Approximately 50 000 women of reproductive age in the United States are currently living after kidney transplantation (KT), and another 2800 undergo KT each year. Although KT improves reproductive function in women with ESRD, studies of post-KT pregnancies are limited to a few voluntary registry analyses and numerous single-center reports. To obtain more generalizable inferences, we performed a systematic review and meta-analysis of articles published between 2000 and 2010 that reported pregnancy-related outcomes among KT recipients. Of 1343 unique studies, 50 met inclusion criteria, representing 4706 pregnancies in 3570 KT recipients. The overall post-KT live birth rate of 73.5% (95%CI 72.1–74.9) was higher than the general US population (66.7%); similarly, the overall post-KT miscarriage rate of 14.0% (95%CI 12.9–15.1) was lower (17.1%). However, complications of preeclampsia (27.0%, 95%CI 25.2–28.9), gestational diabetes (8.0%, 95%CI 6.7–9.4), Cesarean section (56.9%, 95%CI 54.9–58.9) and preterm delivery (45.6%, 95%CI 43.7–47.5) were higher than the general US population (3.8%, 3.9%, 31.9% and 12.5%, respectively). Pregnancy outcomes were more favorable in studies with lower mean maternal ages; obstetrical complications were higher in studies with shorter mean interval between KT and pregnancy. Although post-KT pregnancy is feasible, complications are relatively high and should be considered in patient counseling and clinical decision making.
Women of childbearing age (18–49 years) with end-stage renal disease (ESRD) have fertility rates nearly 10 times lower than their healthy counterparts (1). Previous studies have suggested that kidney transplantation (KT) significantly improves reproductive function in ESRD patients, increasing fertility by approximately fourfold compared to dialysis (2–6). Fertility is restored within a few months after KT and safe conception can probably be achieved as soon as 1 year following KT (3,7). Currently, over half of the approximately 100 000 women in the United States living after KT are of childbearing age (8). Thus, preconception counseling, family planning and contraception are pertinent parts of the transplant counseling process.
It has been suggested that KT recipients might be more susceptible to pregnancy complications than their healthy counterparts (3). KT recipients commonly have comorbidities, such as cardiovascular disease and diabetes, that can put both their pregnancy and their allograft at risk (3). In addition, previous abdominal surgery, long-term exposure to risky pharmacological agents and advanced maternal age are other factors that can lead to an increased risk of pregnancy complications (3). Unfortunately, the risks of pregnancy complications and the prepregnancy factors associated with these complications have not been quantified in a generalizable manner (9,10), nor have the effect of pregnancy on allograft survival and the optimal interval between transplantation and pregnancy.
Current knowledge of pregnancy-related outcomes after KT is limited to a few existing registries, namely a National Transplantation Pregnancy Registry (NTPR) in the United States (founded in 1991), a registry in the United Kingdom (established in 1997 and since discontinued) and one in Australia/New Zealand. These are all voluntary registries of posttransplant pregnancy outcomes, directly and electively reported by transplant coordinators, physicians or recipients. Although these registries are excellent resources and offer invaluable information on posttransplantation pregnancy outcomes, as a whole they only represent one-third of the available information on reported pregnancy outcomes in KT recipients.
About two-thirds of the clinically available information on post-KT pregnancy comes from retrospective single-center cohort studies from across the world. Individually, each study is limited to a small, homogenous sample from a single center, but collectively, these studies represent a rich data set which has been unexplored in its total form that can also provide valuable information on post-KT pregnancy outcomes. We hypothesized that an aggregate analysis of registry and single-center studies will provide more generalizable inferences and also allow evaluation of consistency among post-KT pregnancy outcomes on an international level. As such, the primary goal of this study was to systematically identify all studies of pregnancy-related outcomes in KT recipients and estimate pooled incidences of various pregnancy events, obstetric complications and delivery outcomes through meta-analysis. Our secondary goals were to explore prepregnancy factors that may influence both pregnancy and graft outcomes, to examine the influence of pregnancy on the allograft and to assess recommendations for the ideal interval between KT and pregnancy.
