Intrahepatic cholestasis of pregnancy: what's new.

Abstract

Itching has been reported to occur in 17% of pregnancies. Pruritus without primary skin lesions often results from intrahepatic cholestasis of pregnancy (ICP). The condition manifests itself in severe cases as jaundice (intrahepatic jaundice of pregnancy), but in mild cases itching may be the only symptom (pruritus gravidarum). The terms ‘obstetric cholestasis’, ‘recurrent jaundice of pregnancy’, ‘pruritus gravidarum’, ‘icterus gravidarum’ and ‘idiopathic jaundice of pregnancy’ all refer to the same clinical entity (ICP), characterized by pruritus with or without jaundice due to cholestasis. There is considerable geographical variation in the incidence of ICP.1 The condition is very common in Chile (14%), especially among Araucanian Indian women (24%), and in Bolivia. ICP used to be quite common in Scandinavia. The incidence of ICP is overall lower in Europe (0.1–1.5%), the United States, Canada and Australia.2 This geographical variation is now attributed to environmental and/or dietary factors. ICP recurs in 60–70% of subsequent pregnancies. There is a family history in 50% of ICP cases. ICP has been associated with twin and triplet pregnancies.2 ICP manifests itself in 80% of cases after the 30th week of pregnancy with pruritus affecting the palms and soles and extending to the legs and abdomen. The condition may be preceded by a urinary tract infection. Mild nausea and discomfort in the upper right quadrant may accompany the pruritus. Mild jaundice can be seen in 20% of cases and may be associated with subclinical steatorrhoea and increased risk of haemorrhage. The symptoms and biochemical abnormalities of the condition resolve within 2–4 weeks postpartum, and cases of prolonged postpartum cholestasis are exceptional. An increased risk of cholelithiasis has been reported but remains debatable. Elevation in serum bile acids is the most sensitive indicator of ICP.1 The serum levels of bile acids correlate with the severity of pruritus. Mild abnormalities of the liver function tests are commonly found, namely, elevated cholesterol, triglycerides, phospholipids, transaminases, alkaline phosphatase, 5′-nucleotidase and lipoprotein X. The bilirubin is mildly to moderately elevated (2–5 mg/dL) in jaundiced patients. Malabsorption of fat may cause vitamin K deficiency resulting in a prolonged prothrombin time. Skin biopsy is unnecessary as the condition does not present with primary skin lesions, while liver biopsy shows cholestatic changes namely, dilated bile canaliculi, bile pigment in the parenchyma and minimal inflammatory changes. The differential diagnosis of ICP without icterus (pruritus gravidarum) includes specific dermatoses of pregnancy, allergic reactions and pruritus associated with striae gravidarum. Specific dermatoses of pregnancy, such as prurigo of pregnancy, can be ruled out by the absence of primary skin lesions. However, the distinction between ICP and prurigo of pregnancy can be a challenge. Some authors suggest that prurigo of pregnancy (previously known as ‘Besnier's prurigo gestationis’) and intrahepatic cholestasis of pregnancy may be different stages of severity of the same entity. ICP with icterus (intrahepatic jaundice of pregnancy) should be distinguished from acute liver of pregnancy, pre-eclampsia complicated by increased liver enzymes, hyperemesis gravidarum, viral hepatitis, hyperbilirubinaemic states, drug eruptions, obstructive biliary disease, and haemolytic and metabolic diseases among others. The specific work-up in each case should be directed by the constellation of clinical findings and preliminary biochemical abnormalities. The aetiology of ICP is multifactorial with hormonal, genetic, environmental and probably alimentary factors playing a part.2 Oestrogens interfere with bile acid secretion across the basolateral and canalicular membrane of the hepatocyte. Oestriol-16α-d-glucuronide, the oestrogen metabolite that increases most in pregnancy, was found to be cholestatic in animal studies. This compound decreases the fluidity of the sinusoidal membrane and subsequently lowers the Na+/K+-ATPase activity. As a result, there is a reduction of the sodium gradient, which is necessary for the sodium-dependent bile acid uptake into the hepatocyte. Experiments in rats showed that inhibition of basolateral transport proteins by oestradiol-17-β-d-glucuronide occurs at both transcriptional