Q J Med 2003; 96: 837-844
© 2003 Association of Physicians
Pregnancy in renal transplant recipients: the Royal Free Hospital experience
From the 1Centre for Nephrology and 2Department of Obstetrics and Gynaecology, Royal Free and University College School of Medicine, London, UK
Received 17 March 2003 and in revised form 28 August 2003
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: For women with end-stage renal failure of child-bearing age, renal transplantation offers a chance to start a family. Pregnancies in renal transplant recipients involve risks for graft and fetus, and need to be carefully managed.
Aim: To identify graft, fetal and maternal outcomes in our patients, and compare our results with those of the large national transplant registries.
Design: Retrospective case-note review.
Methods: We assessed the outcomes of 48 pregnancies in 24 renal transplant recipients. Obstetric data and renal parameters were examined in 27–30 pregnancies that progressed to delivery.
Results: Mean time from transplantation to pregnancy was 6.5 years, with an unfavourable outcome in patients who conceived within 1 year. There was a 41% incidence of fetal growth restriction (FGR), and 33% of infants were small for gestational age. FGR was associated with maternal hypertension, a pre-pregnancy serum creatinine (SCr) 
133 µmol/l (1.5 mg/dl), calcineurin inhibitors and the use of cardioselective ß blockers. Two patients with pre-pregnancy SCr > 200 µmol/l lost their grafts within 3 years of delivery. A permanent significant decline in graft function occurred in 20%, by 6 months post delivery.
Discussion: FGR with SGA infants occurs frequently. Atenolol should be avoided in pregnancy and Metoprolol should not be combined with calcineurin inhibitors. Pregnancy appeared to have a deleterious effect on graft function in patients with SCr > 155 µmol (1.75 mg/dl). Patients with pre-pregnancy SCr 200 µmol/l are at greatest risk.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
For women of child-bearing age who require renal replacement therapy, renal transplantation will in most cases restore fertility and offer the chance to start a family. However, pregnancy in renal transplant recipients is not without risks. The potential conflict between fetus, mother and allograft function need to be carefully considered.
Since the first description of successful pregnancy in a renal transplant by Murray et al. in 1963, significant changes in immunosuppression and graft outcomes have occurred. There have been a large number of pregnancies in renal transplant recipients worldwide and data from this experience is available from valuable national registries: the National Transplant Register of the US (NTPR)1 and the UK Transplant Pregnancy Register (UKTPR).2
Our experience of pregnancy in renal transplant recipients spans the period from 1976 to 2001. We performed a retrospective study to analyse fetal, maternal and graft outcomes, and then compared our results with those of the large national transplant registries.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
We identified all pregnancies in our renal transplant recipients between 1976 and 2001. The outcomes of pregnancy, with the percentages of live births, miscarriages, therapeutic abortions, ectopic pregnancies and intrauterine deaths, were determined by retrospective review of the case notes.
The following maternal details were obtained: age at the time of conception; time interval between pregnancy and transplantation (calculated to the nearest 6 months); presence of hypertension or other co-morbid conditions; doses of immunosuppression and dose changes made during pregnancy; other medications.
The highest recorded serum creatinine (SCr) and proteinuria (24 h collection or dipstix urine analysis) were obtained at 3 months pre-pregnancy, in each trimester and at a time 3–6 months post partum. The most recent SCr available was used for long-term follow-up data. An increase in SCr > 20% was used to define episodes of allograft dysfunction.2 To assess the short-term effect of pregnancy on maternal graft function, the 3 months pre-pregnancy SCr was compared with both the maximum recorded SCr during pregnancy and the value 3–6 months post partum. To identify the longer-term effect of pregnancy on renal function, the SCr 3–6 months post partum was compared with the most recent SCr. Length of follow-up was expressed in months. The causes of graft loss were ascertained.
Maternal and obstetric complications of pregnancy were reviewed, e.g. new onset hypertension, gestational diabetes, pre-eclampsia (PET), urinary tract infections (UTIs) and incidence of caesarean section and instrumental delivery.
