British Journal of Obstetrics and Gynaecology December 1992, Vol. 99, pp. 964968
OBSTETRICS
A randomised dacebo controlled trial of labetalol in the treatment of mild to moderate pregnancy induced h w e r t ension I
C. J. P I C K L E S F. B R O U G H T O N P I P K I N E.M. SYMONDS Department of Obstetrics & Gynaecology, University Hospital, Queen’s Medical Centre, Nottingham, NG7 2UH, UK
ABSTRACT Objective To determine the need for, and efficacy of, treatment with labetalol in women with mild-to-moderate pregnancy induced hypertension (PIH). Design Prospective double-blind randomised placebo controlled study. Setting Maternity units of five hospitals in the Trent Region. Subjects 144 women (86 primigravid) who developed PIH after 20 weeks gestation. Interventions Treatment with oral labetalol up to 600 mg daily or placebo with subsequent care of treatment failures in accordance with the attending obstetrician’s practice. Main outcome measures Number of days spent as an antenatal inpatient; the development of proteinuria; the perceived need for induction of labour or elective caesarean section; and gestation age at delivery. Results Labetalol significantly lowered the blood pressure and reduced the incidence of proteinuria. However, neither the number of days spent as an antenatal inpatient, nor the perceived need for induction of delivery o r elective caesarean section, nor the gestation age at delivery differed significantly between the two treatment groups. Post-randomisation consideration of early ( ~ 3 weeks) 2 and late (>32 weeks) onset groups showed the placebo treated early-onset group ( n = 15) to have more patients with severe hypertension (>150/110 mmHg) and a greater requirement for additional antihypertensive therapy prior to labour than the group treated with labetalol ( n = 16). Conclusion Anti-hypertensive intervention therapy in pregnancy induced hypertension has been examined using a placebo controlled randomised double-blind trial of labetalol in pregnancy. The maximum blood pressure prior to labour and the incidence of proteinuria was reduced in women on active therapy. However, the length of gestation was not significantly prolonged and indices of clinical outcome were not significantly altered. The appropriateness of pharmacological therapy for late-onset P I H may be questioned.
There is increasing evidence that the cardiac output is not compromised in pregnancy induced hypertension (PIH) or preeclampsia (PE) until fairly late in the clinically apparent disease process (Wallenburg 1990; Macphail et al. 1991). It is thus supposed that hypertension is driven by an increase in total systemic resistance. While there is some haemoconcentration, the majority of the increase in systemic resistance is assumed to lie in the peripheral arterioles. Blockade of the alpha-adrenergic nervous system lowers peripheral resistance; blockade of beta-adrenergic receptors alters baroreflex sensitivity, blocks peripheral adrenoceptors and can reduce cardiac output. Labetalol is a combined alpha- and beta-blocker, the beta-blocking component being slightly selective for the P,-adrenoceptors of the peripheral Correspondence: E Broughton Pipkin.
964
and pulmonary vasculature (Brittain & Levy 1976). It has been used for more than a decade in the treatment of hypertension in pregnancy (Riley & Symonds 1982). Over that decade views on the treatment of mild to moderate P I H have shifted. While no one would dispute the need for careful monitoring of such patients, a few of whom will progress to the severe disease, the need for pharmacological treatment of mild and moderate PIH is being questioned. We have therefore undertaken a randomised, double-blind placebo controlled study of the use of labetalol in such women. The primary outcome variable assessed was the birthweight of the infant but other, maternal end points were also examined: the number of days spent as an antenatal inpatient; the development of clinically significant proteinuria; the perceived need for induction of labour or elective caesarean section and the
C O N T R O L L E D T R I A L OF LABETALOL I N P R E G N A N C Y
duration of gestation. Factors specifically relating to fetal and neonatal outcome from this study have been previously reported (Pickles et al. 1989).
