The other day we began a discussion of intrahepatic cholestasis of pregnancy, or obstetric cholestasis (OC). The post was prompted by the intrauterine death of an otherwise normal baby in a mother who had OC diagnosed late in pregnancy. Although women with OC can be absolutely miserable from the intense pruritus associated with this condition, they are at relatively low risk for serious complications themselves during the pregnancy. One maternal concern that has been raised is the possibility of increased risk for postpartum hemorrhage in women who have reduced blood-clotting factors secondary either to hepatic impairment or reduced intestinal absorption of vitamin K resulting from steatorrhea associated with OC, however, this rarely occurs.
Most women improve symptomatically within a month following delivery, but there is a recurrence risk as high as 90% with a subsequent pregnancy; and, about one-fourth of these women will develop intermittent pruritus, sometimes coinciding with stage of their menstrual cycle, or associated with oral contraceptive or other estrogen use. An excellent recent retrospective, cohort study out of Finland, looking at 10,504 women who had OC during the years 1972-2000, showed that these women also have significantly greater risk for developing complications related to hepatitis C virus infection, nonalcoholic liver cirrhosis, gallstones and cholecystitis, and nonalcoholic pancreatitis than case-matched controls (Ropponen A, et al., Hepatology 2006;43:647-9).
As part of the evaluation of OC, women should be screened for hepatitis C virus (HCV), since there is a higher prevalence of OC in HCV-positive individuals, and the virus infection itself is associated with short-term pregnancy, and long-term postpartum, complications. Other conditions that should be ruled out in OC are infections with hepatitis A and B, cytomegalovirus, and Epstein-Barr virus, and autoimmune hepatitis. With regard to the latter, antimitochondrial antibodies can be obtained to rule out primary biliary cirrhosis and anti-smooth muscle antibodies to exclude autoimmune chronic active hepatitis. Appropriate imaging studies should also be done to rule out obstructive cholelithiasis.
Although the pregnant woman herself is at fairly low risk for complications, it has been well-established that the baby is not. OC is associated with higher rates of spontaneous premature delivery, meconium passage in utero, fetal distress as assessed by heart rate monitoring, respiratory distress syndrome following delivery, intrauterine fetal demise (IUFD) and neonatal death (ND). Prior to the currently recommended therapeutic intervention of planned, early delivery, OC was accompanied by premature labor and delivery in 30-50% of affected pregnancies and IUFD, or ND, occurred in approximately 10%. In the 1995 report by the CESDI (Confidential Enquiry into Stillbirths and Deaths and Infancy) consortium in the United Kingdom, 5% of all term stillbirths were associated with OC. Since then, recommendations that women identified with OC are delivered, electively, no later 37-38 weeks, rates of IUFD associated with OC have been reduced to 0.5% or less. Of course, late preterm delivery can be associated with neonatal morbidity secondary to respiratory complications, but most pregnant women prefer this over the risk of losing their babies.
Management options prior to elective delivery are somewhat limited and inconsistent in their efficacy. Some providers routinely prescribe vitamin K (10 mg per day) orally once the diagnosis of OC is suspected, although the efficacy for this approach has not been proven in clinical trials. Absolute levels of serum bile acids do not reliably correlate with fetal outcome. Fetal heart rate (FHR) testing may provide some reassurance to providers and patients, but it is also not reliable in predicting fetal outcome. In our case that prompted these posts on OC, the baby had a beautifully reactive nonstress test and no FHR decelerations with maternal contractions to suggest any placental insufficiency within 12 hours of presentation with demise. Serial amniocentesis and transcervical amnioscopy (looking up through the cervix at the fetal membranes using a scope) have been used in OC patients to detect meconium and, if this is present, usually provides an indication for delivery, regardless of gestational age. Unfortunately, if meconium is not detected, this does not assure that the baby is not at risk.
Medications typically prescribed to relieve itching, such as diphenhydramine, hydroxyzine, and topical corticosteroids, seem to have very little effect on the pruritus associated with OC. Ursodeoxycholic acid (UDCA), taken in doses ranging from 500 mg to 2000 mg per day, seems to have the most consistent effect in relieving OC related symptoms. UDCA can take 2 to 3 weeks before an effect is noticed, but in some studies it has been shown to significantly reduce pruritus, serum bile acids, transaminases, and bilirubin. It has also been shown to reduce levels of bile acids in umbilical cord blood and amniotic fluid. Although the mechanism of action of UDCA is not entirely clear, it may protect hepatic and, possibly, placental cells from the toxic effects of bile acids and stimulate impaired hepatocellular secretion. Treatment with UDCA appears to reduce the risk for fetal complications related to OC and, as yet, has not been found to be associated with any short- or long-term fetal morbidity. High-dose dexamethasone therapy, may relieve some symptoms, but has shown inconsistent results in clinical trials and at this point, especially, in view of the recent concerns related to steroid administration and fetal growth and development, it cannot be recommended on a routine basis.
The primary question that remains unanswered in OC is why is the baby at such great risk? Do the elevated maternal serum bile acids cause damage to the placenta? To the fetal hepatobiliary system? To other organs, such as the heart? Do they stimulate the passage of fetal meconium directly? Or as the result of acute placental injury and ‘fetal distress’ secondary to reduced capacity of the placenta to transport oxygen or remove ‘wastes’? Or, is the mechanism of fetal compromise completely unrelated to the elevated bile acids? Once meconium has been passed, do some of the complications arise as the result of meconium aspiration before labor? If there is a genetic basis for OC, as there appears to be, does this put the baby at risk by itself for mechanisms that are unclear at this time?
Until we can elucidate a mechanism, and develop specific therapy to address that issue, or develop a specific test to identify the fetus at risk during an OC pregnancy, our best bet is UDCA and early delivery, knowing the occasional patient will still have a loss while under observation, or will lose a baby at term because OC was not recognized during the pregnancy….