- Email: [email protected]
Pregnancy outcome in liver transplant recipients
Pregnancy Outcome in Liver Transplant Recipients Sandor Nagy, MD, Melissa C. Bush, MD, Richard Berkowitz, MD, Thomas M. Fishbein, MD, and Veronica Gomez-Lobo, MD OBJECTIVE: To evaluate pregnancy course, complications, and outcomes in liver transplant recipients. METHODS: We conducted a retrospective review of 38 pregnancies conceived between 1992 and 2002 in 29 women who underwent liver transplantation at Mount Sinai Medical Center. RESULTS: The most common primary liver disease was autoimmune hepatitis. All patients were on immunosuppressive regimens that included cyclosporine A or tacrolimus. There were four spontaneous first-trimester abortions and ten first-trimester terminations for worsening liver function. The interval from transplantation to pregnancy was shorter in the group that had abortions and terminations (24.4 ⴞ 24.3 months) as compared with the group that had live births (47.8 ⴞ 28.7 months), P ⴝ .02. There were 24 live births to 20 patients. The mean gestational age at delivery was 36.4 weeks, and the mean birth weight was 2762 g. Pregnancy complications included preeclampsia (20.8%), chronic hypertension (20.8%), hemolysis, elevated liver enzymes, low platelets syndrome (8.3%), creatinine 1.3 mg/dL or more (25.0%), anemia (33.3%), diabetes (37.5%), cesarean delivery (45.8%), preterm birth less than 37 weeks (29.2%), intrauterine growth restriction (16.7%), and biopsy-proven graft rejection during pregnancy (16.7%). There were no intrauterine or neonatal deaths. All 5-minute Apgar scores were greater than 7. Four minor congenital anomalies were noted. Before 1997, there were five maternal deaths, 10 –54 months after pregnancy. Pregnancy complications in our population were more common in those patients who delivered from 1992 to 1997 than in those who delivered from 1998 to 2002. CONCLUSION: Pregnancy planned at least 2 years after liver transplantation with stable allograft function can have excellent maternal and neonatal outcome. (Obstet Gynecol 2003;102:121– 8. © 2003 by The American College of Obstetricians and Gynecologists.)
From Petz Aladar County Hospital, Gyo˝r, Hungary; and Mount Sinai Medical Center, New York, New York. The authors thank Rosemarie Gagliardi, EdD, MPH, Lisa McCaskill-Nelson, RN, and Margherita Cassara for their help with data collection.
Liver transplantation is the treatment of choice for many patients with acute and chronic hepatic failure. Approximately 11% of patients who have had liver transplantation are women of reproductive age, and another 15% are female pediatric patients who have greater than 70% chance of surviving long term.1 The first documented pregnancy after orthotopic liver transplantation was hormonally assisted and reported in 1978. As of the year 2000, the US National Transplantation Pregnancy Registry had reports of the outcome of 85 patients who had undergone 142 pregnancies.2 Between June 1988 and May 2002, 2187 orthotopic liver transplantions were performed in 1868 patients at Mount Sinai Medical Center. Of the 771 female recipients, 398 were of reproductive age (ie, between 15 and 55 years old). We report the outcomes of 29 of these patients who had 38 pregnancies between 1992 and 2002 with respect to the safety of the pregnancy for the mother, the fetus, and the transplanted graft.
MATERIALS AND METHODS This is a review of 29 of the 1868 patients who have undergone liver transplantation at Mount Sinai Medical Center between 1992 and 2002. Pregnancy outcomes and complications are reported for a cohort of 38 pregnancies. The pregnancies were self-reported by the patients and logged into the database of the liver transplant coordinator. Patients are counseled before and after transplant regarding pregnancy and contraception, and are encouraged to report all pregnancies. Twenty-six of the 29 patients delivered or terminated their pregnancies at Mount Sinai, whereas three patients delivered at outside institutions. Complications between 1992 and 1997 are compared with those between 1998 and 2002 because immunosuppressive therapy evolved over time. Antenatal management was determined by coordinated care from the maternal–fetal medicine and liver transplant teams in all cases. Routine obstetric evaluations were performed every 2 weeks until 36 weeks’ gestation and weekly thereafter. A first-trimester ultrasound examination was performed to obtain accurate
VOL. 102, NO. 1, JULY 2003 © 2003 by The American College of Obstetricians and Gynecologists. Published by Elsevier.
0029-7844/03/$30.00 doi:10.1016/S0029-7844(03)00369-7
121
Table 1. Liver Transplant Patients With First-Trimester Termination of Pregnancy or Abortion Patient/ pregnancy
Reason for transplant
T-P interval (mo)
Pregnancy outcome
1/A 1/B 2 3* 4 5 6 7/A 7/B 7/C 8 9/A 9/B 10
Autoimmune hepatitis Autoimmune hepatitis Biliary cirrhosis INH toxicity Wilson disease Alcoholic cirrhosis Acetaminophen toxicity Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Biliary atresia
2 60 79 1 47 6 19 2 9 3 22 22 42 28
Missed abortion TOP Spontaneous abortion TOP TOP TOP Missed abortion TOP TOP TOP TOP TOP Spontaneous abortion TOP
Complications 3 tx 3 tx 2 tx Maternal death 2 tx, maternal death 3 tx 3 tx 3 tx 2 tx 2 tx
T-P ⫽ transplantation to pregnancy; INH ⫽ isoniazid; 2 tx and 3 tx ⫽ 2 and 3 liver transplants, respectively, before pregnancy; TOP ⫽ termination of pregnancy. * The patient who had both a TOP and a live birth also is included in Table 2.
dating in all cases. A complete fetal anatomic survey was performed between 18 and 20 weeks, as is standard for all patients in our institution. Fetal growth was measured monthly during the third trimester. Intrauterine growth restriction was defined as estimated fetal weight or birth weight less than 10% for gestational age. If intrauterine growth restriction was detected in utero, fetal surveillance was increased with nonstress tests, Doppler flow measurement of the umbilical vessels, and consideration for early delivery depending on gestational age. Because of an increased risk of carbohydrate intolerance caused by the frequent administration of prednisone to prevent graft rejection, patients were screened with glucose tolerance tests in the first trimester, followed by routine screening between 24 and 28 weeks. Anemia was defined as hemoglobin less than 10 g/dL. Baseline levels of serum creatinine, creatinine clearance, 24-hour urinary protein excretion, and liver function tests, including aspartate transaminase, alanine transaminase, bilirubin, ␥-glutamyltransferase, and lactic dehydrogenase, and a coagulation profile were obtained in the first trimester. Liver function tests were then monitored at least monthly. Hepatitis and cytomegalovirus serologies were also obtained at the first visit along with routine prenatal laboratory tests, human immunodeficiency virus screening, and cervical cytology. Tacrolimus or cyclosporine A trough levels were monitored regularly with adjustment of medication dosage as needed. Graft rejection was detected based on clinical signs, elevations of liver function tests, and liver biopsy results. Patients on prednisone were given stress dose steroids during labor or at the time of cesarean delivery. All patients were delivered by 40 weeks’ gestation to minimize continued fetal exposure to immunosuppres-