Details of the following neonatal outcomes were recorded: pre-term delivery; fetal growth restriction (FGR) and birth weight small for gestational age (SGA); congenital abnormalities; and infant survival. FGR was diagnosed by the obstetricians on clinical grounds, guided by estimations of fetal growth pattern, amniotic fluid volume and associated biophysical features such as redistribution of blood flow in the fetal circulation. Birth weight was interpreted with respect to gestational age using the growth percentile charts (UK cross-sectional reference data 1996/1), but allowance was not made for maternal stature. Infants were described as SGA when birth weight was less than the 10th percentile for gestational age.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
We identified 48 pregnancies in 24 transplanted patients. Obstetric and maternal outcomes are recorded in Table 1. The 33 live births included one twin pregnancy. Six patients had a second successful pregnancy and two patients had a third successful pregnancy. Six pregnancies in four patients were therapeutically aborted (12.5%), all for social reasons, except for one patient in whom an anencephalic fetus was diagnosed at 16 weeks gestation. Two intra-uterine deaths occurred at 31 and 32 weeks, respectively. Details of the first (from 1976) were not available; the second was complicated by severe FGR. Both patients had a subsequent successful pregnancy.
|
The mean age of patients at the time of pregnancy was 30 years (range 19–39 years). The mean interval from transplantation to pregnancy was 6.5 years (range 6 months–24 years). The outcomes in four pregnancies (8% of total) conceived within a year of transplantation were poor: 1 ectopic pregnancy, 1 miscarriage, 1 stillbirth and 1 live infant, delivered at 32 weeks with severe FGR and infant birth weight < 2nd percentile for gestational age.
The following immunosuppression regimens were used: (a) prednisolone and azathioprine (26/48 pregnancies); (b) prednisolone, azathioprine and cyclosporine (CSA) (14/48); (c) prednisolone and CSA (4/48); (d) prednisolone, azathioprine and tacrolimus (2/48); (e) prednisolone and tacrolimus (2/48). The Sandimmune preparation of cyclosporine was used in our unit from 1985 until 1996, when it was replaced by Neoral. Doses of calcineurin inhibitors were adjusted according to trough levels, and prednisolone and azathioprine doses were tapered using a standardized protocol following transplantation.
Maternal complications
Data on obstetric complications were available in 27 pregnancies that progressed to delivery (see Table 1). No patient was diabetic pre-pregnancy. One recipient developed gestational diabetes controlled by diet alone. Seven pregnancies (26%) in total were complicated by one or more episodes of urinary sepsis, with one patient having six UTIs. Two patients developed severe pyelonephritis which precipitated pre-term labour.
Pre-pregnancy hypertension was found in 21/27 (77%) pregnancies that progressed beyond the first trimester, and 31/46 of all pregnancies (67%). A comprehensive list of anti-hypertensive medications was available in 20 pregnancies. Agents used were: hydralazine (5), methyldopa (1), labetolol (2), atenolol (7), metoprolol (5), nifedipine (6), amlodipine (1), frusemide (2) and bumetanide (1). Two patients conceived on enalapril but were converted to hydralazine in the first trimester. One patient (4%) developed new onset hypertension at 15 weeks gestation.
Pre-eclampsia (PET) developed in 8/27 (29%) pregnancies. The diagnosis of PET remains difficult in renal transplant recipients: five were diagnosed by the obstetricians and a further three were presumptively diagnosed, on the basis of worsening hypertension and increasing proteinuria, for which the treatment was early delivery. PET was superimposed on underlying hypertension in 5/8 pregnancies and in all but one case, developed in primigravidas.
Modes of delivery are described in Table 1.