Search strategy and study selection
A comprehensive literature search of PubMed/MEDLINE, EMBASE, and Web of Science entries between January 1, 2000 and November 15, 2010 was conducted (Figure 1). Studies in English reporting pregnancy outcomes, obstetric complications or delivery outcomes among KT recipients were eligible. We excluded one article that specifically studied teratogenic effects of mycophenolate mofetil in 26 pregnancies among 18 KT recipients and effects of sirolimus in four pregnancies among four KT recipients (11). In addition, studies involving cancer patients and multiorgan transplant recipients were excluded. All citations of eligible articles as well as relevant review articles were consulted for supplementary references, although none were found that were not already identified in the primary search. Two reviewers independently screened the abstracts of all eligible studies. If eligibility was indeterminable from the abstract, the manuscript was included in the full-text screen. The same two reviewers then screened the articles at the full-text level. All disagreements were adjudicated by the principal investigator.
One independent reviewer abstracted data from all eligible articles. Data elements were then checked for accuracy by a second reviewer. The following data were extracted from each article: study design (country of location, years of data collection and whether or not a control group was used); maternal demographics (number of KT recipients, number of pregnancies, mean age at pregnancy and mean interval between KT and pregnancy); pregnancy outcomes (number of live births, miscarriages, abortions, stillbirths and ectopic pregnancies); obstetric complications (number of women with hypertension, preeclampsia and gestational diabetes); delivery outcomes (number of Cesarean sections, number of preterm deliveries, mean gestational age and mean birth weight) and KT outcomes (number of acute rejections during pregnancy and graft survival after pregnancy). For the qualitative analysis, the following data were extracted if reported: prepregnancy factors associated with pregnancy and graft outcomes; putative associations between pregnancy and graft survival and clinical recommendations about interval time between KT and pregnancy.
For continuous outcomes (gestational age and birth weight) pooled estimates and 95% confidence intervals were calculated using a weighted Graybill-Deal estimator (12). For binary outcomes, pooled incidence estimates and 95% exact binomial confidence intervals were calculated. Using a two-sample test of proportions (13), the pooled incidence for each analysis was compared to the most recent and updated US general population incidence from the National Vital Statistics Reports from 2005 to 2006 (14–16). All analyses were conducted using Stata 11.1/MP (College Station, TX, USA).
Associations between maternal age, interval between KT and pregnancy and pregnancy outcomes were explored with the caveat that individual-level data were not available, so inferences are limited to associations between study mean and outcomes, as opposed to (the more ideal) associations between patient-level characteristics and outcomes. Categories of mean maternal age (years) included <30 and 30+. Categories of mean interval between transplant and pregnancy (years) included: <2, [2–3), [3–4) and 4+.
Of 1632 citations that were initially retrieved from the three electronic database searches, 1343 unique studies qualified for abstract screening. Among these, 67 full-text articles were reviewed and 50 were selected for inclusion (Figure 1); 18 studies were from Europe, 12 from the Middle East, 8 from Asia, 6 from North America, 4 from South America and 2 from Australia (Table 1).
Table 1. Studies of pregnancy-related outcomes in KT recipients included in meta-analysis
Author (Year published)
1Includes data from five Middle Eastern countries: Saudi Arabia, Lebanon, Syria, Turkey and Oman.
Mean maternal age was 29.0 years and mean interval between transplantation and pregnancy was 3.2 years.
Among 4706 pregnancies in 3570 KT recipients, outcomes were known for 4002 pregnancies. There were 2941 live births (73.5%), 560 miscarriages (14.0%), 370 abortions (9.5%), 100 stillbirths (2.5%) and 24 ectopic pregnancies (0.6%) (Table 2). Live birth rate was higher than that of the US general population (73.5% vs. 66.7%) and favorable across all geographic regions: Asia (69%), Australia (76%), Europe (75%), Middle East (79%), North America (71%) and South America (76%) (Figure 2). Miscarriage rate was lower than that of the US general population (14.0% vs. 17.1%) and lower across all geographic regions: Asia (12%), Australia (11%), Europe (13%), Middle East (14%), North America (15%) and South America (11%) (Figure 3).
