Value
NS NS
the study population had an abnormal oral glucose tolerance test result. In 1987, Main et al. 19 reported on the effects of chronic terbutaline tocolytic therapy. An increase in positive screening results (63%) as well as abnormal oral glucose tolerance tests after abnormal screenings (52.6%) were noted. The results of this prospective study with oral terbutaline contrast to an extent those reported by Main et al. 19 Consistent with their report, 62.8% of our patients treated with oral terbutaline had an abnormal 1hour glucose challenge test value, which was significantly different from both the 26.7% observed in our control population and the 20.8% positive rate observed in low-risk patients ;;..25 years old in the original study of O'Sullivan et al. 20 Statistically significant elevations in fasting blood sugar levels and 1-hour postchallenge glucose values when compared with the control group provide sound evidence for augmentation of glucose intolerance by oral terbutaline used in the third trimester of pregnancy. The incidence of gestational diabetes as demonstrated by an abnormal oral glucose tolerance test result was 11.6% (10 of 86) in the terbutaline group versus 2.3% (2 of 86) in the control group. These numbers achieve statistical significance and differ sharply from the 33.1% incidence of gestational diabetes reported in the terbutaline group by Main et a!. 19 In the present study, once glucose intolerance was revealed by an abnormal glucose challenge test, whether endogenous or related to 13-mimetic therapy, there was no statistical difference in the number of abnormal oral glucose tolerance test results (11.8% and 21.6%, respectively). In addition, comparison of postchallenge oral glucose tolerance test glucose values revealed no difference between the control and terbutaline groups. These percentages are in accordance with the 19% abnormal oral glucose tolerance test results reported by O'Sullivan et al! 0 in 1973. Ritodrine and terbutaline are 132 -specific agonists that, in addition to inhibition of uterine contractility, will with intravenous administration produce maternal hyperglycemia. The pharmacophysiologic mechanisms
Volume 159 Number 3
to explain the glucose intolerance provoked by oral
terbutaline in this study as compared with previous reports• with ritodrine are at present co~ectural. Caritis et al. 2 ' provide data that supports intravenous terbutaline as a more potent stimulator of hyperglycemia than ritodrine. The dosage of terbutaline we used is not considered excessive and is currently used in many centers throughout the United States; it is considered to be the potency equivalent of 120 mg of ritodrine per day. 22 When compared with a control population, there was a 2.4-fold increase in the number of positive screening results and five times the frequency of gestational diabetes in our study population. Patients receiving oral terbutaline should undergo screening for gestational diabetes mellitus. An oral glucose tolerance test should be obtained in patients with abnormal screens. Two deficiencies in our study exist. First, preterbutaline glucose challenge testing would have been helpful in identifying those patients with a predisposition for glucose intolerance. Second, glucose tolerance testing after discontinuation of oral tocolytic therapy to evaluate for resolution of the carbohydrate intolerance would have allowed for the study patients to serve as their own controls. One may want to consider the apparent difference in the effects of terbutaline and ritodrine on carbohydrate tolerance in pregnancy and the proposed diabetogenicity of terbutaline when choosing a specific oral tocolytic agent. One proposal would be to start with terbutaline; if an abnormal glucose tolerance test result is found, then switch to oral ritodrine. A cost-effective analysis comparing oral ritodrine and oral terbutaline was not within the realm of this study, but would be of great interest for future investigation. REFERENCES 1. Bassett JM, Burks AH, Levine DH, et al. Maternal and fetal metabolic effects of prolonged ritodrine infusion. Obstet Gynecol 1985;66:755. 2. Caritis SN, Lin LS, Toig G, Wong LK. Pharmacodynamics of ritodrine in pregnant women during preterm labor. AMJ 0BSTET GYNECOL 1983;147:752. 3. Young DC, Toofanian A, Leveno KJ. Potassium and glucose concentrations without treatment during ritodrine tocolysis. AMj 0BSTET GYNECOL 1983;145:105. 4. Main DM, Main EK, Strong SE, Gabbe SG. The effect of oral ritodrine therapy on glucose tolerance in pregnancy. AMj 0BSTET GYNECOL 1985;152:1031. 5. Blouin D, Murray MAF, Beard RW. The effect of oral ritodrine on maternal and fetal carbohydrate metabolism. Br J Obstet Gynaecol1976;83:7ll. 6. Schreyer P, Caspi E, Snir E, et al. Metabolic effects of intramuscular and oral administration of ritodrine in pregnancy. Obstet Gynecol 1981;57:730. 7. Carpenter MW, Causton DR. Criteria for screening tests for gestational diabetes. AM j 0BSTET GYNECOL 1982; 144:768. 8. National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28:1039.