Short-term effect on graft function (< 6 months post delivery)
Data on the effect of pregnancy on renal function were assessed in 30 pregnancies and compared with the NTPR data in Table 2. There was no significant change in the median creatinine (p = 0.076). Data on serial 24-h proteinuria were available for nine patients, detailed in Table 3.
|
|
There were no proven episodes of transplant rejection associated with pregnancy. Significant renal dysfunction occurred during pregnancy in 5/30 (16.6%) pregnancies, but was reversible in all but two cases. A significant increase in SCr (range 23–60%) that developed in pregnancy or the post partum, persisted beyond 6 months post delivery in 6/30 pregnancies (20%).
Longer-term effect on graft function
The mean length of follow-up was 46 months (range 1–12.6 years). Four grafts failed in this period. The recipient with the highest preconception serum creatinine (235 µmol/l) had a stable creatinine during pregnancy, but this increased by 38% at 6 months and deteriorated inexorably thereafter. A renal biopsy showed end-stage changes, and she recommenced peritoneal dialysis 11 months after delivery. A recipient with biopsy-proven chronic allograft nephropathy 1 year prior to pregnancy and a pre-conception SCr of 203 µmol/l, had transient allograft dysfunction during pregnancy, but a steady deterioration from one year after, and lost her graft within 3 years of pregnancy. Two other recipients lost their grafts: one at 6 years post-delivery due to recurrent UTIs and reflux, and one at 11 years post-pregnancy due to chronic allograft nephropathy.
The allograft function of five further patients deteriorated significantly after the first 6 months. In three patients with good pre-pregnancy graft function (SCr 89–94 µmol/l) and no renal dysfunction during pregnancy, creatinine increased 22–23% within 2–3 years of delivery. In the remaining two cases, the deterioration in graft function was more marked, with SCr increasing by 60% and 68% respectively in the first two years post delivery. One recipient was at risk due to poor pre-conception graft function (SCr 155 µmol/l), but in the other case the deterioration was due to known non-compliance.
Neonatal outcomes (see Table 4)
|
The mean gestational age at birth was 34.9 weeks (range 27–39 weeks). Fifteen of the 17 pre-term (< 37 weeks) deliveries were due to induction of labour or caesarean section for obstetric complications. Four (13.3%) infants were born very pre-term (28–32 weeks) and one at 27 weeks gestation.
Mean infant birth weight was 2204 g (range 537–3230 g) with 50% of infants 'low birth weight' (LBW, < 2500 g). Six out of six mothers of 'very low birth weight' (VLBW, < 1500 g) babies were hypertensive before pregnancy began, and 4/6 had episodes of renal dysfunction during pregnancy.
FGR was diagnosed in 11/27 (40.7%) pregnancies. Ten out of 30 (33.3%) infants were small (< 10th percentile) for gestational age. Cyclosporine/Neoral was associated with FGR in 62% of pregnancies, compared with 20% in prednisolone and azathioprine pregnancies (Table 5). There was an equal incidence of pre-existing hypertension, but more patients with moderate renal impairment in the CSA/Neoral group.
|
The tacrolimus triple therapy group had the lowest mean birth weight and gestational age. This group comprised two infants born at 37–38 weeks gestation, with birth weights above the 25th percentile for gestational age and two very premature babies, born at 27 and 28 weeks gestation respectively, one suffering severe FGR. Both the latter two infants died neonatally.
Maternal factors in pregnancies complicated by FGR or delivering SGA infants (n = 12) are compared with pregnancies without these complications (n = 14) in Table 6. Pre-pregnancy hypertension and cyclosporine immunosuppression were much more frequently identified in the FGR/SGA group
|
The clinical details of 11 pregnancies in eight patients on either atenolol or metoprolol are shown in table 7. Atenolol was used in seven pregnancies, all resulting in LBW infants. There appeared to be an additive effect in recipients on either atenolol or metoprolol, and calcineurin inhibitors.
|
The pregnancies of five recipients with pre-conception SCr > 133 µmol/l (1.5 mg/dl) are reviewed in Table 8. All were complicated by either FGR or PET, and 3/5 had SGA infants. Three patients lost their grafts on long-term follow-up, one within 11 months of delivery.
