As the treatment of cirrhosis improves, pregnancy in patients with cirrhosis is likely to become more common. Although maternal and fetal mortality is expected to similarly improve, pregnant patients with cirrhosis face unique risks. These include higher rates of spontaneous abortion and prematurity and a potential for life-threatening variceal hemorrhage, hepatic decompensation, splenic artery aneurysm rupture, and postpartum hemorrhage. Pregnancy outcome may be influenced by the underlying etiology of liver disease, as in viral and autoimmune hepatitis. Medications also impact the course of pregnancy, and must be tailored appropriately during this time. Liver Transpl 14:1081–1091, 2008. © 2008 AASLD.
INTRODUCTION AND EPIDEMIOLOGY
Pregnancy is a rare event in patients with cirrhosis. The exact incidence is unknown, but it has traditionally been low for 2 reasons. First, advanced liver disease does not typically occur until well after most patients have completed their reproductive years, with only 45 cases of cirrhosis occurring in every 100,000 women of reproductive age. Second, cirrhosis results in metabolic and hormonal derangements that lead to anovulation and amenorrhea.1
Increasing numbers of case reports and case series of pregnancy in this population, however, indicate that improvements in the treatment of chronic liver disease have resulted in higher conception rates and more successful pregnancy outcomes.1 Although previously reported to be as high as 10.5%,2 maternal mortality has likely improved along with better management of variceal hemorrhage and liver failure.
Nevertheless, morbidity and mortality likely remains higher than that of the general pregnant population. It is therefore important for physicians caring for pregnant patients with cirrhosis to understand how to approach potential complications and tailor medication regimens appropriately.
EFFECTS OF CIRRHOSIS AND PORTAL HYPERTENSION
Although it is difficult to distinguish the effects of pregnancy from the natural history of cirrhosis, maternal complications have been described in nearly half of pregnancies affected by cirrhosis and portal hypertension, largely as a result of variceal hemorrhage and liver failure.3
Esophageal variceal bleeding has been reported in 18% to 32% of pregnant women with cirrhosis and in up to 50% of those with known portal hypertension.4, 5 Among those with preexisting varices, up to 78% will have gastrointestinal bleeding during pregnancy, with a mortality rate of 18% to 50%.1 In contrast, pregnant patients with noncirrhotic portal hypertension fare much better. Their mortality rate from variceal bleeding is between 2% to 6%.1 This disparity may be related to the severity of their underlying liver disease, with patients with cirrhosis more likely to be coagulopathic.
Variceal bleeding most commonly occurs during the second and third trimesters when maternal blood volume is maximally expanded and the larger fetus causes increased compression of the inferior vena cava and collateral vasculature.6
As in nonpregnant patients with cirrhosis, endoscopic band ligation remains the mainstay of therapy for acute episodes of hemorrhage. The first case of successful band ligation in a pregnant patient with acute bleeding was reported in 1998 by Starkel et al.,7 but no prospective randomized trials for this treatment currently exist. As in the nongravid population, sclerotherapy was previously looked to as a potential alternative,8 but it has largely been replaced by band ligation. Experts argue that band ligation should be preferred during pregnancy because it avoids any potential risk from chemical instillation.7
Upper endoscopy in general appears to be safe during pregnancy, with the main risk being fetal hypoxia from sedative drugs or positioning. No cases of premature labor or fetal malformations have been reported in patients who have undergone endoscopy during pregnancy.9
Octreotide, designated as pregnancy category B by the Food and Drug Administration, is often used to treat acute variceal bleeding, although its safety has not been well studied in pregnant patients.10 Given its similarity to vasopressin, however, possible concerns include arteriolar vasospasm, which can result in decreased placental perfusion and an increased risk of placental abruption, myocardial infarction, peripheral ischemia, and hypertension.1, 3
Although transjugular intrahepatic portosystemic shunt (TIPS) placement is generally contraindicated during pregnancy because of the risk of radiation exposure to the fetus, it may be an appropriate rescue therapy for failed attempts to control variceal bleeding with band ligation or sclerotherapy.11 The risk of fetal malformations from radiation is thought to be increased at doses above 150 mGy and is considered negligible if doses are below 50 mGy.12 Only 3 cases of TIPS placement in pregnant patients with cirrhosis have been reported thus far, with radiation dosages of 0.1, 5.2, and 5.49 mGy used. In all 3 cases, the mothers survived the episode of acute variceal bleeding.11, 13, 14 In 1 case, the fetus ultimately died as a result of premature delivery and infant respiratory distress syndrome, but this was not thought to be related to the TIPS procedure.11
The prophylactic treatment of varices in pregnant women remains controversial. A case report by Zeeman and Moise15 in 1999 reported a successful outcome for prophylactic banding in a pregnant patient with cryptogenic cirrhosis and portal hypertension, but no prospective, randomized trials have been performed. According to the American Association for the Study of Liver Diseases practice guidelines, all patients diagnosed with cirrhosis should undergo screening endoscopy for the diagnosis of gastric and esophageal varices.16 Although there are no formal guidelines for pregnant women, most experts recommend that a screening endoscopy be done in women with cirrhosis either before pregnancy or in the early second trimester.
