Impact of mass drug administration with ivermectin, diethylcarbamazine, and albendazole for lymphatic filariasis on hookworm and Strongyloides stercoralis infections in Papua New Guinea

Abstract

Background: Persons with lymphatic filariasis (LF) are often co-infected with soil-transmitted helminths. A single co-administered dose of ivermectin/diethylcarbamazine/albendazole (IDA) is recommended by WHO for mass drug administration (MDA) for LF instead of diethylcarbamazine/albendazole (DA) in Papua New Guinea (PNG). We compared the effectiveness of a single round of MDA with IDA or DA on hookworm and strongyloidiasis in PNG.

Methodology/principal findings: This study was conducted as part of a cluster randomized trial of MDA with IDA versus DA for LF in individuals willing to provide stool and blood samples at baseline and 12 months after MDA. Participants from 23 villages were included in the clinical trial. Primary outcomes were changes in hookworm prevalence and infection intensity assessed by Kato Katz and Strongyloides prevalence by serology. Hookworm prevalence at baseline was 78% (91/117) and 80% (119/149) in villages assigned to DA and IDA treatment, respectively. Twelve months post-MDA, hookworm prevalence decreased to 56.5% in DA- and 34.4% in IDA-treated villages, respectively (p<0.001, both comparisons). The proportion of individuals with moderate to heavy infection (>2000 egg per gram (EPG)) similarly decreased from 8.7% to 1.5% after DA (p = 0.001) and from 5.7% to 1.0% after IDA (p = 0.002). Using a logistic regression model adjusting for age, gender, baseline hookworm prevalence, and village drug coverage, IDA resulted in a 45% greater reduction in hookworm prevalence than DA (Odds ratio 0.55, 95% CI [0.31,0.99], p = 0.049). MDA also reduced hookworm transmission. Strongyloides seroprevalence at baseline was 68% (192/283) and 62% (180/290) in IDA and DA villages, respectively, with 49% becoming seronegative in the IDA versus 23% in DA villages at 12 months (p = 0.0001).

Conclusions/significance: MDA with IDA was more effective than DA for reducing hookworm and Strongyloides infections in PNG, extending the benefit of MDA with IDA beyond its effect on LF.

Fig 1. Study Profile.

The number of individuals providing stool samples in each treatment arm is shown at baseline and 12 months post-MDA. Twenty-one individuals in both treatment arms had paired samples at baseline and 12 months follow-up. Stool samples collected at baseline and at 12 months were from 12 of 12 clusters in the IDA treatment arm and 11 of 12 clusters in the DA arm.

Fig 2. The impact of MDA on hookworm prevalence and intensity of infection stratified by treatment with IDA or DA.

Hookworm prevalence was similar across age groups in participants in both treatment arms. IDA significantly reduced hookworm prevalence more than DA in all age groups (Fig 3).

Fig 3. Hookworm prevalence (upper panel) and intensity (lower panel) by age group and treatment arm at baseline (solid circles) and 12 months post-MDA (clear circles with dashed lines).

Blue are DA-treated communities, and red IDA treated communities. Prevalence or geomean epg and 95% Confidence intervals are shown. (S2 Table for sample size stratified by age).

Fig 4. Hookworm prevalence and intensity 12 months after MDA among individuals who received treatment (N

= 124 in DA and N = 158 IDA villages) or those who did not (N = 186 in DA and N = 182 in IDA villages). Values are mean prevalence (95% CI) or geometric mean epg (95% CI). *** p < 0.001.

Fig 5. Receiver operating characteristic curve for anti-NIE antibodies for S. stercoralis.

Using this cut-off, 192 of 283 (68%) in the IDA arm and 180 of 290 (62%) in the DA arm were antibody positive for the S. stercoralis NIE antigen at baseline. Twelve months following MDA, the seroprevalence decreased in the IDA arm to 88 of 212 (41.5%) compared to 100 of 185 (54%) in the DA arm (p = 0.012, chi-square) (Table 4A). The seroprevalence increased with age and was significantly decreased following IDA treatment in each age group (Fig 6). The seroprevalence did not significantly decrease following DA treatment. These results include all individuals sampled, irrespective of whether they took the medications. For participants with paired samples, baseline S. stercoralis prevalence in the IDA arm was 144 of 209 (69%) and 95 of 179 (53%) in the DA arm with slightly higher baseline OD levels (Table 4B and Fig 7). At 12 months following treatment, 22 individuals seroreverted to negative in the DA arm, whereas 71 seroreverted in the IDA arm (to 49%) (Table 4B). There was a reduction in mean OD (p < 0.0001) in paired samples from participants in the IDA arm and no reduction in mean OD for the DA arm (Fig 7). Twenty-one samples that initially were negative became positive at 12 months in the DA arm, whereas 14 became positive in the IDA arm.

Fig 6. The impact of MDA with DA or IDA on seroprevalence of Strongyloides stratified by age.

Solid triangles indicate prevalence at baseline (95% CI) and open triangles (95% CI) 12 months following MDA.

Fig 7. Changes post-treatment in NIE ELISA values in the IDA arm vs.

DA arm. Data are from paired samples shown in Table 4B. The dashed line represents the threshold for cut-off for a positive sample based on the receiver operating characteristic curve shown in Fig 5. Mean ODs are shown in red lines. **** P < 0.0001 (Wilcoxon matched pairs sign ranked test).