and post-transcriptional levels. Some authors postulate that oestrogens regulate the action of actin microfilaments, which mediate bile secretion by hepatocytes. Progesterone metabolites play an important part in the pathogenesis of ICP. Progesterone treatment during the third trimester has been associated with ICP. Progestins inhibit hepatic glucuronyltransferase, thereby reducing the clearance of oestrogens and amplifying their effects. The profile of progesterone metabolites in serum from patients with ICP differs significantly from that seen in normal pregnancy. Monosulphated or disulphated progesterone metabolites, in particular the 3α- and 5α-isomers, are substantially increased in patients with ICP. Additionally, an increased ratio of 3α-hydroxylated steroids to 3β-hydroxylated steroids is found, which is a selective parameter for the diagnosis of ICP. This profile of progesterone metabolites is attributed exclusively to ICP but whether it is a primary or secondary event needs to be resolved. Both biliary and faecal excretion of sulphated and glucuronated progesterone metabolites are decreased in ICP. It has been suggested, that the increased amount of sulphated progesterone metabolites in serum can saturate the maximal transport capacity of membrane transport proteins of the hepatocyte. Familial clustering and geographical variation of cholestasis of pregnancy strongly indicate a genetic predisposition for ICP. A high prevalence of the HLA haplotype Aw31B8 was found in one kindred. Several studies show that, in mothers of patients with progressive familial intrahepatic cholestasis (PFIC) or benign recurrent intrahepatic cholestasis, a higher incidence of ICP is observed. Mutations in genes encoding biliary transport proteins have been identified in patients with PFIC. Patients with PFIC type 3 show mutations of the multidrug resistance 3 (MDR3) gene, which encodes the canalicular phosphatidylcholine translocase. Notably, in a family of PFIC type 3 patients, six women with a history of ICP were heterozygous for the same deletion (1712delT) in the MDR3 gene.3 To make things even more complex, progesterone binds to and modulates the activity of MDR translocases. This finding indicates the interaction between hormonal and genetic factors in ICP. Exogenous factors have been implicated in the pathogenesis of the condition. The decline of ICP in Chile during the past two decades, the higher incidence of the condition during the winter, and the low selenium levels in ICP patients in some studies may indicate a role for environmental and/or alimentary factors in ICP. Nevertheless, the importance of these factors in the pathogenesis of the condition has been debated. ICP has been associated with fetal risks, namely fetal distress, stillbirths and preterm delivery. These complications point out that ICP is a pathological and not a benign condition, as previously thought (‘pregnancy is cholestatic’). The pathogenesis of fetal complications has not been fully determined. Autopsies of the placenta show signs of acute anoxia. Stillborns often lie in meconium-stained amniotic fluid, and infusion of cholic acid in fetal sheep increases the incidence of meconium passage. Meconium can cause acute umbilical vein obstruction. Furthermore, in vitro studies indicate that the increased flux of bile acids from the mother to the fetus may cause vasoconstriction of human placental chorionic veins. It can be concluded from the above data, that the fetal complications in ICP may be caused by decreased fetal elimination of toxic bile acids. Intense pruritus and the risk of fetal adverse effects justify the search for an effective treatment of ICP. Obstetric management varies considerably all over the world, and most authors recommend weekly fetal cardiotocographic monitoring from the 34th week of gestation. Induction of labour has been advocated in the 38th week in mild cases and in the 36th week in severe cases. Some authors have recommended assessment of fetal lung maturity and delivery if a patient presents at or beyond 36 weeks and the cervix is favourable, whereas pharmacological treatment is recommended if the patient presents before 36 weeks. Mild cholestasis responds to symptomatic treatment with soothing baths, topical antipruritics, emollients and primrose oil among others. Antihistamines are rarely effective, while epomediol and silymarine have been helpful in mild cases of cholestasis.1 