Terbutaline-associated glucose intolerance
765
9. Marquette GP, Klein VR, Repke JT, Niebyl JR. Costeffective criteria for glucose screening. Obstet Gynecol 1985;66:181. 10. Phelps RL, Metzger BE, Freinkel N. Carbohydrate metabolism in pregnancy XVIII. Diurnal profiles of plasma glucose, insulin, free fatty acids, triglycerides, cholesterol and individual amino acids in late normal pregnancy. AM j 0BSTET GYNECOL 1981;140:730. 11. Kuhl C. Glucose metabolism during and after pregnancy in normal and gestational diabetic woman: influence of normal pregnancy on serum glucose and insulin concentration during basal fasting conditions and after a challenge with glucose. Acta Endocrinol 1975;79:709. 12. Picard C, Ooms HA, Balasse E, et al. Effect of normal pregnancy on glucose assimilation, insulin and nonesterified fatty acids levels. Diabetologia 1968;4:16. 13. Knapp RH, Montes, Childs M, et al. Metabolic adjustments in normal and diabetic pregnancies. Clin Obstet Gynecol 1981;24:21. 14. Pvavilai G, Drobny EC, Domont LA, et al. Insulin receptors and insulin resistance in human pregnancy: evidence for a post receptor defect in insulin action. J Clin Endocrinol Metab 1982;54:24 7. 15. Hollingsworth DR. Maternal metabolism in normal pregnancy and pregnancy complicated by diabetes mellitus. Clin Obstet Gynecol 1985;28:457. 16. Roberts JM. Current understanding of pharmacologic mechanisms in the prevention of preterm birth. Clin Obstet Gynecol 1984;27:592. 17. Hastwell G, Lambert BE. The effect of oral salbutamol on serum potassium and blood sugar. Br ] Obstet Gynaecol 1978;85:767. 18. Wager J, Fredholm BB, Lunell NO, et al. Metabolic and circulatory effects of oral salbutamol in the third trimester of pregnancy in diabetic and non-diabetic women. Br ] Obstet Gynaecol 1981 ;88:352. 19. Main EK, Main DM, Gabbe SG. Chronic oral terbutaline tocolytic therapy is associated with maternal glucose intolerance. AM J 0BSTET GYNECOL 1987; 157:644. 20. O'Sullivan JB, Charles D, Mahan CM, et al. Screening criteria for high· risk gestational diabetic patients. AM J 0BSTET GYNECOL 1973; 116:895. 21. Caritis SN, Toig G, Reddinger LA, et al. A double blind study comparing ritodrine and terbutaline in the treatment of preterm labor. AM J 0BSTET GYNECOL 1984; 150:7. 22. Creasy RK, Resnik R. Maternal-fetal medicine. Principles and practice. Philadelphia: WB Saunders, 1984:428.
Discussion DR. JoHN H. BAKER, Williamsburg, Virginia (Official Guest). Oral terbutaline appears to be unique among the oralj3-mimetic agents used for tocolysis in its ability to cause carbohydrate intolerance in pregnant patients. Whether this is due to a relative lack of 132 -receptor agonist selectivity on the part of terbutaline or to some other tissue receptor variability remains to be determined. In screening the general obstetric population, we expect to find a 15% to 25% incidence of abnormal 1-hour glucose screening results (glucose challenge test or O'Sullivan screen). There will be a 15% to 20% incidence of abnormal 3-hour glucose tolerance test results in the group with abnormal screening tests, giving us a 2% to 5% incidence of gestational diabetes in the tested patients. This contrasts dramatically and significantly with the 62.4% abnormal 1-hour results and 9.4% abnormal 3-hour glucose tolerance test results in the study of Dr. Angel and his coworkers, whose pa-
766
Angel et al.