Subjects and methods The risk of having a baby of birthweight less than two SD below the mean (i.e. below the 5th centile) in pregnancies complicated by PIH has been calculated as 2.7 times (Ounsted et al. 1985), although her series will have included some women with chronic hypertension since she included those whose blood pressure was 2140/90 mmHg at any time in pregnancy. In order to detect a halving of the risk, a total of 140 patients would be needed ( a = 0.05, p = 0.5) (Altman 1982). In fact, 152 women with singleton pregnancies were recruited to the trial from five hospitals (Queen’s Medical Centre, Nottingham; The City Hospital, Nottingham; King’s Mill Hospital, Mansfield; The City Hospital, Derby; and the North Staffordshire Hospital Centre, Stoke-on-Trent) during the 18 month period of the study. Insofar as was logistically possible, all women fulfilling the entry criteria and booked with one of the participating obstetricians were admitted to the study. Entry criteria were a systolic blood pressure of 140160 mmHg and a diastolic pressure of 90-105 mmHg after 15 min rest on two occasions 24 h apart, occurring between 20 and 38 weeks gestation and without detectable proteinuria. No woman having a past history of hypertension, renal, metabolic, cardiovascular, respiratory or collagen disease was included. Women were identified in the antenatal clinic and initially admitted for at least 24 h rest and assessment but five patients, from an outlying hospital, were not so admitted. Allocation to treatment was double-blind and randomized. Labetalol (Trandate@) and identical placebo tablets were supplied by the manufacturers (Duncan Flockhart, Uxbridge, Middlesex) directly to the pharmacies of participating hospitals. Once recruited, allocation to treatment was made by the pharmacies on a random number basis. The codes were not broken until after delivery. The initial dose was one tablet (100 mg) three times daily. This was increased to a maximum of two tablets three times daily if, in the opinion of the woman’s obstetrician, blood pressure control was unsatisfactory. If control was still inadequate, antihypertensive therapy with drugs other than labetalo1 was initiated, depending on the consultant’s current practice. Recordings of blood pressure were taken at weekly intervals from recruitment until delivery. Patients rested in a semirecumbent position for at least five minutes before the recording was made. Conventional sphygmomanometers were used, and phase IV Korotkoff sounds were used for the diastolic pressure. Blood pressure data shown in the text are expressed in terms of weeks since entry to the trial. Proteinuria was assessed (dipstick) every week. Serial serum oestriol (radioimmunoassay) and urate (Technicon SMA 560 Autonalayser, Basingstoke, Hants) measurements were made from women attending QMC and from the majority of patients attending the other centres. Antenatal care after randomization to treatment was given by individual participating consultants according to their current practice.
965
The study was approved by the ethnics committees of the five hospitals. All patients gave informed verbal consent for inclusion in the study. Routine ultrasound pregnancy dating was not available in the outlying hospitals. Gestation length was thus assessed for all patients using conventional obstetric criteria. The number of days spent antenatally as an inpatient was recorded for each woman, as was the use of additional antihypertensive agents. The perceived need for labour induction and the method of delivery were recorded. The highest blood pressure recorded during labour was noted, as was the blood pressure during the first postnatal day. The outcome with respect to the babies has been previously reported in detail (Pickles et al. 1989). Data measured on an interval scale and normally-distributed are expressed throughout as the mean value with one standard deviation (SD) in parentheses; 95% confidence intervals (CI) are given for the difference between group means. Other data are expressed as medians with interquartiles in parentheses. Specified group data were compared using the Mann-Whitney test.