In women at high risk for bleeding, the potential benefits of primary prophylaxis with nonselective beta blockers such as propranolol and nadolol, designated by the Food and Drug Administration as pregnancy category C, may outweigh the risks of potential fetal harm. The most commonly reported morbidities from these drugs include fetal growth retardation, neonatal hypoglycemia, and neonatal bradycardia.16
Surgical creation of splenorenal or portacaval shunts has also been advocated on the basis of observational studies showing that women with cirrhosis with shunts have a lower risk of spontaneous abortion compared to women without shunts17 and a 7-fold lower risk of severe gastrointestinal hemorrhage.18 However, surgical shunt procedures are generally performed only in the setting of a life-threatening hemorrhage that is refractory to medical and endoscopic treatments.1
Another area of controversy in pregnant patients with portal hypertension is how to approach delivery. The risk of variceal bleeding is thought to be particularly increased at the time of labor because of the need for repetitive Valsalva maneuver. The resultant increased intra-abdominal pressure is postulated to lead to increased portal hypertension and thus a higher risk of variceal rupture. As a result, many experts advocate using elective caesarian section or forceps delivery under extradural analgesia in order to decrease this risk.3, 19 Although this seems reasonable, no randomized controlled trials have been performed to support this recommendation. If a prophylactic caesarian section is performed, a vascular surgeon should be available in the event that bleeding from pelvic or abdominal wall collaterals is encountered.6
Up to 24% of pregnant patients with cirrhosis will also experience hepatic decompensation, which can lead to rapid clinical deterioration.3 This has been described in all stages of pregnancy, but often occurs after episodes of variceal bleeding.20
When fulminant hepatic failure occurs, the only treatment available may be emergent liver transplantation. This has been performed during pregnancy in a small number of cases, with successful outcomes for both mother and fetus.21, 22 Reported complications remain high, however, and have included an increased risk of fetal ischemia, pregnancy-induced hypertension, anemia, caesarian section, and preterm delivery.23, 24 Moreover, these case reports largely involved women without underlying cirrhosis, and it is unclear how underlying liver disease would change the outcome.