Table 2. Maternal1 demographics, pregnancy outcomes, obstetric complications and delivery outcomes among kidney transplant recipients
1To maintain consistency across extracted data, number of pregnant KT recipients was used as the denominator for the “Maternal demographics” analysis. To investigate “Pregnancy outcomes,” the number of pregnancies was used as a denominator. Finally, for both the “Obstetric complications” and “Delivery outcomes” analyses, the number of live births plus stillbirths was used as the denominator.
2Comparison data were retrieved from the 2006 US National Vital Statistics Reports. All comparisons are statistically significant.
3Comparison data were retrieved from the 2005 US. National Vital Statistics Reports. All comparisons are statistically significant.
4Spontaneous abortions including intrauterine fetal death and abnormal product of conception.
5Includes therapeutic abortion and abortion otherwise not specified.
6Hypertension includes chronic hypertension (prepregnancy and during-pregnancy hypertension).
7Preterm is defined as any delivery before 37 weeks gestation.
Age at pregnancy
29.0 years (28.9−29.1)
3.2 years (3.1−3.3)
35.6 weeks (35.5−35.7)
2420 grams (2395−2445)
Among nonterminated pregnancies (live births and still births), there were 1075 cases of hypertension (54.2%), 135 cases of gestational diabetes (8.0%) and 611 cases of preeclampsia (27.0%). Gestational diabetes rate was higher than that of the US general population (8.0% vs. 3.9%) and higher across most geographic regions: Asia (10%), Europe (7%), Middle East (7%), North America (9%) and South America (8%), with the exception of Australia (1%) (Figure 4). Preeclampsia rate was also higher than that of the US general population (27.0% vs. 3.8%) and higher across all geographic regions: Asia (30%), Australia (26%), Europe (32%), Middle East (26%), North America (27%) and South America (21%) (Figure 5).
Of nonterminated pregnancies, 1395 resulted in Cesarean section deliveries (56.9%); of deliveries, 1175 resulted in preterm birth (45.6%) (or before 37 weeks of gestation). Cesarean section rate was higher than that of the general US population (56.9% vs. 31.9%), and higher across all geographic regions: Asia (51%), Europe (66%), Middle East (61%), North America (53%) and South America (57%) (Figure 6). Preterm birth rate was also higher than that of the general US population (45.6% vs. 12.5%) and higher across all geographic regions: Asia (41%), Australia (56%), Europe (51%), Middle East (46%), North America (44%) and South America (46%) (Figure 7). The mean gestational age for newborns was 35.6 weeks (US mean 38.7 weeks) (Figure 8) and the mean birth weight was 2420 g (US mean 3298 g) (Figure 9).
Prepregnancy factors associated with adverse pregnancy outcomes
Three commonly reported prepregnancy factors were described in association with adverse pregnancy outcomes: hypertension, elevated serum creatinine (SCr) and proteinuria. The presence of prepregnancy hypertension was associated with intrauterine growth restriction, preterm delivery, miscarriages and low birth weight (4,21,23,35,42,48,53). Many studies suggested that elevated prepregnancy SCr levels and proteinuria predicted postpartum graft loss and adverse pregnancy outcomes (6,21,20,26,35,36,39,42,53,59,65–69). Yildrim et al. reported significant associations between SCr > 1.5 mg/dL and preterm delivery (p = 0.04) (64); similarly, Keitel et al. reported prepregnancy SCr > 1.5 mg/dL in all six patients who suffered postpartum graft loss within 2 years (40). Little et al. reported five patients with SCr > 1.75 mg/dL, one of who had a miscarriage, and four of who delivered prematurely (43). Cruz et al. report lower birth weight and gestational age outcomes as well as increased incidence of Cesarean section among KT recipients with prepregnancy proteinuria >300 mg/24 h and prepregnancy SCr > 1.5 mg/dL (26). Willis et al. also reported this negative association between maternal SCr and gestational age and birth weight, as well as significant differences in prepregnancy maternal Scr of KT recipients who had live births versus miscarriages (62). Finally, Sibanda et al. found that women with elevated prepregnancy SCr also tended to have higher postpartum SCr levels (p = 0.04) (56).
