Alpha Omega Alpha Honor Medical Society

2013 Research Abstract

The effect of Praziquantel treatment of schistosome-infected pregnant women on maternal anemia risk and infant iron endowment

Investigator: Kelsey Ripp

Mentor: Jennifer Friedman, MD, MPH, PhD

Summary of research year: This randomized controlled trial of Praziquantel (PZQ) treatment of schistosomiasis during pregnancy concluded during the previous summer and I spent the year analyzing data and writing abstracts and drafting manuscripts. The NIH cannot release the primary data including randomization status, until pharmacokinetic analyses are completed by a different laboratory. I conducted all of the analyses relevant to one manuscript that was not focused on maternal treatment status, however, I will still need the randomization codes to finalize the analyses and submit for publication.

Background: The highest prevalence of anemia exists in the developing world where its causes are multi-factorial. Interestingly, few studies have examined the burden of non-iron deficiency anemia (NIDA) among women of reproductive age or in the context of pregnancy. In previous work conducted among non-pregnant subjects residing in Schistosoma japonicum endemic villages in Leyte, The Philippines, we demonstrated that the predominant cause of anemia is NIDA, most commonly related to infection with S. japonicum. Praziquantel is an inexpensive and efficacious treatment for schistosomiasis, however, because there is a lack of studies on use during pregnancy and breastfeeding, many women do not receive treatment during these years, with unknown consequences on both mother and child. Despite lack of human trials but given the strong evidence for the safety of Praziquantel from retrospective studies and animal studies, the WHO recently released a recommendation to treat women with Praziquantel during pregnancy, however, many countries and agencies are reluctant to do this without evidence for its safety and efficacy during pregnancy. This RCT of Praziquantel treatment of pregnant women is critical in order to provide information on the efficacy and importance of the treatment of pregnant women for schistosomiasis and the outcomes in both mother and child.

Prelim trial results: 373 women were enrolled in the study, 370 provided a blood sample at 12 weeks gestation, 364 provided a blood sample at 32 weeks, cord blood samples were captured for 349 women, and newborn blood samples were captured for 359. The mean hemoglobin was 11.7g/dL at 12 weeks and 11.0g/dL at 32 weeks. At 12 weeks, 25% of mothers were anemic, with 53% of them having non-iron deficiency anemia (NIDA) and 47% iron deficiency anemia (IDA). At 32 weeks, 47% of mothers were anemic, with 20% of those having NIDA and 80% IDA. Of those with NIDA at 12 weeks gestation, 38% became non-anemic by 32 weeks, 21% remained with NIDA, and 40% became ID. The relatively high number of women moving from NIDA to no anemia (compared to women who began with IDA), suggests that PZQ treatment may have played a role, however, we cannot determine this until we have the randomization codes available.

Maternal anemia type and newborn inflammatory and iron markers.
There was no significant difference in mean newborn (NB) hemoglobin, cord blood (CB) ferritin, CB serum transferrin receptor (sTfR) or birth weight by maternal anemia type at 32 weeks. However, mean newborn hemoglobin was lower in women with anemia (IDA and NIDA combined) than no anemia (F=6.28, P=0.01). Using Wilcoxon rank-sum tests, median maternal hepcidin level was significantly greater in women with NIDA versus IDA (P <0.0001), and the median is significantly higher in women with NIDA than women with no anemia (P=0.0001), suggesting hepcidin may be a valid bio-marker for inflammation induced anemia during pregnancy, a group for whom this has not yet been studied. For mothers with NIDA, maternal hepcidin and CB ferritin were negatively correlated (ρ= -0.37, P=0.04), suggesting maternal NIDA may limit iron delivery to the developing fetus.
We also assessed whether maternal iron status had an effect on newborn iron status across the cohort and found a positive correlation between maternal ferritin at 32 weeks and CB ferritin (ρ=0.12, P=0.02), and a positive trend between maternal and newborn hemoglobin (ρ=0.10, P=0.06). Maternal sTfR, a marker of tissue iron thirst, and CB ferritin were negatively correlated (ρ= -0.11, P =.03).

Influence of newborn iron status on anemia risk at 6 and 12 months
CB ferritin was positively correlated with 6 month and 12 month hemoglobin (ρ=0.19, P<.005 and ρ=0.16, P=0.006), and when the data was broken down further by maternal anemia yes/no, this correlation only holds true for anemic mothers (IDA and NIDA) (ρ=0.36, P<.00001 at 6 mos and ρ=0.16, P<0.05 at 12 mos). CB and NB (newborn) ferritin do not correlate with infants’ ferritin at 6 or 12 months. CB sTfR is not correlated with 6 or 12 month hemoglobin or ferritin.