Results Eight of the women recruited initially were subsequently excluded. Four were found not to have fulfilled the entry criteria (one with essential hypertension; two with severe hypertension in the first 24 h; one with proteinuria on admission). Two women withdrew themselves; one developed a rash and was withdrawn and one was mistakenly treated with ward stock labetalol. Thus 144 patients were included in the analysis, of whom 70 were treated with labetalol and 74 with placebo. Table 1 summarizes some basal data for the study population by treatment group. There were no significant differences between the study groups for any of the basal variables. The administration of labetalol resulted in statistically highly-significant falls in systolic and diastolic pressures at each treatment week by comparison with placebo (ANOVA P
OBSTETRICS
A randomised dacebo controlled trial of labetalol in the treatment of mild to moderate pregnancy induced h w e r t ension I
C. J. P I C K L E S F. B R O U G H T O N P I P K I N E.M. SYMONDS Department of Obstetrics & Gynaecology, University Hospital, Queen’s Medical Centre, Nottingham, NG7 2UH, UK
ABSTRACT Objective To determine the need for, and efficacy of, treatment with labetalol in women with mild-to-moderate pregnancy induced hypertension (PIH). Design Prospective double-blind randomised placebo controlled study. Setting Maternity units of five hospitals in the Trent Region. Subjects 144 women (86 primigravid) who developed PIH after 20 weeks gestation. Interventions Treatment with oral labetalol up to 600 mg daily or placebo with subsequent care of treatment failures in accordance with the attending obstetrician’s practice. Main outcome measures Number of days spent as an antenatal inpatient; the development of proteinuria; the perceived need for induction of labour or elective caesarean section; and gestation age at delivery. Results Labetalol significantly lowered the blood pressure and reduced the incidence of proteinuria. However, neither the number of days spent as an antenatal inpatient, nor the perceived need for induction of delivery o r elective caesarean section, nor the gestation age at delivery differed significantly between the two treatment groups. Post-randomisation consideration of early ( ~ 3 weeks) 2 and late (>32 weeks) onset groups showed the placebo treated early-onset group ( n = 15) to have more patients with severe hypertension (>150/110 mmHg) and a greater requirement for additional antihypertensive therapy prior to labour than the group treated with labetalol ( n = 16). Conclusion Anti-hypertensive intervention therapy in pregnancy induced hypertension has been examined using a placebo controlled randomised double-blind trial of labetalol in pregnancy. The maximum blood pressure prior to labour and the incidence of proteinuria was reduced in women on active therapy. However, the length of gestation was not significantly prolonged and indices of clinical outcome were not significantly altered. The appropriateness of pharmacological therapy for late-onset P I H may be questioned.
There is increasing evidence that the cardiac output is not compromised in pregnancy induced hypertension (PIH) or preeclampsia (PE) until fairly late in the clinically apparent disease process (Wallenburg 1990; Macphail et al. 1991). It is thus supposed that hypertension is driven by an increase in total systemic resistance. While there is some haemoconcentration, the majority of the increase in systemic resistance is assumed to lie in the peripheral arterioles. Blockade of the alpha-adrenergic nervous system lowers peripheral resistance; blockade of beta-adrenergic receptors alters baroreflex sensitivity, blocks peripheral adrenoceptors and can reduce cardiac output. Labetalol is a combined alpha- and beta-blocker, the beta-blocking component being slightly selective for the P,-adrenoceptors of the peripheral Correspondence: E Broughton Pipkin.
964
and pulmonary vasculature (Brittain & Levy 1976). It has been used for more than a decade in the treatment of hypertension in pregnancy (Riley & Symonds 1982). Over that decade views on the treatment of mild to moderate P I H have shifted. While no one would dispute the need for careful monitoring of such patients, a few of whom will progress to the severe disease, the need for pharmacological treatment of mild and moderate PIH is being questioned. We have therefore undertaken a randomised, double-blind placebo controlled study of the use of labetalol in such women. The primary outcome variable assessed was the birthweight of the infant but other, maternal end points were also examined: the number of days spent as an antenatal inpatient; the development of clinically significant proteinuria; the perceived need for induction of labour or elective caesarean section and the
C O N T R O L L E D T R I A L OF LABETALOL I N P R E G N A N C Y
duration of gestation. Factors specifically relating to fetal and neonatal outcome from this study have been previously reported (Pickles et al. 1989).