122
Nagy et al
Pregnancy After Liver Transplantation
sants. Fetal lung maturity was documented for elective inductions before 39 weeks. Antibiotic prophylaxis was not routinely used for term gestations. Patients were encouraged not to breast-feed because of their immunosuppressive medications. Acyclovir was also widely used to prevent cytomegalovirus infection. Neonatal assessment for effects of maternal immunosuppressive medications was primarily clinical. All infants had a complete blood count checked to rule out bone marrow suppression. Electrolytes were drawn if clinically indicated. Fifteen of the 20 infants born at Mount Sinai had charts that were available for detailed review. Approval for this study was obtained from the Institutional Review Board. Statistical analysis was performed with Student t test and 2 analysis as appropriate, with P ⬍ .05 considered to be significant. RESULTS A total of 29 patients had 38 pregnancies after liver transplantation at Mount Sinai Medical Center between 1992 and 2002 (Tables 1 and 2). The mean maternal age was 27.9 years (range 20 –39 years). The mean number of prenatal visits was 15.7 (range 8 –26 visits). The overall interval from transplantation to pregnancy was 39.2 months (range 1–114 months). The most common primary liver diseases were autoimmune hepatitis (12 patients), primary biliary cirrhosis (five patients), and hepatitis C (three patients) (Table 3). Eight patients had recurrent liver disease requiring more than one liver transplant: Five patients had two liver transplants, and three patients had three liver transplants each. All transplants and ten of 11 retransplants took place before
OBSTETRICS & GYNECOLOGY
Table 2. Liver Transplant Patients With Live Births Patient/ pregnancy
Reason for transplant
T-P interval (mo)
Pregnancy outcome
Birth weight GA at (g) delivery
1
Autoimmune hepatitis
13
NSVD
2435
37 ⫹ 4
2 3 4/A
Autoimmune hepatitis Alcoholic cirrhosis Autoimmune hepatitis
37 42 66
NSVD NSVD LT cesarean
2430 2880 1270
38 ⫹ 0 38 ⫹ 0 28 ⫹ 1
4/B
Autoimmune hepatitis
114
Repeat LT cesarean
3010
38 ⫹ 0
5
Cryptogenic cirrhosis
36
NSVD
1350
30 ⫹ 0
6 7/A
Chronic active hepatitis Primary biliary cirrhosis
100 44
NSVD LT cesarean
3040 4775
39 ⫹ 0 38 ⫹ 0
7/B
Primary biliary cirrhosis
64
Repeat LT cesarean
5160
38 ⫹ 0
7/C
Primary biliary cirrhosis
92
Repeat classic cesarean
3950
37 ⫹ 1
8
INH toxicity
38
LT cesarean
2720
37 ⫹ 6
9 10* 11 12
Chronic hepatitis C INH toxicity Chronic hepatitis C Autoimmune hepatitis
50 39 24 48
NSVD NSVD NSVD Repeat LT cesarean
2630 2930 2770 3940
36 ⫹ 6 37 ⫹ 0 39 ⫹ 6 40 ⫹ 0
13
Chronic hepatitis C
24
LT cesarean
2330
35 ⫹ 2
14
Autoimmune hepatitis
99
Classic cesarean
760
27 ⫹ 2
15
Autoimmune hepatitis
19
LT cesarean
3250
37 ⫹ 1
16
Primary biliary cirrhosis
36
Repeat LT cesarean
1500
34 ⫹ 3
17 18 19 20/A
Acetaminophen toxicity Autoimmune hepatitis Biliary atresia Autoimmune hepatitis
64 11 29 19
NSVD NSVD NSVD NSVD
2100 NA NA NA
20/B
Autoimmune chronic rejection
39
NSVD
NA
Complication Preeclampsia, IUGR, meconium, anemia, rejection, re-tx after pregnancy, maternal death IUGR, hypertension, anemia Anemia Preterm labor, PPROM, HELLP, meconium, anemia, IDDM, moderate rejection during pregnancy Failed attempted VBAC, meconium, BTL, IDDM Fatty liver, rejection before delivery, anemia, diabetes class B, meconium Anemia Diabetes class A2 macrosomia, breech, thrombocytopenia Diabetes class A2 macrosomia, breech Diabetes class A2, placenta previa, BTL Fetal distress, diabetes class B, mild rejection during pregnancy Diabetes class B 3 tx Fetal distress, failed attempted VBAC, meconium, anemia Preeclampsia, anemia, maternal death HELLP, IUGR, absent diastolic flow Induced due to elevated LFT results, hypertension, rejection 6 mo after delivery IUGR, hypertension, gestational DM Preeclampsia
34 ⫹ 1 Full term Full term Hypertension 28 ⫹ 0 Induced for PPROM, severe rejection during pregnancy, 2 tx, maternal death Full term 2 tx, maternal death
GA ⫽ gestational age; NSVD ⫽ normal spontaneous vaginal delivery; tx ⫽ liver transplant; LT cesarean ⫽ low transverse cesarean delivery; IUGR ⫽ intrauterine growth restriction; PPROM ⫽ preterm premature rupture of membranes; HELLP ⫽ hemolysis, elevated liver enzymes, low platelets; IDDM ⫽ insulin-dependent diabetes mellitus; VBAC ⫽ vaginal birth after cesarean; BTL ⫽ bilateral tubal ligation; LFT ⫽ liver function test; DM ⫽ diabetes mellitus; NA ⫽ not applicable. Other abbreviations as in Table 1. The last three patients are those who have missing data regarding birth weight and Apgar scores and exact gestational age in some cases. * The patient who had both a TOP and a live birth also is included in Table 1.
VOL. 102, NO. 1, JULY 2003
Nagy et al
Pregnancy After Liver Transplantation
123
Table 3. Indications for Liver Transplantation
Table 5. Pregnancy Complications Among Those Patients With Live Births
Primary liver disease
Patients who delivered
Patients who had TOP/ abortion
Total n (%)
Autoimmune hepatitis Primary biliary cirrhosis Hepatitis C INH toxicity Alcoholic Acetaminophen toxicity Cryptogenic Wilson disease Chronic active hepatitis Total
8 3 3 2* 1 1 1 0 1 20*
4 2 0 1* 1 1 0 1 0 10
12 (41.3) 5 (17.2) 3 (10.3) 3 (10.3) 2 (6.9) 2 (6.9) 1 (3.5) 1 (3.5) 1 (3.5) 30 (100)
Abbreviations as in Table 1. * One patient who had both a TOP and a live birth had INH toxicity and is included in both columns.
pregnancy. One patient required retransplantation for an episode of rejection 7 months after delivery but died 8 months later. No patient required transplantation during pregnancy or the puerperium. All were maintained on immunosuppressive regimens that included one or more of the following medications: prednisone, cyclosporine A, tacrolimus, azathioprine, and OKT3 (Table 4). Of the 38 pregnancies in this cohort, there were four spontaneous first-trimester abortions and ten terminations of pregnancy in the first trimester (Table 1). All women who terminated their pregnancies had elevated liver function tests that reflected ongoing graft rejection. The ten terminations occurred in eight women, four of whom had undergone more than one liver transplant. Twenty patients had 24 live births (Table 2), one of which involved a patient with three prior liver transplants. There were no ectopic pregnancies. One patient had three pregnancies after her liver transplant, each Table 4. Medication Regimens
Medication regimens Prednisone/CyA/azathioprine Prednisone/tacrolimus Prednisone/CyA Prednisone/tacrolimus/azathioprine CyA/azathioprine Tacrolimus OKT3 Total
Patients Patients who who had delivered TOP/abortion 9 7† 1 1 1 1 1 21*
2 5† 0 1 0 1 1 10†
CyA ⫽ cyclosporine A. Other abbreviations as in Table 1. OKT3 was used only for acute graft rejection and not as maintenance therapy. * One patient changed medication (see text), so total is 21 not 20 patients. † The patient who had both a TOP and a live birth was on prednisone and tacrolimus, and is included in both columns.