Conclusions: While it appears from the preliminary data that maternal anemia type is not correlated with newborn iron acquisition, we are waiting on the randomization of treatment status. This will help us to determine not only if treatment of schistosomiasis can help mitigate anemia of inflammation in mothers, but also if treatment has an effect on newborn iron status. In addition, there appears to be a correlation between newborn iron status, based on CB ferritin, and anemia risk at 6 and 12 months. However, this trend is complicated by lack of correlation between CB ferritin and 6 and 12 month ferritin. Finally, CB sTfR was not a good predictor of anemia risk.

Additional research: While waiting for the NIH and FDA to release the randomization status, I worked on analyzing data from a similar longitudinal pregnancy cohort of mothers with and without S. japonicum from the same region in Leyte. This data looked more specifically at the differences between mothers with and without schistosomiasis and their newborns and was not focused on treatment. This allowed for a greater focus on inflammatory and iron signaling in the fetal compartment.

Background: Schistosomiasis is associated with microbial translocation (MT) as evidenced by higher levels of endotoxin in peripheral and placental blood. There is a paucity of data addressing whether inflammation in the fetal compartment influences iron uptake in utero. Given ferroportin is located on the fetal side of the placenta and hepcidin causes its degradation, we hypothesized that elevated cord blood (CB) hepcidin would be associated with reduced newborn iron stores.

Methods: We enrolled pregnant women in S. japonicum endemic villages in Leyte, The Philippines with singleton pregnancies in the first or second trimester. We captured helminth infection status and SES. At 32 weeks gestation, a blood sample was obtained for a complete hemogram, iron and inflammatory biomarkers. At delivery, birth weight, gestational age and a CB sample were captured. Maternal anemia status at 32 weeks gestation was: 1) nonanemic (hgb ≥11g/dl) or anemic with etiology as 2) iron-deficient (IDA) if ferritin <30μg/l) or 3) non-iron deficient (NIDA) if ferritin ≥30μg/l. The primary outcome, newborn iron endowment, was captured as sTfR:LogFerritin ratio given it is less affected by inflammation. Multivariate models were made to adjust for potential confounders.

Results: The mean hemoglobin was 10.3g/dL. At 32 weeks, 18% of the mothers had NIDA and of those with NIDA, 70% were infected with S. japonicum, suggesting that schistosomiasis may be the predominant cause of NIDA among the pregnant women in this study. After adjusting for SES, S. japonicum infected women showed significantly higher levels of CB endotoxin (β =0.06; P<0.03). Mothers with NIDA exhibited higher levels of placental endotoxin and had babies with higher levels of cord blood endotoxin than mothers without anemia or with IDA (F=10.8, P<0.0001). CB endotoxin levels positively correlated with IL-6 (ρ =0.53; P<0.0001). After adjusting for SES and gestational age, there was a significant relationship between CB sTfR:Logferritin ratio and both maternal and CB hepcidin (β =0.41; P<0.01 and β =0.42; P<0.03 respectively).

Discussion: MT may play a role in the pathogenesis of adverse pregnancy outcomes by 1) promoting maternal inflammatory responses to endotoxin with increased risk of NIDA and decreased iron bioavailability and 2) promoting fetal inflammation, which may lead to degradation of placental ferroportin. CB endotoxin was associated with a pro-inflammatory cytokine profile in the cord blood, including IL-1, IL-6, and TNF-alpha. These findings suggest that newborns born to mothers with NIDA showed signs of inflammation in the fetal compartment. Although there was no significant difference in mean sTfR or sTfR:ferritin ratio by maternal anemia type, there was a trend towards decreased birth weight in mothers with NIDA, and birth weight was negatively correlated with CB stfr:ferritin ratio and a negative trending with CB endotoxin. It is not surprising that maternal anemia did not correlate with newborn iron endowment given the large number of studies to suggest that at mild to moderate anemia, the fetus is able to extract sufficient iron and acts relatively independent of the mother’s iron status. However, it is interesting that there was a trend towards lower birth weight in mothers with NIDA; one possible association being that of increased endotoxin levels.

Updated on July 29, 2014.

© 2017 Alpha Omega Alpha Honor Medical Society