Treatment with phenobarbital, activated charcoal and S-adenosylmethionine have had limited success. Phenobarbital has shown minimal effect on the pruritus of ICP and variable effects on the biochemical abnormalities of the condition.4, 5 Intravenous use of S-adenosylmethionine has shown beneficial effects on both pruritus and biochemical abnormalities in mild ICP. Nevertheless, in a randomized, placebo-controlled study, this treatment was not better than placebo.6 Ultraviolet light B has been variably effective. Dexamethasone suppression of fetoplacental oestrogen production was effective in a small uncontrolled trial.7 Anion exchange resins, such as cholestipol and cholestyramine, bind bile acids and decrease their enterohepatic circulation. Cholestyramine needs to be administered for several days before a clear benefit for pruritus can be obtained, and still, does not improve the biochemical abnormalities of the condition.4, 5 Furthermore, it carries the disadvantage of precipitating vitamin K, and should be administered in conjunction with weekly vitamin K supplementation. A case of severe fetal intracranial haemorrhage during treatment with cholestyramine for ICP has been reported. Vitamin K and cholestyramine should be given at different times of the day so that cholestyramine does not affect vitamin K absorption. The lack of randomized, placebo-controlled cholestyramine trials in ICP renders difficult the evaluation of its clinical efficacy. Small uncontrolled studies undertaken by Laatikainen4 and Heikkinen et al.5 suggested some decrease in pruritus in approximately half of the patients studied. Nevertheless, recurrence of itching occurred in several cases after the first week of treatment. In the study by Laatikainen,4 patients with high serum bile acids did not respond to cholestyramine. The results of these studies as well as several case reports suggest that cholestyramine may be effective only in mild ICP. Notably, several patients with mild ICP who did not respond to cholestyramine subsequently responded to other treatment modalities, such as ursodeoxycholic acid (UCDA) and dexamethasone. The study of Rampone et al. is the first large cholestyramine trial in ICP. In this uncontrolled trial, 80 pregnant women were treated with cholestyramine. The response to cholestyramine (70%) was higher than that suggested by previous small studies. This may be due to the fact that the study included predominantly cases of mild ICP. The authors need to address this by grading the severity of pruritus in their patients. Additional information about the effects of cholestyramine on the biochemical abnormalities in these patients is necessary. Finally, the authors should clarify whether cholestyramine had a sustained effect on pruritus in their patients and whether the medication decreased the incidence of fetal compromise. UDCA, a naturally occurring hydrophilic bile acid, improves the clinical and liver function abnormalities in several cholestatic liver diseases. UDCA protects against injury to bile ducts by hydrophobic bile acids and stimulates the excretion of these and other hepatotoxic compounds as well as sulphated progesterone metabolites. The medication reduces bile acid levels in colostrum, cord blood and amniotic fluid. UDCA has been studied more extensively than cholestyramine in ICP. Four randomized, controlled UDCA trials concurred that UDCA, when administered in doses between 450 and 1200 mg daily, was highly effective in controlling the pruritus and liver dysfunction associated with ICP.8-11 Compared with cholestyramine, UDCA worked faster, had a more sustained effect on pruritus, and showed higher efficacy in improving the biochemical abnormalities of ICP. Nicastri et al. reported that the combination of UDCA with S-adenosylmethionine was more effective than either medication alone.8 Finally, in all studies UDCA was absolutely safe for both mother and fetus. Davies et al. suggested a potential benefit from UDCA in fetal mortality associated with ICP.12 Large randomized, placebo-controlled UDCA trials may be necessary before this can be approved for treatment of ICP by the Food and Drug Administration of the United States. Nevertheless, for the time being, UDCA can be considered as a first-line treatment in severe ICP. Other treatment modalities however, such as dexamethasone, which have shown promise in preliminary uncontrolled studies, also need to be further evaluated.