124
Nagy et al
Pregnancy After Liver Transplantation
Complication
n
%
Cesarean delivery Diabetes Anemia (Hb ⬍ 10 g/dL) Preterm birth ⬍ 37 wks Creatinine ⱖ 1.3 mg/dL Preeclampsia Chronic hypertension IUGR Graft rejection during pregnancy HELLP syndrome
11 9 8 7 6 5 5 4 4 2
45.8 37.5 33.3 29.2 25.0 20.8 20.8 16.7 16.7 8.3
Hb ⫽ hemoglobin. Other abbreviations as in Table 2.
with live births. Another two each had two live births after liver transplantation. The interval from transplantation to pregnancy was much shorter among patients who had abortions and terminations (24.4 ⫾ 24.3 months) than among those who had live births (47.8 ⫾ 28.7 months; P ⫽ .02). One patient had a termination of pregnancy followed by a live birth 3 years later. The following results are for those 20 patients who had 24 live births. Although in four cases, the infants’ birth weights and Apgar scores were missing (all of those delivered at outside institutions), these infants are included in the remaining analysis. The mean gestational age at delivery was 36.4 ⫾ 3.6 weeks (range 27– 40 weeks, n ⫽ 24). The mean birth weight was 2762 ⫾ 1117 g (range 760 –5160 g, n ⫽ 20). Pregnancy complications are described in Table 5. There were eight preterm births (less than 37 weeks, 29.2%), four of which occurred before 32 weeks of pregnancy. Patients who had graft rejection during pregnancy were more likely to deliver preterm (mean gestational age 30.8 weeks) compared with those who did not have graft rejection (mean gestational age 36.7 weeks), but this difference did not reach statistical significance. There were four cases of intrauterine growth restriction at birth (16.7%). Five pregnancies were complicated by chronic hypertension (20.8%) without superimposed preeclampsia. Another five patients had preeclampsia (20.8%), but not chronic hypertension. Three patients developed preeclampsia before term, necessitating delivery at 27, 28, and 35 weeks, respectively. Two of those patients also had hemolysis, elevated liver enzymes, low platelets syndrome. Six patients had renal dysfunction with a creatinine of 1.3 mg/dL or more. Five of the six patients had renal dysfunction before pregnancy, and four had preeclampsia. No patient had cholestasis of pregnancy. Anemia was present in eight pregnancies (33.3%), but no one required transfusion. Insulin-dependent diabetes was a
OBSTETRICS & GYNECOLOGY
Table 6. Indications for Cesarean Delivery Previous cesarean delivery Preeclampsia Fetal distress Breech presentation Failed VBAC Failed induction Total
Table 7. Trend in Pregnancy Complications Over Time 3 3 2 1 1 1 11
VBAC ⫽ vaginal birth after cesarean.
1992–1997
1998–2002
Complication
n ⫽ 14
%
n ⫽ 10
%
Preterm birth ⬍37 wk Graft rejection Chronic hypertension Creatinine ⱖ 1.3 mg/dL Preeclampsia IUGR
5 5 5 4 3 3
35.7 35.7 35.7 28.6 21.4 21.4
2 1 0 2 2 1
20 10 0 20 20 10
IUGR ⫽ intrauterine growth restriction.
comorbidity in nine pregnancies (37.5%), three of which were pregestational. Four patients had cytomegalovirus immunoglobulin G and M in their serum but not in amniotic fluid. Eleven cesarean deliveries were performed for standard obstetric indications (Table 6). Two classic cesarean deliveries were performed: one for hemolysis, elevated liver enzymes, low platelets syndrome and breech presentation at 27 weeks’ gestation and one third repeat cesarean delivery with tubal ligation for a placenta previa. There were 14 spontaneous vaginal deliveries. Four vaginal deliveries were preterm, and ten were at term. Peripartum infections did not occur. All patients who continued their pregnancies past the first trimester delivered live-born infants. There were no intrauterine or neonatal deaths. The mean birth weight was 2762 g. Infants of patients who did not have graft rejection had a mean birth weight of 2947 ⫾ 1153 g, whereas those of mothers with graft rejection during pregnancy had a mean birth weight of 1780 ⫾ 815 g. However, this difference did not reach statistical significance. All 5-minute Apgar scores were greater than 7 (n ⫽ 20). Five preterm infants required neonatal intensive care unit admission; four were intubated because of prematurity. Two infants had meconium staining. There were four congenital anomalies. Three mothers taking tacrolimus had two infants with small membranous ventricular septal defects, and one infant with bilateral hydroceles. One mother taking cyclosporine A had an infant with hypospadias. Despite maternal immunosuppression, none of these infants had adrenal or respiratory suppression or neonatal lymphopenia. Allograft liver function was stable throughout the first trimester in all patients who continued their pregnancies past this point. Four women with rising liver function tests were biopsied after the first trimester. There were no complications caused by the biopsies. Graft rejection was characterized histologically as mild in one patient, moderate in two, and severe in one. After delivery, two other women with autoimmune hepatitis had moderate graft rejection, 2 and 6 months later, respectively. All rejection episodes were successfully treated during pregnancy, and there were no graft losses, retransplantations,
VOL. 102, NO. 1, JULY 2003
or maternal deaths during pregnancy. One patient required retransplantation 7 months after delivery, for worsening chronic rejection. Of the six patients who had graft rejection, five occurred from 1992 to 1997, and only one occurred between 1998 and 2002. Medication regimens are shown in Table 3. Only two patients required alteration in maintenance medications. One was changed from tacrolimus to cyclosporine A and azathioprine because of moderate rejection, and a second woman who was changed from tacrolimus to OKT3 subsequently developed acute fatty liver of pregnancy. The remainder of the patients were maintained on their prepregnancy medications for the duration of their pregnancies, but doses were increased in the other cases of rejection. Medication usage evolved over time. All but two of the women who delivered in 1997 or earlier were on cyclosporine A, whereas all but one who gave birth after 1997 were on tacrolimus. Pregnancy complications in our population were more common in those patients who delivered from 1992 to 1997 than in those who delivered from 1998 to 2002 (Table 7). There were five maternal deaths, with a mean time of 27.6 months after pregnancy (range 10 –54 months). Two deaths occurred in patients who had terminated their pregnancies, and three in women who had live births. Causes of death were graft rejection in four patients, and cytomegalovirus infection with sepsis in the other. No one died during pregnancy or the puerperium. There was a significant difference between the death rate before 1997 and after 1998 (P ⬍ .01). All five of the deaths occurred before 1997. Three patients who died were on cyclosporine A, and the other two were on tacrolimus-based therapy. DISCUSSION The generally good outcomes for the pregnancies of our patients in this series are consistent with previous reports from a MEDLINE search from 1966 to 2002 using the terms “liver transplantation” and “pregnancy.” Table 8 summarizes the eight articles that report outcomes on
Nagy et al
Pregnancy After Liver Transplantation
125
Table 8. Review of the Literature
Mean # mo T-P (range)
Author Scantlebury et al, 19908 Pittsburgh Ville et al, 199312 Paris Laifer and Guido, 19959 Pittsburgh Jain et al, 19976 Pittsburgh Patapis et al, 19973 Birmingham Rayes et al, 199813 Berlin Wu et al, 199814 Hannover Raakow et al, 200115 Berlin Current series, 2003 New York
Preterm delivery (%)
IUGR (%)
Preeclampsia (%)
Hypertension (%)
C/S rate (%)
Mean birth weight (g)
31.2 (5–132)
55
15
20
30
63
2940
26.4 (8–60)
0
10
27.3
N/R
45.4
2990
10.8 (1–24)
90
10
30
10
60
N/R
46.3 (6–121)
52
0
N/R
44
2638
24 (3–94)
N/R
N/R
33
N/R
N/R
N/R
33.2 (6–75)
7.8
15.4
0
43.7
54
2659
43.1 (4–139)
13.6
4
13.6
13.6
31.8
2876
46 (6–84)
19
28.6
4.8
42.9
47.6
2656
39.2 (1–114)
29.2
16.7
20.8
20.8
45.8
2762
3.7
N/R ⫽ not reported. Other abbreviations as in Tables 1 and 2.
