Impact of Sulfadoxine-Pyrimethamine and Dihydroartemisinin-Piperaquine as Intermittent Preventive Treatment in Pregnancy on Stool Antimicrobial Resistance Gene Abundance

ABSTRACT.

Increasing antimicrobial resistance (AMR) is a global public health emergency. Although chemoprevention has improved malaria-related pregnancy outcomes, the downstream effects on AMR have not been characterized. We compared the abundance of 10 AMR genes in stool samples from pregnant women receiving sulfadoxine-pyrimethamine (SP) as intermittent preventive treatment against malaria in pregnancy (IPTp) to that in samples from women receiving dihydroartemisinin-piperaquine (DP) for IPTp. All participants had at least one AMR gene at baseline. Mean quantities of the antifolate gene dfrA17 were increased after two or more doses of SP (mean difference = 1.6, 95% CI: 0.4–2.7, P = 0.008). Antimicrobial resistance gene abundance tended to increase from baseline in SP recipients compared with a downward trend in the DP group. Overall, IPTp-SP had minimal effects on the abundance of antifolate resistance genes (except for dfrA17), potentially owing to a high starting prevalence. However, the trend toward increasing AMR in SP recipients warrants further studies.

INTRODUCTION

Antimicrobial resistance (AMR) represents a global threat to public health.^1–3 The WHO’s multifaceted plan to combat AMR calls for generating knowledge to provide evidence-based insights on controlling AMR emergence.⁴ Based on compelling evidence of AMR being driven by the volume of antimicrobial use, a major objective of the WHO’s plan hinges on antimicrobial stewardship.⁴

To combat stagnation in malaria control, the WHO recently expanded chemoprevention recommendations to include not only pregnancy and seasonal malaria chemoprophylaxis but also perennial malaria chemoprophylaxis and intermittent preventive treatment in school-aged children, which use sulfadoxine-pyrimethamine (SP) alone or in combination with other agents. Sulfadoxine-pyrimethamine works by inhibiting intermediate reactions in the folate synthesis pathway in Plasmodium species; sulfadoxine inhibits dihydropteroate synthase, whereas pyrimethamine inhibits dihydrofolate reductase.⁵ The WHO recommends providing a dose of SP to pregnant women residing in areas of moderate- to high-malaria endemicity as intermittent preventive treatment against malaria in pregnancy (IPTp) at every scheduled visit in the second and third trimesters at least 1 month apart.⁶ Despite increasing resistance to SP as an antimalarial, IPTp-SP continues to reduce maternal anemia, treat common concurrent maternal infections, and increase fetal birth weight; thus, it is still recommended for IPTp.^7–10

Although bacteria and protozoa have different isoforms of folate metabolism enzymes, SP has demonstrated broad-spectrum antibiotic activity in in vitro studies.⁵ This raises concerns that routine IPTp-SP administration could select for AMR genes in intestinal bacteria (e.g., dfrA17), which confers resistance to trimethoprim.^10–12 Carriage of antimicrobial-resistant bacteria has been linked to infection with AMR bacteria.¹³

MATERIALS AND METHODS

Using longitudinally collected stool samples from an open-label, double-blind, randomized controlled trial of IPTp-SP versus IPTp with dihydroartemisinin-piperaquine (DP), we evaluated the dynamics of AMR genes to determine selective effects of the IPTp regimens. Pregnant women between 16 and 28 weeks’ gestation attending their first antenatal care appointment at the Machinga District Hospital in the southern region of Malawi were enrolled in a randomized controlled trial comparing the efficacy and safety of monthly administration of SP versus DP as IPTp [ClinicalTrials.gov #NCT03009526]. The study was described in detail previously.⁸ A subset of 100 women were enrolled in a substudy to explore the effects of SP on the gut microbiome; all of these participants except those who were unable to provide a baseline stool sample were included in this analysis. Subsequent monthly stool samples were collected within 24 hours from the time of their study visit using a pre-provided stool collection kit. Stool samples were processed, stored, and transported as previously described.⁸ Institutional review board approval was obtained from the University of North Carolina, Centers for Disease Control and Prevention, and the University of Malawi College of Medicine, Blantyre, Malawi.

DNA was extracted as previously described.^8,10 An abundance of 10 bacterial AMR genes, Sul1, Sul2, ErmB, Oxa-48, CatB3, CTX-M-15, Oxa-1, dfra17, TetA, and TetB, were assessed by quantitative polymerase chain reaction (qPCR) using previously described primers (Supplemental Table 1).¹² Details of PCR assays are described in the Supplemental Materials. Assays consisted of a duplex reaction for the AMR gene and a universal 16S ribosomal RNA gene target to allow normalization for bacterial quantity in the extract. The PCR assays were run with two nontemplate controls (water) and a serial dilution of synthesized plasmids (in duplicate) containing both targets.