Ascites and Spontaneous Bacterial Peritonitis
Ascites rarely occurs during pregnancy because of increased intra-abdominal pressure, which acts to resist the extravasation of fluid from splanchnic vessels and organs. If therapy is required, however, sodium restriction and diuretics can be used, as in nonpregnant patients with cirrhosis. Although cases of spontaneous bacterial peritonitis have not been reported, other causes of peritonitis have been known to increase the risk of premature delivery and placental abruption. If spontaneous bacterial peritonitis does occur, treatment should consist of a third-generation cephalosporin.1
Like nonpregnant patients with cirrhosis, pregnant patients with cirrhosis can develop encephalopathy if they are given predisposing medications or experience hypotension, hypoxia, infection, hypoglycemia, or gastrointestinal hemorrhage. Of note, spinal and general anesthesia should be avoided during delivery because of the potential for hypotension and the risk of precipitating encephalopathy. The mainstay of treatment remains lactulose and/or antibiotic therapy.1
Splenic Artery Aneurysm Rupture
Pregnant patients with cirrhosis also have an increased risk of splenic artery aneurysm rupture, which occurs in 2.6%. In fact, 20% of all splenic artery aneurysm ruptures occur during pregnancy, with 70% occurring during the third trimester. Rapid intra-abdominal bleeding and hypovolemic shock often ensue, resulting in substantial maternal and fetal mortality rates of 70% and 80%, respectively.3 The mechanism for the development of these aneurysms is unclear but may be related to increased splenic blood flow from both pregnancy and portal hypertension. High estrogen levels during pregnancy may also have effects on the elastic tissue of the tunica media. Management options consist of emergency splenectomy, transcatheter embolization of the aneurysm, or stent-graft placement. The latter 2 options are usually preferred in cases of portal hypertension, as an extensive collateral circulation in these patients makes surgery more difficult.25
Postpartum Uterine Hemorrhage
Postpartum uterine hemorrhage is another potential source of maternal morbidity and mortality, occurring in 7% to 10% of pregnancies in patients with cirrhosis. This is likely related to a higher incidence of coagulopathy and thrombocytopenia.9 Treatment is similar to that in patients without cirrhosis. Blood and coagulation factors should be transfused and administered oxytocin or other uterine contractile agents. Surgical therapy to ligate the bleeding vessels or hysterectomy is indicated when these measures fail.26
The spontaneous abortion rate in patients with cirrhosis is significantly higher than that of the general population, with a rate of 30% to 40%27 versus 15% to 20%.28 This is not true for patients with extrahepatic portal obstruction unrelated to cirrhosis, in whom the rate of spontaneous abortion is 3% to 6%.29 Moreover, patients with cirrhosis who have undergone portal decompressive procedures prior to conception have spontaneous abortion rates comparable to patients with extrahepatic obstruction.30 Termination of pregnancy most often occurs as a result of maternal death, variceal hemorrhage, stillbirth, intrauterine growth retardation, and maternal complications during delivery.6
In those pregnancies that do result in live births, the risk of prematurity is significantly increased, with a rate of up to 25%.30 In contrast, the general population, as of 2006, had a preterm birth rate of 12.8%.31 The perinatal death rate is likewise elevated and may be as high as 18%30 versus 1.08% in the noncirrhotic population.32 The majority of these data are from older literature, however, and may reflect complications of prematurity at a time when routine measures such as corticosteroids, surfactants, and modern neonatal intensive care management were unavailable.1
MEDICATIONS USED IN PATIENTS WITH LIVER DISEASE
Data regarding the use of medications commonly prescribed for patients with cirrhosis during pregnancy are scarce, with the majority of medications falling into pregnancy category C (Tables 1 and 2). Notable exceptions are neomycin, azathioprine, and penicillamine, which have been designated pregnancy category D because of animal studies showing a possible risk of fetal harm (Tables 1 and 2). Octreotide, cefotaxime, lactulose, telbivudine, prednisone, and ursodeoxycholic acid (UDCA) have been deemed pregnancy category B, as animal studies have shown no harmful effects (Tables 1 and 2). Ribavirin is pregnancy category X, with significant teratogenic effects in all animal studies at 0.01 times the maximum human dose (Tables 1 and 2).