more than ten pregnancies in liver transplant patients between 1990 and 2001. Our series is the largest singlecenter experience reported to date in the above-mentioned literature. Chronic hepatitis C virus infection represents the single leading indication for orthotopic liver transplantation in the United States. In our series, however, autoimmune hepatitis was the most frequent reason for transplantation (Table 3). This is not surprising because autoimmune hepatitis is a more common indication for transplantation in younger, reproductive-age patients. The most frequent maternal complication after orthotopic liver transplantation is pregnancy-induced hypertension, which has been reported in patients taking cyclosporine A, tacrolimus, and/or steroids. Prior reports cite incidences ranging from 14% to 44%, which is considerably higher than rates of 4% to 10% in the general population. In addition, the incidence of preeclampsia is also elevated (up to 33% in some series) among pregnant women after orthotopic liver transplantation.3 In our series, preeclampsia and hypertension each occurred in 21% of our patients, and four of the five preeclamptics were on cyclosporine-based immunosuppression. The incidence of chronic hypertension decreased in our center in the group delivered between 1998 and 2002 when compared with those cared for from 1992 to 1997 (Table 7). This finding suggests that the type of immunosuppressive therapy used may be responsible for the increased incidence of hypertension after liver transplantation.4 It is known that cyclosporine can cause endothelial
126
Nagy et al
Pregnancy After Liver Transplantation
cell dysfunction, decrease endogenous nitric oxide production,5 and be associated with renal dysfunction. In our series, renal dysfunction was more common in patients treated with cyclosporine A than tacrolimus, and four of five preeclamptics had renal dysfunction. A series of 27 pregnancies in 21 patients from the University of Pittsburgh also found that mothers using tacrolimus had a lower incidence of hypertension and preeclampsia than those using cyclosporine A.6 The hemolysis, elevated liver enzymes, low platelets syndrome occurs in 0.25– 0.8% of all pregnancies, but was evident in two of 24 (8%) of our patients. It is not clear whether hepatic transplant per se is responsible for the increased risk of this complication, or whether it is caused by the underlying liver disease, immunosuppressive regimens, or associated chronic hypertension. The rate of cesarean deliveries among orthotopic liver transplantation patients has been reported to be increased (30 – 63%) over that of the general population (20 –25%). A review of the indications for cesarean delivery in our population indicates that most are for obstetric reasons (eg, indicated preterm delivery for severe preeclampsia). This suggests that liver transplant per se does not independently increase the cesarean delivery rate (Table 6). Our cesarean delivery rate remained constant over the 10-year course of the study, but nationally the rate appears to be declining.7 This may reflect greater comfort in managing these women. Other pregnancy-related problems that have been reported in liver transplant recipients include anemia, hy-
OBSTETRICS & GYNECOLOGY
perbilirubinemia,8 and infection,9 but not gestational diabetes. The incidence of diabetes in our series is higher than that in the 18% National Transplantation Pregnancy Registry (38% versus 18%),2 although a similar percentage of the patients in each study were treated with prednisone. This difference might be explained by the implementation of stricter criteria for the diagnosis of gestational diabetes (all the diabetics in the National Transplantation Pregnancy Registry report were pregestational) or by different population demographics. The incidence of intrauterine growth restriction and preterm delivery in our study was consistent with previous reports (Table 8). We had no fetal or neonatal deaths, and all neonatal intensive care unit admissions were because of prematurity. The incidence of malformations in the general population is 2–3%. There is a slightly increased risk of malformations reported in previous studies of liver transplant recipients: 4% in patients taking corticosteroids, 6% in those taking tacrolimus,10 and 7% in those taking azathioprine.7 Our rate of 27%, four of 15, is higher than previously reported. However, the distribution of anomalies could not be attributed to a particular pattern of medication usage, and the anomalies were minor, reparable, and not life threatening. Pregnancy does not appear to alter hepatic allograft function and is not a contraindication to liver biopsy if rejection is suspected. In our series, acute rejection was successfully managed with adjustment of immunosuppressive medications. None of our patients needed liver transplantation during pregnancy, but rejection was higher than that reported by other centers. We must emphasize that all of the rejections occurred before 1997 when older immunosuppressive regimens were the standard of care. The overall patient and graft survival rates 1 year after transplantation at Mount Sinai Hospital are 83% and 73%, respectively, and 56% and 45% at 10 years.11 The Kaplan–Meier 1-year survival rate for women aged 20 –39 transplanted at Mount Sinai between 1992 and 1997 was 84%, which improved to 89% between 1998 and 2002. Maternal death in liver transplant patients related to pregnancy is uncommon. There were no anteor peripartum deaths in our series or in the National Transplantion Pregnancy Registry report. Five of the 29 women in this series (17%) died between 10 and 54 months after delivering, and all of them died before 1997. Only seven maternal deaths were reported in the 85 patients included in the National Transplantation Pregnancy Registry database (8%).7 These patients died 8 –54 months after pregnancy, five of seven of recurrent liver disease. It is difficult to compare these mortality rates given the different methodologies between the two
VOL. 102, NO. 1, JULY 2003
studies. Although pregnancy does not increase maternal mortality in liver transplant patients, the overall longterm outcome needs to be taken into consideration as a couple contemplates parenthood. Although amenorrhea occurs in up to 50% of women with chronic liver disease, after transplantation, many enjoy return to normal health and fertility.10 An ongoing study at Mount Sinai Hospital has revealed that 67% of women younger than age 50 are sexually active both before and after transplantation (unpublished data). In the reproductive-age woman, who has undergone a liver transplant, obstetric care should ideally begin before conception. At our institution, all transplant patients are extensively counseled by obstetrics– gynecologists and the transplant team regarding contraception and pregnancy both before and after transplantation. The National Transplantation Pregnancy Registry advises female organ transplant recipients to wait 1 or 2 years after transplantation before attempting to conceive to insure that the transplanted organ is functioning adequately and to allow for stabilization of the immunosuppressive regimen.7 We strongly advocate delaying conception for at least 24 months after transplantation because only one of our patients who conceived less than 1 year after transplantation had a successful outcome. The remainder either had spontaneous abortions or were advised to terminate in the first trimester for graft rejection. Coordinated efforts between maternal–fetal medicine and surgical transplantation specialists are needed to encourage appropriate contraceptive planning, even pretransplantation, with consideration of concomitant tubal ligation at the time of liver transplantation in those women who have completed childbearing. In summary, pregnancy planned at least 2 years after liver transplantation with stable allograft function can have excellent maternal and neonatal outcome when there is coordination of prenatal care between maternal– fetal medicine and surgical specialists.
REFERENCES 1. Casele HL, Laifer SA. Pregnancy after liver transplantation. Semin Perinatol 1998;22:149–55. 2. Armenti VT, Radomski JS, Moritz MJ, Philips LZ, McGrory CH, Coscia LA. Report from the National Transplantation Pregnancy Registry (NTPR): Outcomes of pregnancy after transplantation. Clin Transpl 2000: 123–34. 3. Patapis P, Irani S, Mirza DF, Gunson BK, Lupo L, Mayer AD. Outcome of graft function and pregnancy following liver transplantation. Transplant Proc 1997;29:1565–6. 4. Roberts M, Brown AS, James OF, Davison JM. Interpretation of cyclosporin A levels in pregnancy following
Nagy et al
Pregnancy After Liver Transplantation
127
5.
6.
7.
8.
9.
10.
11.
128
orthotopic liver transplantation. Br J Obstet Gynaecol 1995;102:570–2. Laifer S, Darby MJ, Scantlebury VP, Harger JH, Caritis SN. Pregnancy and liver transplantation. Obstet Gynecol 1990;76:1083–8. Jain A, Venkataramanan R, Fung JJ, Gartner JC, Lever J, Balan V, et al. Pregnancy after liver transplantation under tacrolimus. Transplantation 1997;64:559–65. Armenti VT, Herrine SK, Radomski JS, Moritz MJ. Pregnancy after liver transplantation. Liver Transpl 2000;6: 671–85. Scantlebury V, Gordon R, Tzakis A, Koneru B, Bowman J, Mazzaferro V, et al. Childbearing after liver transplantation. Transplantation 1990;49:317–21. Laifer SA, Guido RS. Reproductive function and outcome of pregnancy after liver transplantation. Mayo Clin Proc 1995;70:388–94. Kainz A, Harabacz I, Cowlrick IS, Gadgil S, Hagiwara D. Analysis of 100 pregnancy outcomes in women treated systemically with tacrolimus. Transpl Int 2000;13 Suppl 1:S299–S300. Kim-Schluger L, Florman SS, Gondolesi G, Emre S, Sheiner PA, Fishbein TM, et al. Liver transplantation at Mount Sinai. Clin Transpl 2000:247–53.