Acute rejection and graft loss
Among 2412 pregnant recipients in cohorts where acute rejection was studied, 102 (4.2%) experienced an episode of acute rejection during their pregnancy, with similar rates in most regions: Asia (4%), Europe (3%), North America (3%), South America (5%) and only slightly higher in the Middle East (8%) (Figure 10). Among 103 pregnant recipients in cohorts where 1-year postpregnancy graft loss was studied, graft loss was 5.8% (Figure 11). Among 1353 recipients followed for 2 years, graft loss was 8.1%. And among 465 recipients followed for 5 years, graft loss was 6.9%. Although the impact of pregnancy on allograft function remains controversial (31), studies suggested that pregnancy did not have a deleterious effect upon the allograft (36,70,71). Several studies compared graft survival of pregnant KT recipients to matched nonpregnant controls and found that no statistically significant difference in graft survival between the two groups (9, 35, 39, 30,47,54). Pour-Reza-Gholi et al. reported that graft outcomes were more favorable in the pregnant group with a 1-year graft survival of 100% in 60 KT recipients, possibly representing a selection bias compared with the general post-KT population (52).
Interval between KT and pregnancy
It is suggested that the interval between transplantation and conception can affect both graft survival as well as maternal-fetal outcomes (65,66,72). Many KT recipients have been counseled to wait 2 years between transplantation and pregnancy (41) and this recommendation was discussed in many studies (5,35,54,64,68). Most recently, the 2003 American Society of Transplantation (AST) Women's Health Committee consensus conference concluded that post-KT pregnancy is safe after 1 year, provided the patient has “adequate and stable graft function, is at low for opportunistic infections,[and] is not taking teratogenic medications” (73). This recommendation of waiting a minimum of 1 year was also echoed by several studies (17,41,74). In one controlled comparison of KT recipients who were pregnant after <1 year and >1 year, Moon et al. found no statistically significant differences in long-term graft survival and kidney function (48).
There were 24 studies with a mean maternal age of <30 years and 15 studies with a mean maternal age of 30+ years (Figure 12A). Outcomes were stratified by age, with more favorable pregnancy outcomes in studies of younger women (live birth rate of 78.3% vs. 70.3% and miscarriage rate of 10.2% vs. 16.0%) (Table 3A). There were 3 studies with a study mean interval <2 years, 10 studies between 2–3 years, 14 studies between 3–4 years and 14 studies with a study mean interval of 4+ years (Figure 12B). Outcomes were also stratified by interval, with more favorable pregnancy outcomes in the <2-year interval following KT (live birth rate of 80.1% vs. 64.2%, 76.1% and 75.4% and miscarriage rate of 8.3% vs. 9.7%, 14.7% and 10.9%) (Table 3B). However, obstetric complications were highest in the <2-year interval following KT (preeclampsia rate of 39.4% vs. 35.5%, 30.1% and 20.6% and gestational diabetes rate of 21.1% vs. 6.7%, 9.2% and 4.3%), and delivery outcomes were also less favorable in this interval (Cesarean section rate of 71.7% vs. 49.1%, 54.7% and 66.1% and preterm birth rate of 65.4% vs. 33.0%. 48.4% and 42.1%).
