Ferric carboxymaltose versus standard-of-care oral iron to treat second-trimester anaemia in Malawian pregnant women: a randomised controlled trial

Abstract

Background: Anaemia affects 46% of pregnancies in Africa; oral iron is recommended by WHO but uptake and adherence are suboptimal. We tested a single dose of a modern intravenous iron formulation, ferric carboxymaltose, for anaemia treatment in Malawian pregnant women.

Methods: In this open-label, individually randomised controlled trial, we enrolled women with a singleton pregnancy of 13-26 weeks' gestation in primary care and outpatient settings across two regions in southern Malawi. Women were eligible if they had capillary haemoglobin of less than 10·0 g/dL and negative malaria rapid diagnostic test. Participants were randomised by sealed envelope 1:1. Assessors for efficacy outcomes (laboratory parameters and birthweight) were masked to intervention; participants and study nurses were not masked. Participants were given ferric carboxymaltose up to 1000 mg (given once at enrolment in an outpatient primary care setting), or standard of care (60 mg elemental iron twice daily for 90 days), along with intermittent preventive malaria treatment. The primary maternal outcome was anaemia at 36 weeks' gestation. The primary neonatal outcome was birthweight. Analyses were performed in the intention-to-treat population for mothers and liveborn neonates, according to their randomisation group. Safety outcomes included incidence of adverse events during infusion and all adverse events from randomisation to 4 weeks' post partum. The trial is registered with ANZCTR, ACTRN12618001268235. The trial has completed follow-up.

Findings: Between Nov 12, 2018, and March 2, 2021, 21 258 women were screened, and 862 randomly assigned to ferric carboxymaltose (n=430) or standard of care (n=432). Ferric carboxymaltose did not reduce anaemia prevalence at 36 weeks' gestation compared with standard of care (179 [52%] of 341 in the ferric carboxymaltose group vs 189 [57%] of 333 in the standard of care group; prevalence ratio [PR] 0·92, 95% CI 0·81 to 1·06; p=0·27). Anaemia prevalence was numerically lower in mothers randomly assigned to ferric carboxymaltose compared with standard of care at all timepoints, although significance was only observed at 4 weeks' post-treatment (PR 0·91 [0·85 to 0·97]). Birthweight did not differ between groups (mean difference -3·1 g [-75·0 to 68·9, p=0·93). There were no infusion-related serious adverse events or differences in adverse events by any organ class (including malaria; ≥1 adverse event: ferric carboxymaltose 183 [43%] of 430 vs standard of care 170 [39%] of 432; risk ratio 1·08 [0·92 to 1·27]; p=0·34).

Interpretation: In this malaria-endemic sub-Saharan African setting, treatment of anaemic pregnant women with ferric carboxymaltose was safe but did not reduce anaemia prevalence at 36 weeks' gestation or increase birthweight.

Funding: Bill & Melinda Gates Foundation (INV-010612).

Figure 1

Trial profile IPTp=intermittent preventative therapy. RDT=rapid diagnostic test. *Reasons for not meeting eligibility were assessed on the questions on the eligibility data collection forms; woman can be ineligible on more than one of the eligibility criteria. †Defined as those who answered no to the question “Accepts study procedures?” at enrolment. ‡Reasons for not receiving the treatment were collected on the participant randomisation form. §Defined as those who received IPTp with suldadoxine pyrimethamine at enrolment. ¶There were four twins born in the intravenous iron group and four in the standard of care oral iron group. ||The intention-to-treat basis indicates maternal and neonatal outcomes were analysed according to randomly allocated group of the woman, and included all available data. Home visits collecting capillary haemoglobin and adherence discontinued due to COVID-19 restrictions.

Figure 2

Subgroup analyses of the treatment effect on maternal anaemia at 36 weeks' gestation and birthweight according to baseline characteristics (A) A prevalence ratio of ferric carboxymaltose versus standard of care is displayed for anaemia at 36 weeks' gestation using a Poisson model with robust standard errors. Subgroup (main effect) and subgroup-by-treatment-by-visit interaction (and subgroup-by-treatment and subgroup-by-visit interaction) have been added to the model to evaluate how the treatment effect differs between subgroup categories. (B) An absolute mean difference for birthweight between ferric carboxymaltose and standard of care is displayed following fitting a linear regression model. Subgroup (main effect) and subgroup-by-treatment interactions terms have been added to the models to evaluate how the treatment effect differs between subgroup categories. The p values and 95% CIs presented have not been adjusted for multiple comparisons. The intervals cannot be used in place of hypothesis testing. FCM=ferric carboxymaltose. SOC=standard of care.