Nagy et al
Pregnancy After Liver Transplantation
12. Ville Y, Fernandez H, Samuel D, Bismuth H, Frydman R. Pregnancy in liver transplant recipients: Course and outcome in 19 cases. Am J Obstet Gynecol 1993;168:896–902. 13. Rayes N, Neuhaus R, David M, Steinmuller T, Bechstein WO, Neuhaus P. Pregnancies following liver transplantation—How safe are they? A report of 19 cases under cyclosporine A and tacrolimus. Clin Transpl 1998;12: 396–400. 14. Wu A, Nashan B, Messner U, Schmidt HH, Guenther HH, Niesert S, et al. Outcome of 22 successful pregnancies after liver transplantation. Clin Transpl 1998;12: 454–64. 15. Raakow R, Neuhaus R, Bu¨scher U, Schmidt S, Rayes N, Glanemann M, et al. Parenthood following liver transplantation. Transplant Proc 2001;33:1450–2. Address reprint requests to: Veronica Gomez-Lobo, MD, Department of Obstetrics, Gynecology, and Reproductive Science, One Gustave Levy Place, Box 1170, New York, NY 10029; E-mail: [email protected]. Received September 17, 2002. Received in revised form February 3, 2003. Accepted February 6, 2003.
OBSTETRICS & GYNECOLOGY
From Petz Aladar County Hospital, Gyo˝r, Hungary; and Mount Sinai Medical Center, New York, New York. The authors thank Rosemarie Gagliardi, EdD, MPH, Lisa McCaskill-Nelson, RN, and Margherita Cassara for their help with data collection.
Liver transplantation is the treatment of choice for many patients with acute and chronic hepatic failure. Approximately 11% of patients who have had liver transplantation are women of reproductive age, and another 15% are female pediatric patients who have greater than 70% chance of surviving long term.1 The first documented pregnancy after orthotopic liver transplantation was hormonally assisted and reported in 1978. As of the year 2000, the US National Transplantation Pregnancy Registry had reports of the outcome of 85 patients who had undergone 142 pregnancies.2 Between June 1988 and May 2002, 2187 orthotopic liver transplantions were performed in 1868 patients at Mount Sinai Medical Center. Of the 771 female recipients, 398 were of reproductive age (ie, between 15 and 55 years old). We report the outcomes of 29 of these patients who had 38 pregnancies between 1992 and 2002 with respect to the safety of the pregnancy for the mother, the fetus, and the transplanted graft.
MATERIALS AND METHODS This is a review of 29 of the 1868 patients who have undergone liver transplantation at Mount Sinai Medical Center between 1992 and 2002. Pregnancy outcomes and complications are reported for a cohort of 38 pregnancies. The pregnancies were self-reported by the patients and logged into the database of the liver transplant coordinator. Patients are counseled before and after transplant regarding pregnancy and contraception, and are encouraged to report all pregnancies. Twenty-six of the 29 patients delivered or terminated their pregnancies at Mount Sinai, whereas three patients delivered at outside institutions. Complications between 1992 and 1997 are compared with those between 1998 and 2002 because immunosuppressive therapy evolved over time. Antenatal management was determined by coordinated care from the maternal–fetal medicine and liver transplant teams in all cases. Routine obstetric evaluations were performed every 2 weeks until 36 weeks’ gestation and weekly thereafter. A first-trimester ultrasound examination was performed to obtain accurate
VOL. 102, NO. 1, JULY 2003 © 2003 by The American College of Obstetricians and Gynecologists. Published by Elsevier.
0029-7844/03/$30.00 doi:10.1016/S0029-7844(03)00369-7
121
Table 1. Liver Transplant Patients With First-Trimester Termination of Pregnancy or Abortion Patient/ pregnancy
Reason for transplant
T-P interval (mo)
Pregnancy outcome
1/A 1/B 2 3* 4 5 6 7/A 7/B 7/C 8 9/A 9/B 10
Autoimmune hepatitis Autoimmune hepatitis Biliary cirrhosis INH toxicity Wilson disease Alcoholic cirrhosis Acetaminophen toxicity Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Autoimmune hepatitis Biliary atresia
2 60 79 1 47 6 19 2 9 3 22 22 42 28
Missed abortion TOP Spontaneous abortion TOP TOP TOP Missed abortion TOP TOP TOP TOP TOP Spontaneous abortion TOP
Complications 3 tx 3 tx 2 tx Maternal death 2 tx, maternal death 3 tx 3 tx 3 tx 2 tx 2 tx
T-P ⫽ transplantation to pregnancy; INH ⫽ isoniazid; 2 tx and 3 tx ⫽ 2 and 3 liver transplants, respectively, before pregnancy; TOP ⫽ termination of pregnancy. * The patient who had both a TOP and a live birth also is included in Table 2.
dating in all cases. A complete fetal anatomic survey was performed between 18 and 20 weeks, as is standard for all patients in our institution. Fetal growth was measured monthly during the third trimester. Intrauterine growth restriction was defined as estimated fetal weight or birth weight less than 10% for gestational age. If intrauterine growth restriction was detected in utero, fetal surveillance was increased with nonstress tests, Doppler flow measurement of the umbilical vessels, and consideration for early delivery depending on gestational age. Because of an increased risk of carbohydrate intolerance caused by the frequent administration of prednisone to prevent graft rejection, patients were screened with glucose tolerance tests in the first trimester, followed by routine screening between 24 and 28 weeks. Anemia was defined as hemoglobin less than 10 g/dL. Baseline levels of serum creatinine, creatinine clearance, 24-hour urinary protein excretion, and liver function tests, including aspartate transaminase, alanine transaminase, bilirubin, ␥-glutamyltransferase, and lactic dehydrogenase, and a coagulation profile were obtained in the first trimester. Liver function tests were then monitored at least monthly. Hepatitis and cytomegalovirus serologies were also obtained at the first visit along with routine prenatal laboratory tests, human immunodeficiency virus screening, and cervical cytology. Tacrolimus or cyclosporine A trough levels were monitored regularly with adjustment of medication dosage as needed. Graft rejection was detected based on clinical signs, elevations of liver function tests, and liver biopsy results. Patients on prednisone were given stress dose steroids during labor or at the time of cesarean delivery. All patients were delivered by 40 weeks’ gestation to minimize continued fetal exposure to immunosuppres-
122
Nagy et al
Pregnancy After Liver Transplantation
sants. Fetal lung maturity was documented for elective inductions before 39 weeks. Antibiotic prophylaxis was not routinely used for term gestations. Patients were encouraged not to breast-feed because of their immunosuppressive medications. Acyclovir was also widely used to prevent cytomegalovirus infection. Neonatal assessment for effects of maternal immunosuppressive medications was primarily clinical. All infants had a complete blood count checked to rule out bone marrow suppression. Electrolytes were drawn if clinically indicated. Fifteen of the 20 infants born at Mount Sinai had charts that were available for detailed review. Approval for this study was obtained from the Institutional Review Board. Statistical analysis was performed with Student t test and 2 analysis as appropriate, with P ⬍ .05 considered to be significant. RESULTS A total of 29 patients had 38 pregnancies after liver transplantation at Mount Sinai Medical Center between 1992 and 2002 (Tables 1 and 2). The mean maternal age was 27.9 years (range 20 –39 years). The mean number of prenatal visits was 15.7 (range 8 –26 visits). The overall interval from transplantation to pregnancy was 39.2 months (range 1–114 months). The most common primary liver diseases were autoimmune hepatitis (12 patients), primary biliary cirrhosis (five patients), and hepatitis C (three patients) (Table 3). Eight patients had recurrent liver disease requiring more than one liver transplant: Five patients had two liver transplants, and three patients had three liver transplants each. All transplants and ten of 11 retransplants took place before