STATISTICAL ANALYSES

Quantification of AMR gene abundance was done using the delta delta cycle threshold (ΔΔCT) method (as described in the Supplemental Materials). t-Tests were used for the log mean relative quantities of AMR genes after two and three doses of IPTp. We correlated log mean relative quantities of AMR genes with month after enrollment to establish trends in the mean relative abundance of AMR genes. Two-sided alternative hypotheses were used for all analyses, and α was set at a significance level of 0.05. Data were analyzed using Microsoft Excel and Stata 18.

Ninety-three participants provided a baseline stool sample and were included in this substudy; 47 received IPTp-SP, and 46 received IPTp-DP at each monthly antenatal care visit. All baseline samples analyzed (93/93) had at least one bacterial AMR gene detected. Sul1 had the highest baseline prevalence at 97.8% (91/93); Oxa48 and CatB3 had the lowest baseline prevalence at 1.1% (1/93) and 4.3% (4/93), respectively. We saw no differences in baseline prevalence by study arm (Table 1).

Table 1.

Cohort baseline detection of AMR genes in stool samples

AMR Gene	Baseline Prevalence	Baseline AMR Prevalence (SP)	Baseline AMR Prevalence (DP)
Sul 1	97.8% (91/93)	97.9% (46/47)	97.8% (45/46)
Sul 2	93.5% (87/93)	91.4% (43/47)	95.7% (44/46)
Tet A	88.1% (82/93)	87.2% (41/47)	89.1% (41/46)
Tet B	86.0% (80/93)	89.4% (42/47)	82.6% (38/46)
dfrA17	86.0% (80/93)	91% (43/47)	80.4% (37/46)
Oxa-1	36.6% (34/93)	38.3% (18/47)	34.8% (16/46)
Oxa-48	1.1% (1/93)	2.1% (1/47)	0% (0/46)
CTX-M-15	8.6% (8/93)	8.5% (4/47)	8.7% (4/46)
CatB3	4.3% (4/93)	4.3% (2/47)	4.3% (2/46)
ErmB	11.8% (11/93)	12.8% (6/47)	10.8% (5/46)

AMR = antimicrobial resistance; DP = dihydroartemisinin-piperaquine; SP = sulfadoxine-pyrimethamine. Real-time quantitative polymerase chain reaction was used to detect the presence and quantities of AMR genes in baseline stool samples. At least one tested AMR gene was detected in all participants.

To evaluate the potential impact of IPTp on bacterial AMR gene emergence, we compared gene quantities at baseline and the participants’ third and fourth visits (after two and three monthly doses of IPTp, respectively). Two and three monthly doses of IPTp-SP, but not IPTp-DP, were associated with an increase in mean dfrA17 quantities compared with baseline (mean difference = 1.6, 95% CI: 0.4–2.7, P = 0.008; mean difference = 2.2, 95% CI: 0.8–3.5, P = 0.003, respectively). Two SP doses were associated with a reduction in mean Sul2 quantities (P = 0.015). However, this was not observed after three SP doses. The mean quantity of all other genes in the IPTp-SP arm was similar to baseline. Sparsely amplified genes were not reliable given their low numbers and were not included in Figure 1. We did not observe a strong association between study arm and mean AMR gene quantities over time (Figure 2). However, we observed a weak positive correlation between IPTp-SP administration and mean AMR gene quantities for genes that directly affect the bacterial folate synthesis pathway (Sul1, Sul2, and dfrA17); these genes tended to decline over time in IPTp-DP recipients. The beta lactamase gene Oxa-1 was selected for in SP but not in DP recipients (Supplemental Figure 1). However, the overall prevalence was low.

Figure 1.

Mean difference between baseline log ΔΔCT expression of AMR genes and log ΔΔCT expression after three doses of SP versus DP. (A and B) We performed a paired t-test to compare the relative abundance of the mean AMR gene at baseline with the relative abundance of the mean AMR gene after (A) two and (B) three monthly doses of SP or DP. ΔΔCT = delta delta CT; AMR = antimicrobial resistance; DP = dihydroartemisinin-piperaquine; SP = sulfadoxine-pyrimethamine.

Figure 2.