ttents were tested after 48 hours of oral terbutaline therapy according to a standard dosage schedule. The data presented by Angel et al. are clear, well stated, and essentially confirm the findings of a similar study presented to the Society of Perinatal Obstetricians in February 1987. 1 Brown and Tejani! reporting on 23 patients receiving oral terbutaline therapy for ~3 weeks after initial successful tocolysis with ethanol, described no significant adverse maternal or fetal effects, but blood sugar status was not mentioned. Obstetric care providers are highly sensitized now to the prevention of preterm birth, both through prenatal risk assessment and aggressive management of preterm labor. Oral tocolytic agents, particularly terbutaline, are being used quite liberally in offices and clinics everywhere. Do the findings of Angel et al. pose a serious challenge to our abiding philosophic dictum of "first do no harm?" My personal bias as a private practitioner is that we can manage maternal hyperglycemia far easier than we can the complications of preterm labor and delivery. No discussion would be complete without a few questions for the presenter. Do you think that glucose screening after 48 hours of oral terbutaline, as compared with Main et al., 1 who screened after 7 or more days, would account for the difference in gestational diabetes in the two study groups (9.4% versus 33%)? Do you have any comments about the 16 patients in the study group for whom 3-hour oral glucose tolerance test results are not known? Were there any patients in the study group who were screened for gestational diabetes before receiving oral terbutaline, thereby becoming their own controls? Was terbutaline stopped when gestational diabetes was diagnosed in the study group? Also, was there any difference between the patients with gestational diabetes in the two groups in terms of the need for insulin therapy? Was there any difference in adverse maternal or perinatal outcomes between the two groups? Finally, do you have any recommendations regarding chronic oral tocolytic therapy in patients with diabetes? REFERENCES I. Main EK, Main DM, Gabbe SG. Chronic oral terbutaline
tocolytic therapy is associated with maternal glucose intolerance. AM J 0BSTET GYNECOL 1987;157:644. 2. Brown SM, Tejani NA. Terbutaline sulfate in the prevention of recurrence of premature labor. Obstet Gynecol 1981;57:22.
September 1988 Am J Obstet Gynecol
DR. ANGEL (Closing). Our protocol required glucose tolerance testing after at least 48 hours of tocolytic dosing. Some of the patients were tested at 3 and 4 days. Due to the logistics of the testing protocol, in that patients went home and came back to the clinic, many were tested after a 7-day period. I do not believe that there is an augmentation of glucose intolerance that has a dose- or a time-related frequency, such that a week would result in a more common or a more frequent incidence of gestational diabetes. In fact, we did retest some women later, and the time lag did not make any difference. However, Main's group did look at the effects on glucose challenge test results 7 days after the start of terbutaline dosing. I think what should be done are serial tests, but I have found very few patients who will agree to serial 3-hour glucose tolerance testing. We did not test any patients in our study group before they went into preterm labor. Main's group did, as a method partially related to the retrospective nature of their study. These patients were in the high-risk clinic and they were tested per protocol; then, if they had preterm labor, they were tested again. We did not have that setup. However, in those patients with gestational diabetes while taking terbutaline who achieved 36 weeks' gestation and progressed to where we could discontinue the tocolytic drugs, we did so and retested them. The numbers were not sufficient to add up to something we could present. There were three patients, one of whom was insulin dependent, and she no longer required insulin therapy. The two other patients who controlled their disease by diet no longer required dietary therapy. These were the controls that we had. The diabetes was not harder to control in the terbutaline group versus the control group. In the patients who had gestational diabetes, the numbers were small but the perinatal outcomes, primarily birth weight and cesarean section rates, were the same. As to patients with overt diabetes and the agents for tocolysis, I do not have the answer. We proceed with ritodrine or terbutaline and watch the blood sugar levels closely, readjusting insulin therapy as necessary. Dr. Morrison has some data supporting magnesium gluconate as an oral tocolytic, and I think we need some more information on that, as is the case with the magnesium oxide data from Dr. Petrie. My closing statement would be that I am not sure where to go as far as testing terbutaline versus ritodrine. If we use terbutaline, we test all of our patients for gestational diabetes; if a positive glucose tolerance test results, they are treated as if the diabetes were endogenous.