OBSTETRICS & GYNECOLOGY
Table 2. Liver Transplant Patients With Live Births Patient/ pregnancy
Reason for transplant
T-P interval (mo)
Pregnancy outcome
Birth weight GA at (g) delivery
1
Autoimmune hepatitis
13
NSVD
2435
37 ⫹ 4
2 3 4/A
Autoimmune hepatitis Alcoholic cirrhosis Autoimmune hepatitis
37 42 66
NSVD NSVD LT cesarean
2430 2880 1270
38 ⫹ 0 38 ⫹ 0 28 ⫹ 1
4/B
Autoimmune hepatitis
114
Repeat LT cesarean
3010
38 ⫹ 0
5
Cryptogenic cirrhosis
36
NSVD
1350
30 ⫹ 0
6 7/A
Chronic active hepatitis Primary biliary cirrhosis
100 44
NSVD LT cesarean
3040 4775
39 ⫹ 0 38 ⫹ 0
7/B
Primary biliary cirrhosis
64
Repeat LT cesarean
5160
38 ⫹ 0
7/C
Primary biliary cirrhosis
92
Repeat classic cesarean
3950
37 ⫹ 1
8
INH toxicity
38
LT cesarean
2720
37 ⫹ 6
9 10* 11 12
Chronic hepatitis C INH toxicity Chronic hepatitis C Autoimmune hepatitis
50 39 24 48
NSVD NSVD NSVD Repeat LT cesarean
2630 2930 2770 3940
36 ⫹ 6 37 ⫹ 0 39 ⫹ 6 40 ⫹ 0
13
Chronic hepatitis C
24
LT cesarean
2330
35 ⫹ 2
14
Autoimmune hepatitis
99
Classic cesarean
760
27 ⫹ 2
15
Autoimmune hepatitis
19
LT cesarean
3250
37 ⫹ 1
16
Primary biliary cirrhosis
36
Repeat LT cesarean
1500
34 ⫹ 3
17 18 19 20/A
Acetaminophen toxicity Autoimmune hepatitis Biliary atresia Autoimmune hepatitis
64 11 29 19
NSVD NSVD NSVD NSVD
2100 NA NA NA
20/B
Autoimmune chronic rejection
39
NSVD
NA
Complication Preeclampsia, IUGR, meconium, anemia, rejection, re-tx after pregnancy, maternal death IUGR, hypertension, anemia Anemia Preterm labor, PPROM, HELLP, meconium, anemia, IDDM, moderate rejection during pregnancy Failed attempted VBAC, meconium, BTL, IDDM Fatty liver, rejection before delivery, anemia, diabetes class B, meconium Anemia Diabetes class A2 macrosomia, breech, thrombocytopenia Diabetes class A2 macrosomia, breech Diabetes class A2, placenta previa, BTL Fetal distress, diabetes class B, mild rejection during pregnancy Diabetes class B 3 tx Fetal distress, failed attempted VBAC, meconium, anemia Preeclampsia, anemia, maternal death HELLP, IUGR, absent diastolic flow Induced due to elevated LFT results, hypertension, rejection 6 mo after delivery IUGR, hypertension, gestational DM Preeclampsia
34 ⫹ 1 Full term Full term Hypertension 28 ⫹ 0 Induced for PPROM, severe rejection during pregnancy, 2 tx, maternal death Full term 2 tx, maternal death
GA ⫽ gestational age; NSVD ⫽ normal spontaneous vaginal delivery; tx ⫽ liver transplant; LT cesarean ⫽ low transverse cesarean delivery; IUGR ⫽ intrauterine growth restriction; PPROM ⫽ preterm premature rupture of membranes; HELLP ⫽ hemolysis, elevated liver enzymes, low platelets; IDDM ⫽ insulin-dependent diabetes mellitus; VBAC ⫽ vaginal birth after cesarean; BTL ⫽ bilateral tubal ligation; LFT ⫽ liver function test; DM ⫽ diabetes mellitus; NA ⫽ not applicable. Other abbreviations as in Table 1. The last three patients are those who have missing data regarding birth weight and Apgar scores and exact gestational age in some cases. * The patient who had both a TOP and a live birth also is included in Table 1.
VOL. 102, NO. 1, JULY 2003
Nagy et al
Pregnancy After Liver Transplantation
123
Table 3. Indications for Liver Transplantation
Table 5. Pregnancy Complications Among Those Patients With Live Births
Primary liver disease
Patients who delivered
Patients who had TOP/ abortion
Total n (%)
Autoimmune hepatitis Primary biliary cirrhosis Hepatitis C INH toxicity Alcoholic Acetaminophen toxicity Cryptogenic Wilson disease Chronic active hepatitis Total
8 3 3 2* 1 1 1 0 1 20*
4 2 0 1* 1 1 0 1 0 10
12 (41.3) 5 (17.2) 3 (10.3) 3 (10.3) 2 (6.9) 2 (6.9) 1 (3.5) 1 (3.5) 1 (3.5) 30 (100)
Abbreviations as in Table 1. * One patient who had both a TOP and a live birth had INH toxicity and is included in both columns.
pregnancy. One patient required retransplantation for an episode of rejection 7 months after delivery but died 8 months later. No patient required transplantation during pregnancy or the puerperium. All were maintained on immunosuppressive regimens that included one or more of the following medications: prednisone, cyclosporine A, tacrolimus, azathioprine, and OKT3 (Table 4). Of the 38 pregnancies in this cohort, there were four spontaneous first-trimester abortions and ten terminations of pregnancy in the first trimester (Table 1). All women who terminated their pregnancies had elevated liver function tests that reflected ongoing graft rejection. The ten terminations occurred in eight women, four of whom had undergone more than one liver transplant. Twenty patients had 24 live births (Table 2), one of which involved a patient with three prior liver transplants. There were no ectopic pregnancies. One patient had three pregnancies after her liver transplant, each Table 4. Medication Regimens
Medication regimens Prednisone/CyA/azathioprine Prednisone/tacrolimus Prednisone/CyA Prednisone/tacrolimus/azathioprine CyA/azathioprine Tacrolimus OKT3 Total
Patients Patients who who had delivered TOP/abortion 9 7† 1 1 1 1 1 21*
2 5† 0 1 0 1 1 10†
CyA ⫽ cyclosporine A. Other abbreviations as in Table 1. OKT3 was used only for acute graft rejection and not as maintenance therapy. * One patient changed medication (see text), so total is 21 not 20 patients. † The patient who had both a TOP and a live birth was on prednisone and tacrolimus, and is included in both columns.
124
Nagy et al
Pregnancy After Liver Transplantation
Complication
n
%
Cesarean delivery Diabetes Anemia (Hb ⬍ 10 g/dL) Preterm birth ⬍ 37 wks Creatinine ⱖ 1.3 mg/dL Preeclampsia Chronic hypertension IUGR Graft rejection during pregnancy HELLP syndrome
11 9 8 7 6 5 5 4 4 2
45.8 37.5 33.3 29.2 25.0 20.8 20.8 16.7 16.7 8.3
Hb ⫽ hemoglobin. Other abbreviations as in Table 2.
with live births. Another two each had two live births after liver transplantation. The interval from transplantation to pregnancy was much shorter among patients who had abortions and terminations (24.4 ⫾ 24.3 months) than among those who had live births (47.8 ⫾ 28.7 months; P ⫽ .02). One patient had a termination of pregnancy followed by a live birth 3 years later. The following results are for those 20 patients who had 24 live births. Although in four cases, the infants’ birth weights and Apgar scores were missing (all of those delivered at outside institutions), these infants are included in the remaining analysis. The mean gestational age at delivery was 36.4 ⫾ 3.6 weeks (range 27– 40 weeks, n ⫽ 24). The mean birth weight was 2762 ⫾ 1117 g (range 760 –5160 g, n ⫽ 20). Pregnancy complications are described in Table 5. There were eight preterm births (less than 37 weeks, 29.2%), four of which occurred before 32 weeks of pregnancy. Patients who had graft rejection during pregnancy were more likely to deliver preterm (mean gestational age 30.8 weeks) compared with those who did not have graft rejection (mean gestational age 36.7 weeks), but this difference did not reach statistical significance. There were four cases of intrauterine growth restriction at birth (16.7%). Five pregnancies were complicated by chronic hypertension (20.8%) without superimposed preeclampsia. Another five patients had preeclampsia (20.8%), but not chronic hypertension. Three patients developed preeclampsia before term, necessitating delivery at 27, 28, and 35 weeks, respectively. Two of those patients also had hemolysis, elevated liver enzymes, low platelets syndrome. Six patients had renal dysfunction with a creatinine of 1.3 mg/dL or more. Five of the six patients had renal dysfunction before pregnancy, and four had preeclampsia. No patient had cholestasis of pregnancy. Anemia was present in eight pregnancies (33.3%), but no one required transfusion. Insulin-dependent diabetes was a
OBSTETRICS & GYNECOLOGY
Table 6. Indications for Cesarean Delivery Previous cesarean delivery Preeclampsia Fetal distress Breech presentation Failed VBAC Failed induction Total
Table 7. Trend in Pregnancy Complications Over Time 3 3 2 1 1 1 11
VBAC ⫽ vaginal birth after cesarean.