Correlation between mean log relative abundance of AMR genes to 16S rRNA gene and monthly ANC visit stool sample. A weak positive correlation was observed in all folate synthesis pathway AMR genes with monthly SP administration. A general decline in mean folate synthesis pathway AMR gene quantities was observed with monthly IPTp-DP administration. The error bars represent the spread of the gene quantity estimates around the mean at each time point. Fractions on the time axis represent the numbers of AMR (gene-specific) and 16S gene–positive samples relative to the total number of samples available at that time point. AMR = antimicrobial resistance; ANC = antenatal care; DP = dihydroartemisinin-piperaquine; IPTp = intermittent preventive treatment against malaria in pregnancy; rRNA = ribosomal RNA; SP = sulfadoxine-pyrimethamine.

DISCUSSION

The emergence of bacterial AMR genes represents a critical global health concern that poses multifaceted challenges to both clinical medicine and public health.³ In 2015, the WHO put together a global action plan to combat the emergence of AMR aimed at safeguarding treatment efficacy.⁴ A key objective was to strengthen the knowledge and evidence base on AMR to improve understanding of potential causative factors and to mitigate potential downstream effects.⁴ This study examined the interplay between malaria chemoprophylactic interventions and the potential increase of AMR gene abundance in intestinal organisms by comparing the relative quantities of AMR genes in pregnant women receiving IPTp-SP versus IPTp-DP. We observed weak evidence supporting selection of antifolate resistance genes (e.g., sul1, sul2, and dfrA17), but the association seems to be inversely associated with the starting prevalence of the mutation. Impacts on other AMR genes were not significant.

The increased quantities of dfrA17 after two doses of monthly SP administration indicates the potential for repeated SP dosing to increase resistance against folate inhibition–based antibiotic therapy. Although the widespread use of trimethoprim-sulfamethoxazole for prevention or treatment of multiple conditions has been demonstrated to lower the occurrence of SP resistance in previous studies, it is important to monitor associated genetic resistance markers as countries consider expanding malaria prophylaxis with SP to broader populations.^14–18 Moreover, cotrimoxazole resistance markers have been shown to persist despite discontinuation of medication in children receiving prophylactic doses, potentially explaining the high starting prevalence of sul1 and sul2 in this cohort.¹⁹ No increase in gene quantity associated with monthly SP exposure was observed among the genes with higher starting prevalence in the population (e.g., sul1 and sul2). Although the effects of preventive SP appear to be minimal on AMR gene emergence, the trends are interesting and support a wider evaluation, especially given the increasing use of SP for chemoprevention in children.²⁰

Details on the characteristics of gut pathogens after assigned IPTp regimen administration in this cohort have been reported elsewhere.⁸ Importantly, there was no significant difference in amounts of various gut pathogens between arms, except for a dose-dependent reduction in enteroaggregative Escherichia coli observed in SP but not DP recipients, potentially attributable to the inherent antibiotic properties of SP.⁸ There also did not appear to be a significant reduction in bacterial diversity. However, normalization of AMR gene quantities to 16S inherently mitigated selective effects of bacterial load reduction, if any existed. Although we did not explore sequencing of samples to assess bacterial diversity more definitively, future projects in this domain could benefit from 16S gene sequencing at baseline to better characterize the potential of malaria chemoprevention with SP to disrupt the gut microbiome.²¹

Strengths of this study include using samples from a randomized controlled trial. In addition, inclusion of non–folate metabolism resistance genes allowed us to examine effects of SP administration on other clinically significant AMR genes. However, the relatively small sample size and the fact that not all women provided a third stool sample leading to less-precise estimates, as well as the fact that we were only able to perform experiments in singlets because of limited material, remain limitations.

CONCLUSION

In conclusion, we examined the effects of chemoprophylactic SP administration and potential emergence of bacterial AMR genes in stool. By highlighting the potential consequences of specific interventions, such as IPTp-SP, on AMR gene emergence, we have contributed to the growing body of knowledge aimed at addressing the global challenge of AMR. This pilot showed that IPTp-SP was associated with an increase in dfrA17 gene content in stool and a trend toward increases in other antifolate genes. Genes associated with resistance to other antimicrobials were not significantly affected. This suggests that SP may not cause selection for other resistance phenotypes (potentially from multiple genes on the same plasmid). Additional larger studies assessing other populations should be conducted to confirm these findings and provide additional data to policymakers for assessing the risk and benefits of the use of SP as a malaria chemoprophylactic agent.

Note: Supplemental material appears at www.ajtmh.org.