Table 3. Incidence of pregnancy-related outcomes stratified by (A) study mean maternal age and (B) study mean interval between transplant and pregnancy
Study mean maternal age (years)
(n = 24)1
(n = 15)1
Mean maternal age
Study mean interval between transplant and pregnancy (years)
1n represents the number of studies within each category and refers to study-level, not patient-level data.
n = 31
n = 101
n = 151
n = 151
Mean maternal age
This international systematic review of 50 studies from 25 countries, representing 4706 pregnancies in 3570 KT recipients, confirms that successful pregnancies are viable in KT recipients. The overall live birth rate for KT recipients exceeded the general US population (73.5% vs. 66.7%) and this favorable trend was seen worldwide. Similarly, the chance of miscarriage was low both overall (14.0% vs. 17.1%) and across all geographic locations. Though post-KT pregnancies were viable, the risk of obstetrical complications were high: a higher proportion than the US average of nonterminated pregnancies reported gestational diabetes (8.0%), preeclampsia (27.0%), Cesarean section delivery (56.9%) and preterm birth (45.6%). Overall, KT recipients delivered late preterm (35.6 weeks) and low birth weight (2420 g) babies.
In our analysis, there appeared to be an association between maternal age and pregnancy outcomes, as studies with a younger mean maternal age had greater live birth outcomes and lower incidences of miscarriage and stillbirth. Similarly, studies with a mean interval between KT and pregnancy of <2 years had the most favorable pregnancy outcomes. However, there were only three studies identified in this <2-year interval range. Unfortunately, these findings cannot be applied to the individual (patient) level and are limited to three studies. Furthermore, we were limited in our analysis because no studies had a mean study interval of less than 1 year. Whether or not the initial period immediately following KT is in fact a period particularly sensitive to high-risk maternal-fetal complications, and the physiological and pharmacological reasons behind this sensitivity, merits further investigation.
The conclusion that pregnancies are viable in KT recipients is consistent with data from large voluntary registries in the United States and United Kingdom, both of which have reported that live birth occurs in 71–79% and miscarriage occurs in 12–24% of KT recipients (56,75). Across all locations included in this meta-analysis, the live birth and miscarriage incidences fell within this reported range, suggesting that the NTPR and the UK registry are indeed reasonable clinical reference points. It is critical that transplant and obstetric centers in North America continue to share information with the NTPR. We support and encourage the reporting of posttransplantation pregnancy data with the NTPR in order to ensure its continued success.
Several limitations merit discussion, including patient overlap between studies, unmeasured confounding, differences in classification criteria and reporting bias in studies analyzed. First, while multiple reports of the same KT recipient cohort were purposefully excluded to prevent overlap (i.e. counting the same patient twice if she appears in more than one study), it was not possible to take this precaution with the studies involving registry analyses. It is possible that some KT recipients and pregnancies may be counted multiple times. Second, since this was an unadjusted, international meta-analysis, there are baseline differences in perinatal care, malnutrition, socioeconomics and healthcare infrastructure among the various geographic regions that were unaccounted for. Third, the diagnostic distinction between preeclampsia and hypertension may vary by geographic location. It was unclear if gestational hypertension was reported as unique from hypertension, so all reports of hypertension were grouped together. Also, it was also unclear whether abortions were performed for therapeutic or contraceptive reasons. Finally, the major registries included in our analysis were all voluntary, potentially introducing selection or reporting bias in the studies that were analyzed.
In conclusion, live birth outcomes are possible among KT recipients, and this favorable trend is consistent on the international level. However, special attention should be given to obstetric complications such as hypertension, preeclampsia, gestational diabetes as well as delivery outcomes such as Cesarean section and preterm delivery. The high incidence of these complications supports a high-risk classification of post-KT pregnancies. Furthermore, KT recipients tend to deliver preterm and low birth weight babies. Thus, it is necessary for a multidisciplinary team to be involved in the monitoring and counseling of KT recipients both before and during pregnancy.
We would like to thank Victoria Goode from the Johns Hopkins Welch Medical Library for reviewing the strategies used to search the three electronic clinical databases.
The authors have no conflicts of interest to disclose as described by the American Journal of Transplantation. This study was not funded in any way by a commercial organization.