|
Two infants died neonatally (6%). One died from cerebral haemorrhage due to accidental heparin administration. The other was the smallest (537 g) and most premature infant, delivered at 27 weeks, for severe FGR (< 0.4th percentile) and died from these complications. The mother had pre-pregnancy SCr of 81 µmol/l but creatinine clearance of only 53 ml/min. The pregnancy was complicated by episodes of pyelonephritis and significant graft dysfunction. In addition, antenatal scanning suggested that this fetus had maternally inherited autosomal dominant nephronophthisis with associated short limbs and stature. A further baby, who overcame severe FGR, prematurity and neonatal bowel surgery, succumbed to meningococcaemia at 4 months of age.
Other congenital abnormalities included one baby with a ventricular septal defect and two siblings with spherocytosis, inherited from their father. A further infant had a dysplastic right kidney, ectopic left kidney and bilateral reflux, in the absence of a family history of similar abnormalities.
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Renal transplant recipients may be optimistic about a successful outcome to their pregnancy if they meet the criteria for considering pregnancy, as set out in the European Best Practice guidelines IV.3 In general, our results support these recommendations, but a few issues raised by our data warrant further discussion.
Although our percentage of live births (69%) may seem lower than reported elsewhere (Table 9), this is largely due to the higher percentage of ectopic pregnancies (4%) and terminations (12.5%). The high percentage of terminations highlights the importance of contraceptive counselling in female transplant patients of reproductive age. This is further emphasized by the poor outcome of those pregnancies conceived prior to the recommended 1–2 years post transplant.
|
Pre-pregnancy graft function has a significant influence on both neonatal outcome and maternal graft survival. Important markers of graft function are pre-pregnancy SCr, stability of graft function and levels of proteinuria before, during and after pregnancy.
In our experience a pre-pregnancy SCr 1.5 mg/dl (133 µmol/l) was associated with obstetric complications and VLBW infants in 4/5 cases (Table 8). Four out of five patients with renal dysfunction during pregnancy had very pre-term (< 32 weeks) infants with birth weights less than the 25th percentile for gestational age, including one IUD.
Kozlowska-Boszko suggested that a pre-pregnancy proteinuria 0.3 g/day identifies grafts at risk of progressive deterioration after delivery.4 Three of the six patients with a permanent decline in graft function following pregnancy had 0.3 g proteinuria/day pre-pregnancy.
Considerable uncertainty exists regarding the effect of pregnancy on graft function,5–9 and the mechanisms surrounding this. Some 30% of pregnancies were complicated by episodes of graft dysfunction, either during pregnancy or in the 6 months post-delivery. In 20%, there was a permanent decline in graft function. Two patients were identifiably at risk of graft loss, due to pre-conception SCr > 200 µmol/l.10,11 However, the effect on graft function in the other patients, was less predictable. These patients all had SCr < 125 µmol/l and comprised a disparate group where several mechanisms could be implicated. Proteinuria may be a marker of subclinical chronic rejection4 or alternatively, proteinuria per se may have a detrimental effect on renal function. However, high levels of proteinuria in our series, were not consistently associated with graft deterioration. New-onset hypertension or worsening hypertension from early pregnancy, certainly could have contributed to the decline in renal function in at least two cases. By contrast, PET caused only transient increases in serum creatinine. Urosepsis was associated with graft dysfunction in three cases. In one patient, hypercalcaemia, secondary to hyperparathyroidism could have played a role. No identifiable reason could be found in one case of transient graft dysfunction. Similarly, no clear cause could be found for the permanent decline in renal function in a further pregnancy with FGR, ending in IUD. One could postulate this was due to relative calcineurin nephrotoxicity. Histological examination of the placenta implicated CSA in the extensive obliteration of both fetal and maternal placental vasculature, resembling CSA vasculopathy.