Subjects and methods The risk of having a baby of birthweight less than two SD below the mean (i.e. below the 5th centile) in pregnancies complicated by PIH has been calculated as 2.7 times (Ounsted et al. 1985), although her series will have included some women with chronic hypertension since she included those whose blood pressure was 2140/90 mmHg at any time in pregnancy. In order to detect a halving of the risk, a total of 140 patients would be needed ( a = 0.05, p = 0.5) (Altman 1982). In fact, 152 women with singleton pregnancies were recruited to the trial from five hospitals (Queen’s Medical Centre, Nottingham; The City Hospital, Nottingham; King’s Mill Hospital, Mansfield; The City Hospital, Derby; and the North Staffordshire Hospital Centre, Stoke-on-Trent) during the 18 month period of the study. Insofar as was logistically possible, all women fulfilling the entry criteria and booked with one of the participating obstetricians were admitted to the study. Entry criteria were a systolic blood pressure of 140160 mmHg and a diastolic pressure of 90-105 mmHg after 15 min rest on two occasions 24 h apart, occurring between 20 and 38 weeks gestation and without detectable proteinuria. No woman having a past history of hypertension, renal, metabolic, cardiovascular, respiratory or collagen disease was included. Women were identified in the antenatal clinic and initially admitted for at least 24 h rest and assessment but five patients, from an outlying hospital, were not so admitted. Allocation to treatment was double-blind and randomized. Labetalol (Trandate@) and identical placebo tablets were supplied by the manufacturers (Duncan Flockhart, Uxbridge, Middlesex) directly to the pharmacies of participating hospitals. Once recruited, allocation to treatment was made by the pharmacies on a random number basis. The codes were not broken until after delivery. The initial dose was one tablet (100 mg) three times daily. This was increased to a maximum of two tablets three times daily if, in the opinion of the woman’s obstetrician, blood pressure control was unsatisfactory. If control was still inadequate, antihypertensive therapy with drugs other than labetalo1 was initiated, depending on the consultant’s current practice. Recordings of blood pressure were taken at weekly intervals from recruitment until delivery. Patients rested in a semirecumbent position for at least five minutes before the recording was made. Conventional sphygmomanometers were used, and phase IV Korotkoff sounds were used for the diastolic pressure. Blood pressure data shown in the text are expressed in terms of weeks since entry to the trial. Proteinuria was assessed (dipstick) every week. Serial serum oestriol (radioimmunoassay) and urate (Technicon SMA 560 Autonalayser, Basingstoke, Hants) measurements were made from women attending QMC and from the majority of patients attending the other centres. Antenatal care after randomization to treatment was given by individual participating consultants according to their current practice.
965
The study was approved by the ethnics committees of the five hospitals. All patients gave informed verbal consent for inclusion in the study. Routine ultrasound pregnancy dating was not available in the outlying hospitals. Gestation length was thus assessed for all patients using conventional obstetric criteria. The number of days spent antenatally as an inpatient was recorded for each woman, as was the use of additional antihypertensive agents. The perceived need for labour induction and the method of delivery were recorded. The highest blood pressure recorded during labour was noted, as was the blood pressure during the first postnatal day. The outcome with respect to the babies has been previously reported in detail (Pickles et al. 1989). Data measured on an interval scale and normally-distributed are expressed throughout as the mean value with one standard deviation (SD) in parentheses; 95% confidence intervals (CI) are given for the difference between group means. Other data are expressed as medians with interquartiles in parentheses. Specified group data were compared using the Mann-Whitney test.
Results Eight of the women recruited initially were subsequently excluded. Four were found not to have fulfilled the entry criteria (one with essential hypertension; two with severe hypertension in the first 24 h; one with proteinuria on admission). Two women withdrew themselves; one developed a rash and was withdrawn and one was mistakenly treated with ward stock labetalol. Thus 144 patients were included in the analysis, of whom 70 were treated with labetalol and 74 with placebo. Table 1 summarizes some basal data for the study population by treatment group. There were no significant differences between the study groups for any of the basal variables. The administration of labetalol resulted in statistically highly-significant falls in systolic and diastolic pressures at each treatment week by comparison with placebo (ANOVA P