Appendix A: Keyword Search Strategy
1(“Pregnancy Outcome” OR “Pregnancy Complications” OR “Pre-Eclampsia” OR “Premature Birth” OR “Infant Mortality” OR “Neonatal Mortality” OR “Fetal Death” OR “Perinatal Mortality” OR “Obstetric Labor Complications” OR “Gestational Diabetes” OR “Pregnancy Outcome” OR “Pregnancy Outcomes” OR “Birth Outcome” OR “Obstetric Outcome” OR “Pregnancy complication” OR “Pregnancy complications” OR “Gestational complication” OR “Pre-Eclampsia” OR “Pre Eclampsia” OR “Pre-eclampsia” OR “Pre-eclamptic” OR “Pre eclamptic” OR “Preeclamptic” OR “Eclampsia” OR “Pre-eclamptic toxemia” OR “Pre eclamptic toxemia” OR “Preeclamptic toxemia” OR “Pregnancy Toxemia” OR “Pregnancy Toxemias” OR “Toxemia” OR “Hypertension-Edema-Proteinuria Gestosis” OR “Hypertension Edema Proteinuria Gestosis” OR “Proteinuria-Edema-Hypertension Gestosis” OR “Proteinuria Edema Hypertension Gestosis” OR “EPH Toxemia” OR “EPH Toxemias” OR “EPH Gestosis” OR “Edema-Proteinuria-Hypertension Gestosis” OR “Edema Proteinuria Hypertension Gestosis” OR “Maternal Hypertension” OR “Gestational Hypertension” OR “Hypertension in pregnancy” OR “Pregnancy induced hypertension” OR “Pregnancy-induced hypertension” OR “Premature Birth” OR “Premature Births” OR “Preterm Birth” OR “Preterm Births” OR “Prematurity” OR “Premature labor” OR “Preterm delivery” OR “Infant Mortality” OR “Infant Mortalities” OR “Infant Death” OR “Infant Deaths” OR “Postneonatal Mortality” OR “Neonatal Mortality” OR “Neonatal Mortalities” OR “Neonatal Death” OR “Neonatal Deaths” OR “Newborn Mortality” OR “Newborn Mortalities” OR “Newborn Death” OR “Newborn Deaths” OR “Fetal Mortality” OR “Fetal Mortalities” OR “Fetal Death” OR “Fetal Deaths” OR “Perinatal Mortality” OR “Perinatal Mortalities” OR “Perinatal Death” OR “Perinatal Deaths” OR “Obstetric Labor Complication” OR “Obstetric Labor Complications” OR “Uterine Complication” OR “Uterine Complications” OR “Labor Complication” OR “Labor Complications” OR “Gestational diabetes” OR “Pregnancy-induced diabetes” OR “Pregnancy induced diabetes” OR “Gestational diabetes mellitus” OR “Pregnancy diabetes” OR “Diabetes mellitus gravidarum” OR “Maternal diabetes mellitus”)
2(“Nephrectomy” OR “Kidney Transplantation” OR “Living Donors” OR “Graft Survival” OR “Kidney Failure” OR “Nephrectomy” OR “Nephrectomies” OR “Kidney Transplantation” OR “Kidney Transplantations” OR “Kidney Allograft Transplantation” OR “Kidney Allograft Transplantations” OR “Kidney Allotransplantation” OR “Kidney Allotransplantations” OR “Renal Transplantation” OR “Renal Transplantations” OR “Kidney Graft” OR “Kidney Grafts” OR “Kidney Allograft” OR “Kidney Allografts” OR “Kidney Grafting” OR “Kidney Transplant” OR “Kidney Transplants” OR “Renal Transplant” OR “Renal Transplants” OR “Renal Graft” OR “Renal Grafts” OR “Living Donors” OR “Living Donor” OR “Living Related Donor” OR “Living Related Donors” OR “Live donor” OR “Graft Survival” OR “Graft Survivals” OR “Allograft Survival” OR “Allograft Survivals” OR “Kidney Failure” OR “Kidney Failures” OR “Renal Failure” OR “Renal Failures”)