Table 1. The Food and Drug Administration Pregnancy Categories
|A||Controlled studies indicate no risk.|
|B||Animal studies show no risk, and there are no human controlled studies, or animal studies may show an adverse effect that was not reproduced in human controlled studies.|
|C||Risk cannot be ruled out; animal studies show an adverse effect, and there are no human controlled studies, or there are no studies available.|
|D||Risk is evident in human studies; use may be considered if the benefit outweighs the risk.|
|X||Risk is evident in animal and human studies and outweighs any benefit; use is contraindicated in women who are or may be pregnant.|
Table 2. Medications Used for Liver Disease
|Furosemide||C||None||Electrolyte disturbances, decreased hepatic perfusion||Decreased uteroplacental perfusion and hypovolemia, electrolyte disturbances||Unknown; no reports available||76–78|
|Spironolactone||C||None||Electrolyte disturbances||Decreased uteroplacental perfusion and hypovolemia, feminization of males in rat studies||Probably; negligible amount in breast milk||76, 79|
|Propranolol||C||None||Bradycardia, bronchospasm, hypotension, fatigue||Intrauterine growth retardation, bradycardia, hypoglycemia, neonatal respiratory depression during labor||Probably; negligible amount in breast milk||80–87|
|Nadolol||C||None||Bradycardia, bronchospasm, hypotension, fatigue||Unknown||Possibly unsafe; concentrations in breast milk can be significant if high doses are used or in younger infants||88–90|
|Octreotide||B||Unknown||Gastrointestinal side effects||No teratogenic effects in animal models||Unknown||14|
|Lactulose||B||None||Diarrhea, electrolyte disturbances, abdominal cramps||No adverse effects in animal models||Unknown; no reports available||91|
|Rifaximin||C||None||Nausea, abdominal pain||Teratogenic in high doses in animals; no human studies||Unknown; no reports available||92|
|Neomycin||D||None||Nephrotoxicity, ototoxicity||Like other aminoglycosides, may cause fetal ototoxicity||Unknown; no reports, although other aminoglycosides are considered safe||93–96|
|Cefotaxime||B||None||Diarrhea, abdominal pain||No fetal harm in animal studies||Probably; negligible amount in breast milk||97–101|
|Ciprofloxacin||C||None||Nausea, dyspepsia||Teratogenic in animal studies; congenital anomalies reported in 1 human study, but none in 3 other studies||Probably; arthropathy in juvenile animals; 1 case report of pseudomembranous colitis||102–109|
|Norfloxacin||C||None||Nausea, dyspepsia||Teratogenic in animal studies; no harm in human observational study||Probably; no reports, although other quinolones are considered safe||102–103, 107|
|Lamivudine||C||Unknown||Lactic acidosis, pancreatitis, peripheral neuropathy, anemia||No increases in birth defects overall; reports of mitochondrial dysfunction||Unsafe; possible severe adverse reactions in infant||110–113|
|Interferon alfa 2a and 2b||C||Unknown||Fatigue, mood changes, pancytopenia, myalgias||Abortifacient in monkeys at 90 times human dose; case reports of transient thrombocytopenia, intrauterine growth restriction, and neonatal lupus||Unknown but not recommended; excretion into milk reported in mice||114, 115|
|Adefovir||C||Unknown||Nephrotoxicity, lactic acidosis||Unknown; no studies or case reports in either women or animals||Unknown but not recommended||116|
|Entecavir||C||Unknown||Lactic acidosis||No studies in humans; in rats, lower body weights, tail and vertebral malformations, and reduced ossification were seen at 3100 times human dose||Unsafe; secreted into milk of lactating rats||117|
|Telbivudine||B||Unknown||Lactic acidosis, neutropenia, thrombocytopenia||No adverse effects in animals at doses as high as 37 times human dose||Possibly unsafe; found to be excreted into the milk of lactating rats||118|
|Ribavirin||X||Unknown||Pancytopenia, nephrotoxicity, myalgias||Embryocidal and teratogenic in all animals studied; should also be avoided in men whose female partners are or may become pregnant||Possibly unsafe; unknown whether it is excreted into milk||119|
|Azathioprine||D||Unknown||Leukopenia, gastrointestinal side effects||22% malformations, 45% first trimester abortion||Unsafe||120|