1992–1997
1998–2002
Complication
n ⫽ 14
%
n ⫽ 10
%
Preterm birth ⬍37 wk Graft rejection Chronic hypertension Creatinine ⱖ 1.3 mg/dL Preeclampsia IUGR
5 5 5 4 3 3
35.7 35.7 35.7 28.6 21.4 21.4
2 1 0 2 2 1
20 10 0 20 20 10
IUGR ⫽ intrauterine growth restriction.
comorbidity in nine pregnancies (37.5%), three of which were pregestational. Four patients had cytomegalovirus immunoglobulin G and M in their serum but not in amniotic fluid. Eleven cesarean deliveries were performed for standard obstetric indications (Table 6). Two classic cesarean deliveries were performed: one for hemolysis, elevated liver enzymes, low platelets syndrome and breech presentation at 27 weeks’ gestation and one third repeat cesarean delivery with tubal ligation for a placenta previa. There were 14 spontaneous vaginal deliveries. Four vaginal deliveries were preterm, and ten were at term. Peripartum infections did not occur. All patients who continued their pregnancies past the first trimester delivered live-born infants. There were no intrauterine or neonatal deaths. The mean birth weight was 2762 g. Infants of patients who did not have graft rejection had a mean birth weight of 2947 ⫾ 1153 g, whereas those of mothers with graft rejection during pregnancy had a mean birth weight of 1780 ⫾ 815 g. However, this difference did not reach statistical significance. All 5-minute Apgar scores were greater than 7 (n ⫽ 20). Five preterm infants required neonatal intensive care unit admission; four were intubated because of prematurity. Two infants had meconium staining. There were four congenital anomalies. Three mothers taking tacrolimus had two infants with small membranous ventricular septal defects, and one infant with bilateral hydroceles. One mother taking cyclosporine A had an infant with hypospadias. Despite maternal immunosuppression, none of these infants had adrenal or respiratory suppression or neonatal lymphopenia. Allograft liver function was stable throughout the first trimester in all patients who continued their pregnancies past this point. Four women with rising liver function tests were biopsied after the first trimester. There were no complications caused by the biopsies. Graft rejection was characterized histologically as mild in one patient, moderate in two, and severe in one. After delivery, two other women with autoimmune hepatitis had moderate graft rejection, 2 and 6 months later, respectively. All rejection episodes were successfully treated during pregnancy, and there were no graft losses, retransplantations,
VOL. 102, NO. 1, JULY 2003
or maternal deaths during pregnancy. One patient required retransplantation 7 months after delivery, for worsening chronic rejection. Of the six patients who had graft rejection, five occurred from 1992 to 1997, and only one occurred between 1998 and 2002. Medication regimens are shown in Table 3. Only two patients required alteration in maintenance medications. One was changed from tacrolimus to cyclosporine A and azathioprine because of moderate rejection, and a second woman who was changed from tacrolimus to OKT3 subsequently developed acute fatty liver of pregnancy. The remainder of the patients were maintained on their prepregnancy medications for the duration of their pregnancies, but doses were increased in the other cases of rejection. Medication usage evolved over time. All but two of the women who delivered in 1997 or earlier were on cyclosporine A, whereas all but one who gave birth after 1997 were on tacrolimus. Pregnancy complications in our population were more common in those patients who delivered from 1992 to 1997 than in those who delivered from 1998 to 2002 (Table 7). There were five maternal deaths, with a mean time of 27.6 months after pregnancy (range 10 –54 months). Two deaths occurred in patients who had terminated their pregnancies, and three in women who had live births. Causes of death were graft rejection in four patients, and cytomegalovirus infection with sepsis in the other. No one died during pregnancy or the puerperium. There was a significant difference between the death rate before 1997 and after 1998 (P ⬍ .01). All five of the deaths occurred before 1997. Three patients who died were on cyclosporine A, and the other two were on tacrolimus-based therapy. DISCUSSION The generally good outcomes for the pregnancies of our patients in this series are consistent with previous reports from a MEDLINE search from 1966 to 2002 using the terms “liver transplantation” and “pregnancy.” Table 8 summarizes the eight articles that report outcomes on
Nagy et al
Pregnancy After Liver Transplantation
125
Table 8. Review of the Literature
Mean # mo T-P (range)
Author Scantlebury et al, 19908 Pittsburgh Ville et al, 199312 Paris Laifer and Guido, 19959 Pittsburgh Jain et al, 19976 Pittsburgh Patapis et al, 19973 Birmingham Rayes et al, 199813 Berlin Wu et al, 199814 Hannover Raakow et al, 200115 Berlin Current series, 2003 New York
Preterm delivery (%)
IUGR (%)
Preeclampsia (%)
Hypertension (%)
C/S rate (%)
Mean birth weight (g)
31.2 (5–132)
55
15
20
30
63
2940
26.4 (8–60)
0
10
27.3
N/R
45.4
2990
10.8 (1–24)
90
10
30
10
60
N/R
46.3 (6–121)
52
0
N/R
44
2638
24 (3–94)
N/R
N/R
33
N/R
N/R
N/R
33.2 (6–75)
7.8
15.4
0
43.7
54
2659
43.1 (4–139)
13.6
4
13.6
13.6
31.8
2876
46 (6–84)
19
28.6
4.8
42.9
47.6
2656
39.2 (1–114)
29.2
16.7
20.8
20.8
45.8
2762
3.7
N/R ⫽ not reported. Other abbreviations as in Tables 1 and 2.
