The safety of double- and triple-drug community mass drug administration for lymphatic filariasis: A multicenter, open-label, cluster-randomized study

Abstract

Background: The Global Programme to Eliminate Lymphatic Filariasis (GPELF) provides antifilarial medications to hundreds of millions of people annually to treat filarial infections and prevent elephantiasis. Recent trials have shown that a single-dose, triple-drug treatment (ivermectin with diethylcarbamazine and albendazole [IDA]) is superior to a two-drug combination (diethylcarbamazine plus albendazole [DA]) that is widely used in LF elimination programs. This study was performed to assess the safety of IDA and DA in a variety of endemic settings.

Methods and findings: Large community studies were conducted in five countries between October 2016 and November 2017. Two studies were performed in areas with no prior mass drug administration (MDA) for filariasis (Papua New Guinea and Indonesia), and three studies were performed in areas with persistent LF despite extensive prior MDA (India, Haiti, and Fiji). Participants were treated with a single oral dose of IDA (ivermectin, 200 μg/kg; diethylcarbamazine, 6 mg/kg; plus albendazole, a fixed dose of 400 mg) or with DA alone. Treatment assignment in each study site was randomized by locality of residence. Treatment was offered to residents who were ≥5 years of age and not pregnant. Adverse events (AEs) were assessed by medical teams with active follow-up for 2 days and passive follow-up for an additional 5 days. A total of 26,836 persons were enrolled (13,535 females and 13,300 males). A total of 12,280 participants were treated with DA, and 14,556 were treated with IDA. On day 1 or 2 after treatment, 97.4% of participants were assessed for AEs. The frequency of all AEs was similar after IDA and DA treatment (12% versus 12.1%, adjusted odds ratio for IDA versus DA 1.15, 95% CI 0.87-1.52, P = 0.316); 10.9% of participants experienced mild (grade 1) AEs, 1% experienced moderate (grade 2) AEs, and 0.1% experienced severe (grade 3) AEs. Rates of serious AEs after DA and IDA treatment were 0.04% (95% CI 0.01%-0.1%) and 0.01% (95% CI 0.00%-0.04%), respectively. Severity of AEs was not significantly different after IDA or DA. Five of six serious AEs reported occurred after DA treatment. The most common AEs reported were headache, dizziness, abdominal pain, fever, nausea, and fatigue. AE frequencies varied by country and were higher in adults and in females. AEs were more common in study participants with microfilaremia (33.4% versus 11.1%, P < 0.001) and more common in microfilaremic participants after IDA than after DA (39.4% versus 25.6%, P < 0.001). However, there was no excess of severe or serious AEs after IDA in this subgroup. The main limitation of the study was that it was open-label. Also, aggregation of AE data from multiple study sites tends to obscure variability among study sites.

Conclusions: In this study, we observed that IDA was well tolerated in LF-endemic populations. Posttreatment AE rates and severity did not differ significantly after IDA or DA treatment. Thus, results of this study suggest that IDA should be as safe as DA for use as a MDA regimen for LF elimination in areas that currently receive DA.

Trial registration: Clinicaltrials.gov registration number: NCT02899936.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. CONSORT diagram.

Treatment allocation was by cluster randomization at each study site. DA, diethylcarbamazine plus albendazole; FAS, full analysis set; IDA, ivermectin plus diethylcarbamazine and albendazole.

Fig 2. Mf prevalence (%) by sex and country.

Error bars indicate 95% CIs. CI, confidence interval; Mf, microfilaremia; PNG, Papua New Guinea.

Fig 3. Forest plot showing adjusted odds ratios for factors associated with AEs following treatment for lymphatic filariasis.

Odds ratios were assessed relative to the listed reference groups. P values for comparisons to reference group: *<0.05, **<0.01, ***<0.001. Note that both unadjusted and adjusted models contain a random effect to account for correlation among subjects within a locality. Only the adjusted odds ratios and 95% CIs are plotted. All univariable models had a total N of 26,186 (3,147 with AEs) except for the infection group model, which had an N of 25,978 (3,122 with AEs) because of missing values for CFA or Mf. The multivariable model excluded all subjects with missing values for a total N of 25,978. Indonesian participants with Mf were considered to be CFA(+) for this analysis, although sometimes this was not the case. That is because Brugia infections are not detected by the CFA test. (-), negative results; (+), positive results; AE, adverse event; CFA, circulating filarial antigenemia; CI; confidence interval; DA, double-drug therapy (albendazole, diethylcarbamazine); IDA, triple-drug therapy (albendazole, diethylcarbamazine, ivermectin); MF, microfilaremia; PNG, Papua New Guinea.

Fig 4. Frequencies of the most commonly observed AEs by treatment regimen expressed as percentages of participants who were assessed for AEs after treatment.

In this analysis, participants can only count once for each AE type (e.g., a subject can only have a single headache). However, participants may have multiple different AE types (e.g., if a subject experiences a headache and fatigue, then they will be included in the numerator for both of these AE categories). AE, adverse event; DA, double-drug therapy (diethylcarbamazine, albendazole); IDA, triple-drug therapy (ivermectin, diethylcarbamazine, albendazole).