|Prednisone||B||None||Hypertension, gestational diabetes, cushingoid appearance, osteonecrosis, weight gain, dyslipidemia, infection||4% malformation rate, adrenal insufficiency, premature rupture of membranes||Safe||121|
|Penicillamine||D||Unknown||Anemia, agranulocytosis, proteinuria||Possible cutaneous defects||Unknown||122, 123|
|Trientine||C||Unknown||Anemia, neurologic worsening, iatrogenic copper deficiency||Chromosomal defects||Unknown||124|
|Ursodeoxycholic acid||B||Unknown||Headache, dizziness, constipation||Effects during first trimester unclear; presumed safe during third trimester||Unknown||64, 66|
|Cholestyramine||C||None in rat models||Vitamin malabsorption, particularly K||No studies in pregnant humans, possible vitamin deficiencies||Not excreted into breast milk||125|
|Hydroxyzine||C||Unknown||Central nervous system depression||Neonatal withdrawal syndrome||Unknown||67|
EFFECT OF PREGNANCY ON SPECIFIC LIVER DISEASES
Chronic Viral Hepatitis B and C
The chief potential risk posed by viral hepatitis during pregnancy is transmission of the virus from mother to child. In areas endemic for hepatitis B, vertical transmission of the virus is responsible for the vast majority of infections. When a mother is a hepatitis B surface antigen carrier with viral DNA present in her serum, the neonate has an 80% to 90% risk of also becoming a chronic carrier.33 High levels of viremia are the most important risk factor for infection of the infant, with transmission rates decreasing to as low as 10% to 30% when viral DNA is absent from the mother's serum.34 Intraplacental transmission is the most common means of vertical spread, although infection can also occur during delivery and breastfeeding.35
In contrast, rates of transmission of hepatitis C from mother to infant are typically low, ranging from 4% to 10%.35 Risk is increased by high levels of viral RNA in the mother and by prolonged time from rupture of membranes to delivery. Most children who acquire hepatitis C from their mothers do so during delivery, when they are exposed to a high volume of their mother's blood and vaginal fluids. Additional risk factors for the vertical spread of hepatitis C are genotypes 1 and 3, elevated alanine aminotransferase levels, and coinfection with human immunodeficiency virus.35
Laboratory parameters, are expected to fluctuate during pregnancy, although the clinical significance is unknown. Because of the expansion of extracellular fluid, all markers of liver function except for alkaline phosphatase, which is produced by the placenta, tend to decrease. This is particularly true during the second and third trimesters.3 Hepatitis B virus DNA levels typically increase late in pregnancy or in the early postpartum period, with 1 retrospective study showing a mean increase of 0.4 log.36 Several studies have likewise shown hepatitis C RNA levels to increase during the second and third trimesters.37–39
Hepatitis B viral infection itself does not appear to alter fertility, conception, or pregnancy beyond the effects of cirrhosis or liver failure.40 Case reports have reported delivery before 34 weeks, antepartum hemorrhage,41 fetal distress, and meconium peritonitis,42 but these have been isolated experiences.
In mothers affected by hepatitis C, 1 study demonstrated an increased incidence of intrahepatic cholestasis of pregnancy.43 Another suggested that pregnant women with hepatitis C may experience worsening necroinflammatory and fibrotic changes on liver biopsy after pregnancy.44 Other studies, however, have shown no adverse effects on pregnancy or neonatal outcome.45
Caesarean delivery is not advocated in order to reduce transmission of hepatitis B, as studies have not shown it to significantly decrease infection rates.46 Rather, prevention of transmission is accomplished through administration of hepatitis B vaccine and hepatitis B immunoglobulin within 12 hours of birth. The vaccine series must then be completed according to the standard schedule.47, 48
Caesarian delivery remains controversial for mothers with hepatitis C, with 1 multicenter European study demonstrating no protective effect of Caesarean delivery49 and another demonstrating a significantly higher rate of vertical transmission in vaginally delivered infants.50 There are no vaccines or immunoglobulins available at this time for hepatitis C.