more than ten pregnancies in liver transplant patients between 1990 and 2001. Our series is the largest singlecenter experience reported to date in the above-mentioned literature. Chronic hepatitis C virus infection represents the single leading indication for orthotopic liver transplantation in the United States. In our series, however, autoimmune hepatitis was the most frequent reason for transplantation (Table 3). This is not surprising because autoimmune hepatitis is a more common indication for transplantation in younger, reproductive-age patients. The most frequent maternal complication after orthotopic liver transplantation is pregnancy-induced hypertension, which has been reported in patients taking cyclosporine A, tacrolimus, and/or steroids. Prior reports cite incidences ranging from 14% to 44%, which is considerably higher than rates of 4% to 10% in the general population. In addition, the incidence of preeclampsia is also elevated (up to 33% in some series) among pregnant women after orthotopic liver transplantation.3 In our series, preeclampsia and hypertension each occurred in 21% of our patients, and four of the five preeclamptics were on cyclosporine-based immunosuppression. The incidence of chronic hypertension decreased in our center in the group delivered between 1998 and 2002 when compared with those cared for from 1992 to 1997 (Table 7). This finding suggests that the type of immunosuppressive therapy used may be responsible for the increased incidence of hypertension after liver transplantation.4 It is known that cyclosporine can cause endothelial
126
Nagy et al
Pregnancy After Liver Transplantation
cell dysfunction, decrease endogenous nitric oxide production,5 and be associated with renal dysfunction. In our series, renal dysfunction was more common in patients treated with cyclosporine A than tacrolimus, and four of five preeclamptics had renal dysfunction. A series of 27 pregnancies in 21 patients from the University of Pittsburgh also found that mothers using tacrolimus had a lower incidence of hypertension and preeclampsia than those using cyclosporine A.6 The hemolysis, elevated liver enzymes, low platelets syndrome occurs in 0.25– 0.8% of all pregnancies, but was evident in two of 24 (8%) of our patients. It is not clear whether hepatic transplant per se is responsible for the increased risk of this complication, or whether it is caused by the underlying liver disease, immunosuppressive regimens, or associated chronic hypertension. The rate of cesarean deliveries among orthotopic liver transplantation patients has been reported to be increased (30 – 63%) over that of the general population (20 –25%). A review of the indications for cesarean delivery in our population indicates that most are for obstetric reasons (eg, indicated preterm delivery for severe preeclampsia). This suggests that liver transplant per se does not independently increase the cesarean delivery rate (Table 6). Our cesarean delivery rate remained constant over the 10-year course of the study, but nationally the rate appears to be declining.7 This may reflect greater comfort in managing these women. Other pregnancy-related problems that have been reported in liver transplant recipients include anemia, hy-
OBSTETRICS & GYNECOLOGY
perbilirubinemia,8 and infection,9 but not gestational diabetes. The incidence of diabetes in our series is higher than that in the 18% National Transplantation Pregnancy Registry (38% versus 18%),2 although a similar percentage of the patients in each study were treated with prednisone. This difference might be explained by the implementation of stricter criteria for the diagnosis of gestational diabetes (all the diabetics in the National Transplantation Pregnancy Registry report were pregestational) or by different population demographics. The incidence of intrauterine growth restriction and preterm delivery in our study was consistent with previous reports (Table 8). We had no fetal or neonatal deaths, and all neonatal intensive care unit admissions were because of prematurity. The incidence of malformations in the general population is 2–3%. There is a slightly increased risk of malformations reported in previous studies of liver transplant recipients: 4% in patients taking corticosteroids, 6% in those taking tacrolimus,10 and 7% in those taking azathioprine.7 Our rate of 27%, four of 15, is higher than previously reported. However, the distribution of anomalies could not be attributed to a particular pattern of medication usage, and the anomalies were minor, reparable, and not life threatening. Pregnancy does not appear to alter hepatic allograft function and is not a contraindication to liver biopsy if rejection is suspected. In our series, acute rejection was successfully managed with adjustment of immunosuppressive medications. None of our patients needed liver transplantation during pregnancy, but rejection was higher than that reported by other centers. We must emphasize that all of the rejections occurred before 1997 when older immunosuppressive regimens were the standard of care. The overall patient and graft survival rates 1 year after transplantation at Mount Sinai Hospital are 83% and 73%, respectively, and 56% and 45% at 10 years.11 The Kaplan–Meier 1-year survival rate for women aged 20 –39 transplanted at Mount Sinai between 1992 and 1997 was 84%, which improved to 89% between 1998 and 2002. Maternal death in liver transplant patients related to pregnancy is uncommon. There were no anteor peripartum deaths in our series or in the National Transplantion Pregnancy Registry report. Five of the 29 women in this series (17%) died between 10 and 54 months after delivering, and all of them died before 1997. Only seven maternal deaths were reported in the 85 patients included in the National Transplantation Pregnancy Registry database (8%).7 These patients died 8 –54 months after pregnancy, five of seven of recurrent liver disease. It is difficult to compare these mortality rates given the different methodologies between the two
VOL. 102, NO. 1, JULY 2003
studies. Although pregnancy does not increase maternal mortality in liver transplant patients, the overall longterm outcome needs to be taken into consideration as a couple contemplates parenthood. Although amenorrhea occurs in up to 50% of women with chronic liver disease, after transplantation, many enjoy return to normal health and fertility.10 An ongoing study at Mount Sinai Hospital has revealed that 67% of women younger than age 50 are sexually active both before and after transplantation (unpublished data). In the reproductive-age woman, who has undergone a liver transplant, obstetric care should ideally begin before conception. At our institution, all transplant patients are extensively counseled by obstetrics– gynecologists and the transplant team regarding contraception and pregnancy both before and after transplantation. The National Transplantation Pregnancy Registry advises female organ transplant recipients to wait 1 or 2 years after transplantation before attempting to conceive to insure that the transplanted organ is functioning adequately and to allow for stabilization of the immunosuppressive regimen.7 We strongly advocate delaying conception for at least 24 months after transplantation because only one of our patients who conceived less than 1 year after transplantation had a successful outcome. The remainder either had spontaneous abortions or were advised to terminate in the first trimester for graft rejection. Coordinated efforts between maternal–fetal medicine and surgical transplantation specialists are needed to encourage appropriate contraceptive planning, even pretransplantation, with consideration of concomitant tubal ligation at the time of liver transplantation in those women who have completed childbearing. In summary, pregnancy planned at least 2 years after liver transplantation with stable allograft function can have excellent maternal and neonatal outcome when there is coordination of prenatal care between maternal– fetal medicine and surgical specialists.
REFERENCES 1. Casele HL, Laifer SA. Pregnancy after liver transplantation. Semin Perinatol 1998;22:149–55. 2. Armenti VT, Radomski JS, Moritz MJ, Philips LZ, McGrory CH, Coscia LA. Report from the National Transplantation Pregnancy Registry (NTPR): Outcomes of pregnancy after transplantation. Clin Transpl 2000: 123–34. 3. Patapis P, Irani S, Mirza DF, Gunson BK, Lupo L, Mayer AD. Outcome of graft function and pregnancy following liver transplantation. Transplant Proc 1997;29:1565–6. 4. Roberts M, Brown AS, James OF, Davison JM. Interpretation of cyclosporin A levels in pregnancy following
Nagy et al
Pregnancy After Liver Transplantation
127
5.
6.
7.
8.
9.
10.
11.
128
orthotopic liver transplantation. Br J Obstet Gynaecol 1995;102:570–2. Laifer S, Darby MJ, Scantlebury VP, Harger JH, Caritis SN. Pregnancy and liver transplantation. Obstet Gynecol 1990;76:1083–8. Jain A, Venkataramanan R, Fung JJ, Gartner JC, Lever J, Balan V, et al. Pregnancy after liver transplantation under tacrolimus. Transplantation 1997;64:559–65. Armenti VT, Herrine SK, Radomski JS, Moritz MJ. Pregnancy after liver transplantation. Liver Transpl 2000;6: 671–85. Scantlebury V, Gordon R, Tzakis A, Koneru B, Bowman J, Mazzaferro V, et al. Childbearing after liver transplantation. Transplantation 1990;49:317–21. Laifer SA, Guido RS. Reproductive function and outcome of pregnancy after liver transplantation. Mayo Clin Proc 1995;70:388–94. Kainz A, Harabacz I, Cowlrick IS, Gadgil S, Hagiwara D. Analysis of 100 pregnancy outcomes in women treated systemically with tacrolimus. Transpl Int 2000;13 Suppl 1:S299–S300. Kim-Schluger L, Florman SS, Gondolesi G, Emre S, Sheiner PA, Fishbein TM, et al. Liver transplantation at Mount Sinai. Clin Transpl 2000:247–53.
Nagy et al
Pregnancy After Liver Transplantation
12. Ville Y, Fernandez H, Samuel D, Bismuth H, Frydman R. Pregnancy in liver transplant recipients: Course and outcome in 19 cases. Am J Obstet Gynecol 1993;168:896–902. 13. Rayes N, Neuhaus R, David M, Steinmuller T, Bechstein WO, Neuhaus P. Pregnancies following liver transplantation—How safe are they? A report of 19 cases under cyclosporine A and tacrolimus. Clin Transpl 1998;12: 396–400. 14. Wu A, Nashan B, Messner U, Schmidt HH, Guenther HH, Niesert S, et al. Outcome of 22 successful pregnancies after liver transplantation. Clin Transpl 1998;12: 454–64. 15. Raakow R, Neuhaus R, Bu¨scher U, Schmidt S, Rayes N, Glanemann M, et al. Parenthood following liver transplantation. Transplant Proc 2001;33:1450–2. Address reprint requests to: Veronica Gomez-Lobo, MD, Department of Obstetrics, Gynecology, and Reproductive Science, One Gustave Levy Place, Box 1170, New York, NY 10029; E-mail: [email protected]. Received September 17, 2002. Received in revised form February 3, 2003. Accepted February 6, 2003.
OBSTETRICS & GYNECOLOGY