The use of corticosteroids for the treatment of hyperemesis gravidarum was advocated by Kemp more that 60 years ago based on the theory that "relative adrenal insufficiency" in early pregnancy caused Addison-like nausea and vomiting. In the years that followed, several reports supported their use, but there were no analytical studies. In 1979, Ylikorkala published the first randomized double-blind placebo-controlled trial testing this hypothesis. He gave intramuscular corticotropin to stimulate endogenous cortisol for four days in 32 women (16 in each group) with hyperemesis gravidarum.1 All of the patients stopped vomiting while in hospital but corticotropin and placebo fared equally well in stopping vomiting and providing relief to patients. Nelson-Piercy, in 1994, described four patients with intractable hyperemesis.2 All four responded promptly to treatment with prednisolone 40 to 45 mg per day.
We have reported on the findings of a study in which 18 women with hyperemesis gravidarum were treated with oral methylprednisolone.3 Hyperemesis gravidarum was defined as > 5% weight loss, persistent vomiting and a large degree of ketonuria. All of the women had had failure of previous antiemetic therapy (prochlorperazine, trimethobenzamide, promethazine). Excluded were women with twin pregnancies, foetal or placental anomalies, and suspected medical or surgical problems related to the vomiting.
All women were admitted to hospital and given intravenous thiamine 100 mg and intravenous hydration. The methylprednisolone dose was 48 mg per day for three days, followed by a tapering dose over two weeks. Nothing was taken by mouth, other than the medication, until day 2 of treatment. If a regular diet could not be tolerated without vomiting within three days, the medication was discontinued. If vomiting recurred after two weeks of treatment or during tapering, the medication was restarted or extended, but not longer than for one month in total.
Seventeen of the 18 women, including four receiving parenteral nutrition, ceased vomiting in three days with oral methylprednisolone. Seven did not have further symptoms during their pregnancies. Nine vomited during or after tapering, but seven of these responded to extension or reinstitution of therapy. Four of six patients on total parenteral nutrition at the start of therapy had a complete response within three days. (See Figure 1.) Some patients noted that they had a dramatic response to the corticosteroid that was different from the "containment" of vomiting by conventional antiemetics. This may be attributable to the steroid itself, since we used large doses.
Sixteen patients delivered at term (> 37 weeks' gestation) with no anomalies reported at delivery. Two of the 16 infants weighed less than 2500 g at birth. Of the remaining two patients, one decided to have an abortion and the other was lost to follow-up.
Following the observational study, we conducted a randomized controlled trial comparing oral methylprednisolone with oral promethazine.4 Forty patients were enrolled over 11 months (20 in each group). Women with hyperemesis gravidarum (defined as any patient requiring hospitalization for NVP), a normal-appearing pregnancy, and at less than or equal to 16 weeks' gestation were admitted to the study. There were no significant differences between the groups with respect to maternal age, gravidity, parity, weeks gestational age at entry, number of prior admissions, or greater than 5% body weight loss. There was one difference in the baseline factors: the duration of hyperemesis in the promethazine group was significantly longer than in the methylprednisolone group (28 days [range 5-75], 14 days [range 6-64], p < 0.03). Treatment was with either methylprednisolone 16 mg or promethazine 25 mg three times daily for three days. The methylprednisolone was then tapered over two weeks, and the promethazine dose remained the same for two weeks. Patients who continued to vomit after two days had the study medication discontinued.
Three patients in the methylprednisolone group and two patients in the promethazine group failed treatment. One promethazine patient was lost to follow-up. No women on methylprednisolone, but five of 17 on promethazine, were readmitted for hyperemesis within two weeks of discharge (p = 0.0001). There were no adverse effects of either drug noted. In this study we were primarily concerned about maternal welfare and used the methylprednisolone as a "rescue" therapy in patients after their 9th or 10th week of gestation.
Birth information was available for 12 patients in the methylprednisolone group and 11 in the promethazine group. The neonates in both groups did not differ with respect to birth weight or 1- and 5-minute Apgar scores. One patient in the methylprednisolone group delivered a male infant at 35 weeks' gestation with a genetic disorder (Smith-Lemli-Opitz syndrome) who died on the second day after birth.
Parenteral corticosteroids may be substituted for patients unable to tolerate the oral medication, but more than 90% of the reported patients have been able to tolerate the oral regimen. This makes steroid therapy an ideal outpatient regimen. After the first weeks of therapy we have limited the use of corticosteroids to low dosages (12 mg methylprednisolone per day or less) for no more than six weeks of total treatment. Others have used lower doses from the start with reported efficacy. Some have continued the corticosteroids for as long as 10 weeks. It is our view that usage of this duration begins to pose a significant risk of maternal complications and should be strictly limited. Since the vast majority of hyperemesis, even if refractory to antiemetic therapy, can be managed with supportive care, a reasonable risk-benefit ratio demands a high degree of confidence regarding the safety of the regimen.
Even chronic steroid use has not been shown to pose significant foetal risk. Although it is still debated whether corticosteroids could have any teratogenic effect in humans, therapy can usually be delayed until at least 10 weeks' gestation. Recently, an increased risk of preterm birth and preterm premature rupture of membranes has been attributed to corticosteroids; this effect has been described only with use throughout pregnancy, however.5
Serious maternal morbidity has not been reported with exposure comparable to that which our patients received. Reported complications from steroid therapy for hyperemesis are clearly related to prolonged treatment at very high dosages. This emphasizes the need to taper the medication and limit the duration of usage. In our experience, corticosteroids for the treatment of hyperemesis are safe and effective and can significantly reduce the need for hospitalization.
RELATED PUBLICATION BY THE SPEAKER:
Goodwin TM. Hyperemesis gravidarum. Clin Obstet Gynecol 1998;41:597-605.
Back to home