According to guidelines published by the American College of Obstetricians and Gynecologists, breastfeeding is considered safe for mothers with hepatitis C.51 In mothers with hepatitis B, breastfeeding is also encouraged, although mothers are advised to hold off until the neonate receives hepatitis B immunoglobulin.52
Antiviral treatment of hepatitis B during pregnancy is controversial. Because the rate of mother-to-child transmission of hepatitis B is dependent on the level of viral DNA in the mother's serum, it has been argued that treating mothers with lamivudine, or more recently, telbivudine results in decreased rates of neonatal infection.53, 54 Lamivudine is designated pregnancy category C. However, extensive experience in pregnant women with human immunodeficiency virus, as reported in a national antiviral pregnancy registry, shows the drug to be well tolerated and safe in both mothers and infants.55 Telbivudine has been deemed category B because animal studies have shown no increase in fetal harm (Table 1). With no long-term studies as of yet, however, antiviral treatment during pregnancy remains a subjective decision between mothers and physicians.
Antiviral treatment of hepatitis C is contraindicated during pregnancy. Although case reports of patients who became pregnant while on interferon showed relative safety of the medication,56 ribavirin is considered pregnancy category X. Animal studies of ribavirin have shown a high incidence of birth defects and spontaneous abortion (Table 1).
Autoimmune Hepatitis (AIH)
Although earlier studies of pregnancy in patients with AIH showed an increased risk of prematurity, low birth weight, and fetal loss,57 recent advances in therapy have modified the prognosis of pregnancy in these patients.
Theoretically, AIH should improve during pregnancy because of changes in female sex hormones and cytokine profiles, resulting in a shift from T helper 1 cells to T helper 2 cells. Indeed, there are multiple reported cases of improvement of AIH during pregnancy.58 However, in a significant number of patients, AIH may by exacerbated by pregnancy, with a rate of intrapartum flare of 12.5% to 21%.59, 60 Postpartum flares of AIH also occur, with exacerbation rates ranging widely, anywhere from 12.5% to 86%.58–61 This discrepancy in data may be due to inconsistent alteration of therapy during pregnancy, with dosing being maintained, reduced, or even discontinued.
Maternal and fetal complications do appear to be slightly increased, although evidence is limited. Patients with AIH have been shown to undergo caesarian section at higher rates than the general population, ranging from 16.1% to 43%.60, 61 The rate of serious maternal complications was as high as 9% in 1 study, but this may have been biased by 1 mother who died of multiorgan failure from a septic abortion and another who developed fulminant hepatic failure requiring emergent liver transplantation and hysterectomy.58 Fetal loss was reported to be as high as 33% in earlier studies,57 but more recent studies have shown rates between 14% and 24%.59, 60
A recent review of pregnancy in AIH, published by Schramm et al.60 in 2006, examined 42 pregnancies in 22 women with AIH. The spontaneous abortion rate in this population was 17%, and the rate of preterm delivery was 17%. Twenty-one percent of patients had AIH flares during pregnancy, and 52% suffered postpartum AIH flares. Interestingly, the presence of anti-Ro/SSA and anti-soluble liver antigen/liver-pancreas antibodies was associated with adverse pregnancy outcomes.
Because of the risk of intrapartum and postpartum flares, it is generally recommended that immunosuppressive therapy be continued during pregnancy. A modest decrease in immunosuppression can often be performed after the third month of pregnancy, when liver test typically improve. However, this is usually followed by a need for increased doses just before delivery and in the postpartum period. Azathioprine has been used successfully during pregnancy in patients with inflammatory bowel disease and rheumatoid arthritis, with benefits that far outweighing its teratogenic potential.62, 63
Although successful pregnancy is possible in patients with AIH, careful monitoring is required because of the higher rate of maternal and fetal complications the unpredictable course of disease during pregnancy.
Other Liver Diseases
Reports of pregnancy outcomes in patients with primary biliary cirrhosis (PBC) are rare, largely because of the negative effect of PBC on fertility.64 It is thought that high levels of estrogen during pregnancy may result in impaired biliary function and altered aminotransferase levels, but the impact of these changes on maternal and fetal outcomes is unknown.
The safety of medications commonly used to treat PBC is also a matter of debate. Although UDCA is thought to be safe during the third trimester, it is unclear whether use in the first trimester can result in fetal harm. The American Association for the Study of Liver Diseases recommends avoiding UDCA in women considering pregnancy.65 Furthermore, breastfeeding is not advised, as it is unknown whether UDCA is excreted into breast milk.66
No well-controlled studies exist on the effects of bile acid sequestrants such as cholestyramine, and caution is advised regarding their use. Cholestyramine is particularly known to interfere with the absorption of fat-soluble vitamins.