Our long-term follow-up data suggest that a SCr 1.75 mg/dl (155 µmol/l) is associated with a risk of graft deterioration or graft loss (Table 8). Two grafts were lost in relation to pregnancy, one at 11 months post-delivery, and the other at 3 years. Although patients had an increase in serum creatinine during follow-up, and a further graft was lost at 11 years post-pregnancy, these problems may have been time-related and would probably have occurred irrespective of pregnancy. Pregnancy did, however, seem more closely linked to a decline in graft function in the following two cases. The first was a recipient with recurrent UTIs, where episodes triggered a stepwise increase in serum creatinine post-pregnancy, ending in graft loss at 6 years. The other was a patient with pre-pregnancy SCr 155 µmol/l and proteinuria of 0.4 g/day, whose renal function remained stable until 1 year post pregnancy and then increased by 68% over the next year. Renal histology confirmed chronic allograft nephropathy.
The high prevalence of pre-conception hypertension in our series is similar to that from other series (Table 9). Maternal hypertension is one of the most important predictors of low birth weight12 and pre-term delivery.2 Although pre-existing hypertension has previously been found to be more common in CSA-treated patients,12,13 it was as prevalent in our CSA/Neoral as in the azathioprine group of patients (Table 5).
Antihypertensive drug regimens should be adjusted prior to conception, as certain drugs or drug combinations may have detrimental effects on pregnancy. Atenolol is known to increase the vascular resistance in the umbilical artery and fetal aorta,14 and affect fetal growth.15 Metoprolol, also a cardioselective ß-blocker, which is generally considered to be safe in pregnancy, could presumably have a similar effect. The outcomes of the 11 pregnancies in our series seem to reflect this, particularly when either drug was combined with an calcineurin inhibitor. CSA can cause significantly decreased expression of endothelial constitutive nitric oxide synthase activity, with a coexistent increased in placental endothelin-1 expression, which would further compromise uteroplacental blood flow.16
The literature suggests that FGR occurs in some 9–40% of pregnancies.17–19 Despite the high prevalence we observed (40.7%), FGR may have been clinically under-diagnosed in our series. Two infants, whose birth weights were less than the second percentile for gestational age, were not diagnosed antenatally. The 62% prevalence in our CSA/Neoral recipients was particularly high, and could not be attributed solely to pre-pregnancy hypertension. The higher proportion of patients on CSA/Neoral with SCr > 1.5 mg/dl would certainly have played a role (Table 5), as would the combination of CSA and atenolol or metoprolol (Table 6).
We have reported our experience of pregnancy in a cohort of renal transplant recipients. Our results support the observation of others that recipients should wait 1–2 years before contemplating pregnancy. FGR and small size for gestational age were frequent findings (FGR 41%, SGA 33%) and were associated with maternal hypertension, pre-pregnancy creatinine 133 µmol/l (1.5 mg/dl) and the use of calcineurin inhibitors and cardioselective ß-blockers. We recommend that atenolol should be avoided in pregnancy and that metoprolol should not be combined with calcineurin inhibitors.
Pregnancy appeared to have a deleterious effect on graft function in patients with pre-pregnancy serum creatinine exceeding 155 µmol/l (1.75 mg/dl). Recipients with serum creatinine > 200 µmol/l (2.25 mg/dl) were at increased risk of graft loss. These patients should consider very carefully the risk to graft survival, if contemplating pregnancy.
|
Acknowledgments |
|---|
We would like to acknowledge Vanessa Brown, Catherine O’Malley and Simon Reid for their help with the data collection.
|
Footnotes |
|---|
Address correspondence to Dr B.C. Thompson, Centre for Nephrology, Royal Free Hospital, Pond Street, London NW3 2QG. e-mail: sbjbi{at}hotmail.com
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Armenti VT, Radomski JS, Moritz MJ, Gaughan WJ, Phillips LZ, McGrory CH, Coscia LA: Report from the National Transplant Pregnancy Registry (NTPR): Outcomes of pregnancy after transplantation. Clin Transplant 2001; (no volume number):97–105.
2. Briggs D, Davison JM, Redman C, et al. United Kingdom Transplant Pregnancy Register – First Report. United Kingdom Transplant Support Service Authority (UKTSSA) Users’ Bulletin 1999; 33:8–9.