Hydroxyzine, often used to treat pruritus in PBC patients, is contraindicated in early pregnancy because of animal studies showing increased fetal abnormalities at high doses. Human studies also show an increased risk of neonatal withdrawal symptoms.
On the other hand, continuation of therapy is advocated in pregnant patients with Wilson's disease. Cessation of therapy can lead to hemolytic crisis, with possible hepatic failure and maternal death.68 Although penicillamine is pregnancy category D, successful outcomes have been reported in patients treated throughout their pregnancies.69 Current guidelines recommend that pregnant women be continued on penicillamine at a reduced dose of 25% to 50% their prepregnancy dose.70 The safety of trientine during pregnancy is less clear, but it can also be considered at reduced doses.
Pregnancy data in women with hemochromatosis are exceedingly rare, given that women are generally diagnosed during their postmenopausal years and that those in their reproductive years often present with secondary amenorrhea.
Pregnancy data are also rare in women with alcoholic liver disease (ALD), as women with ALD are often infertile. Infants born to mothers with ALD who continue to drink are at high risk for fetal-alcohol syndrome, which consists of several abnormalities, including poor growth, central nervous system defects, dysmorphic facial features, and cognitive and behavioral impairments.71
PREGNANCY AFTER LIVER TRANSPLANTATION
The advent of liver transplantation has further changed the course of pregnancy in patients with liver disease. Although rare, like pregnancy in the context of cirrhosis, successful pregnancy after liver transplantation is achievable. As of January 2006, 202 pregnancies and 205 outcomes have been reported in 121 female liver transplant recipients in the National Transplantation Pregnancy Registry.72 Although children born to female liver transplant recipients have a greater risk of prematurity (35% versus 11.0%–12.7%) and low birth weight (34% versus 8.2%) compared to the general population, overall outcome is similar, with no associated malformation patterns reported thus far.72 Recent data, however, indicate that mycophenolate may be associated with first trimester pregnancy loss and an increased risk of congenital malformations.73
Maternal complications have been shown to be increased in liver transplant recipients, particularly pregnancy-induced hypertension (34% versus 4%–10%), preeclampsia (22% versus 6%–8%), and caesarian section (35% versus 20%–25%).72 The incidence of pregnancy-induced or exacerbated hypertension may relate to the type of immunosuppressive regimen used, with the highest risk posed by cyclosporine, followed by tacrolimus and then corticosteroids.74
The rate of acute rejection in pregnant liver recipients is not substantially different, from that of their nonpregnant counterparts. In a 2005 National Transplantation Pregnancy Registry report evaluating 121 liver recipients, 7% developed acute rejection during pregnancy, and 8% suffered graft loss within 2 years of delivery. This risk was highest in women who conceived within 6 months of their transplant.72
Experts recommend that pregnancy be postponed for at least 1 year post-transplant, when graft function is optimal on lower doses of immunosuppression and the risks of acute rejection and opportunistic infection are lower.75 Until then, contraception is advised, preferably using barrier methods, which confer a lower risk of infection or potential drug interaction. Physicians should discuss the timing of pregnancy, methods of contraception, and safety and adverse effects of immunosuppressants with all patients considering pregnancy as part of pretransplant counseling.
Although pregnancy in patients with cirrhosis remains rare, recent improvements in the treatment of cirrhosis have resulted in increased life expectancy and quality of life, making pregnancy a more common occurrence. As growing numbers of patients with cirrhosis become pregnant, physicians will need to become aware of the unique risks and complications that they face. These include a potential for life-threatening variceal bleeding, hepatic decompensation, splenic artery aneurysm rupture, and postpartum hemorrhage. Medications may need to be adjusted during this time period, as several drugs commonly used for liver disease have potential harmful effects on the fetus or may be transmitted to neonates during breastfeeding. Finally, one must consider risks specific to the underlying liver disease, as in chronic viral hepatitis and AIH.