3. European Best Practice Guidelines IV. 10 Pregnancy in renal transplant recipients. Nephrol Dial Transplant 2002; 17 (Suppl. 4):50–5.[Medline]
4. Kozlowska-Boszko B, Lao M, Gaciong Z, et al. Chronic rejection as a risk factor for deterioration of renal allograft function following pregnancy. Transplant Proc 1997; 29:1522–3.[CrossRef][Web of Science][Medline]
5. Sturgiss SN, Davison JM. Effect of pregnancy on long-term function of renal allografts. Am J Kidney Dis 1992; 19:167–72.[Web of Science][Medline]
6. Sturgiss SN, Davison JM. Effect of pregnancy on long-term function of renal allografts: an update. Am J Kidney Dis 1995; 26:54–6.[Web of Science][Medline]
7. First MR, Coombs CA, Weiskittel P, Miodovnik M. Lack of effect of pregnancy on renal allograft survival or function. Nephrology 1995; 59:472–6.
8. Salmela KT, Kyllonen LEJ, Holmberg C, Gronhagen-Riska C. Impaired renal function after pregnancy in renal transplant recipients. Transplantation 1993; 56:1372–5.[Web of Science][Medline]
9. Tanabe K, Kobayashi C, Takahashi K, et al. Long-term renal function after pregnancy in renal transplant recipients. Transplant Proc 1997; 29:1567–8.[CrossRef][Web of Science][Medline]
10. Armenti VT, McGrory CH, Cater JS, et al. Pregnancy outcomes in female renal transplant recipients. Transplant Proc 1998; 30:1732–4.[CrossRef][Web of Science][Medline]
11. Crowe A, Rustom R, Gradden C, et al. Pregnancy does not adversely affect renal transplant function. Q J Med 1999; 92:631–5.
12. Armenti VT, Ahlswede KM, Ahlswede BA, et al. National Transplantation Pregnancy Registry- Outcomes of 154 pregnancies in cyclosporine-treated female kidney transplant recipients. Transplantation 1994; 57:502–6.[Web of Science][Medline]
13. Armenti VT, Ahlswede KM, Ahlswede BA, et al. Variables affecting birth weight and graft survival in 197 pregnancies in cyclosporine-treated female kidney transplant recipients. Transplantation 1995; 59:476–9.[Web of Science][Medline]
14. Montan S, Ingermarsson I, Mar
ál K, Sjoberg NO. Randomised controlled trial of atenolol and pindolol in human pregnancy: effects on fetal haemodynamics. Br Med J 1992; 304:946–9.
15. Butters L, Kennedy S, Rubin PC. Atenolol in essential hypertension in pregnancy. Br Med J 1990; 301:587–9.
16. Di Paolo S, Monno R, Stallone G, et al. Placental imbalance of vasoactive factors does not affect pregnancy outcome in patients treated with Cyclosporine A after transplantation. Am J Kidney Dis 2002; 39:776–83.[Web of Science]
17. Davison JM. Pregnancy in renal allograft recipients: problems, prognosis and practicalities. Baillieres Clin Obstet Gynaecol 1994; 8:501–25.[CrossRef][Web of Science][Medline]
18. Toma H, Tanabe K, Tokumoto T, Kobayashi C, Yagisawa T. Pregnancy in women receiving renal dialysis or transplantation in Japan: a nationwide survey. Nephrol Dial Transplant 1999; 14:1511–16.
19. Jungers P, Chauveau D. Pregnancy in renal disease. Kidney Int 1997; 52:871–85.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  V. Levidiotis, S. Chang, and S. McDonald Pregnancy and Maternal Outcomes Among Kidney Transplant Recipients J. Am. Soc. Nephrol., November 1, 2009; 20(11): 2433 - 2440. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. F. Ross Ethical considerations related to pregnancy in transplant recipients. N. Engl. J. Med., March 23, 2006; 354(12): 1313 - 1316. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||