Community deworming alleviates geohelminth-induced immune hyporesponsiveness

Abstract

In cross-sectional studies, chronic helminth infections have been associated with immunological hyporesponsiveness that can affect responses to unrelated antigens. To study the immunological effects of deworming, we conducted a cluster-randomized, double-blind, placebo-controlled trial in Indonesia and assigned 954 households to receive albendazole or placebo once every 3 mo for 2 y. Helminth-specific and nonspecific whole-blood cytokine responses were assessed in 1,059 subjects of all ages, whereas phenotyping of regulatory molecules was undertaken in 121 school-aged children. All measurements were performed before and at 9 and 21 mo after initiation of treatment. Anthelmintic treatment resulted in significant increases in proinflammatory cytokine responses to Plasmodium falciparum-infected red blood cells (PfRBCs) and mitogen, with the largest effect on TNF responses to PfRBCs at 9 mo-estimate [95% confidence interval], 0.37 [0.21-0.53], P value over time (P_time) < 0.0001. Although the frequency of regulatory T cells did not change after treatment, there was a significant decline in the expression of the inhibitory molecule cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) on CD4⁺ T cells of albendazole-treated individuals, -0.060 [-0.107 to -0.013] and -0.057 [-0.105 to -0.008] at 9 and 21 mo, respectively; P_time = 0.017. This trial shows the capacity of helminths to up-regulate inhibitory molecules and to suppress proinflammatory immune responses in humans. This could help to explain the inferior immunological responses to vaccines and lower prevalence of inflammatory diseases in low- compared with high-income countries.

Keywords: Indonesia; albendazole; cytokine responses; deworming; helminths.

Fig. S1.

Trial consort diagram. The current study is nested within the ImmunoSPIN trial, with a total of 4,004 individuals in two participating villages who were allocated to placebo and albendazole treatment. For the immunological studies, a random selection was made. Cytokines were assessed for 1,059 subjects, of which 572 were in the placebo and 487 in the albendazole arm. After 9 mo, 504 and 428 and, after 21 mo, 434 and 364 individuals were analyzed in the placebo and albendazole group, respectively. Availability of CBCs and parasitological data are indicated at the different time points for both groups.

Fig. S2.

Prevalence and intensity infections at posttreatment time points. Helminth prevalence was assessed at baseline and 9 and 21 mo posttreatment by PCR (Ascaris and hookworms) or microscopy (Trichuris). Helminth infection intensity is given for baseline, 9-mo, and 21-mo time points, based on cycle threshold values derived from PCR analysis of stool samples. Positive Ct values were grouped into three categories—Ct < 30.0, 30.0 ≤ Ct < 35.0, and ≥35.0—representing a high (dark gray bars), moderate (medium gray bars), and low (light gray bars) DNA load, respectively. The white bars represent the uninfected. (Left) Data for N. americanus infection; (Right) A. lumbricoides infections. The P values were generated from linear mixed models of the combined effect of albendazole treatment over time, which were significant (P < 0.001) for any helminth and for each of the species separately.

Fig. 1.

The effect of anthelminthic treatment on cytokine responses to AscAg, PfRBCs, and PHA. TNF, IFN-γ, IL-2, IL-5, and IL-10 concentrations were assessed in supernatants of 72 h-stimulated whole-blood cultures. The values on the y axis (the spider web lines) represent the estimated outcome (beta) of the effect of albendazole treatment on cytokine responses to PHA (blue), PfRBCs (red), and AscAg (green). By comparing the responses in the albendazole versus placebo group, the estimates of the treatment effect in the whole study population after 9 (A) and 21 (B) mo of albendazole treatment were obtained using linear mixed models, and positive values were plotted in a spider chart. Statistically significant estimates at 9 mo were IL-2 responses to AscAg (estimated effect of treatment [95% CI], 0.17 [0.05–0.28]), TNF (0.37 [0.21–0.53]), and IFN-γ (0.14 [0.03–0.24]) responses to PfRBCs and TNF (0.14 [0.05–0.24]), IFN-γ (0.10 [0.01–0.19]), and IL-2 (0.12 [0.01–0.23]) responses to PHA. At 21 mo posttreatment, PHA-induced IL-5 (0.10 [0.01–0.19]) and IL-10 (0.12 [0.05–0.19]) were significantly enhanced. As an indication of the magnitude of change in level of cytokines that were significantly different between the placebo and albendazole group, geometric mean and SE for TNF to PfRBCs at 9 mo (C) are given as an example.

Fig. S3.

Effect of deworming on cytokine responses to AscAg, PfRBCs, and PHA. Cytokine concentrations were assessed in supernatants of 72 h-stimulated whole-blood cultures. The estimated effects of albendazole treatment on the production of the significant cytokines over the whole trial period [IL-2 to AscAg (A), TNF (B) and IFN-γ (C) to PfRBCs, and TNF (D) and IL-10 (E) to PHA] are shown (Left). The estimates are calculated based on the differences in treatment versus placebo data but taking into account the clustered design and missing at random. The estimate, beta value is the unit change in cytokine as a result of treatment. (Right) The corresponding geometric means and SE of these cytokine levels [IL-2 to AscAg (F), TNF (G) and IFN-γ (H) to PfRBCs, TNF (I) and IL-10 (J) to PHA] in pg/mL. It should be noted that these geometric means were calculated only in subjects for whom we have data at those follow-up time points. Interpreting these means is valid only under missing completely at random.

Fig. 2.

Effect of deworming on cell subsets and marker expression. Flow cytometry was performed on PBMCs from a subset of school children. Gating strategy is shown for (A) lymphocytes and CD4⁺ T cells, from which (B) CD25^hiFOXP3⁺ Treg cell, (C) PD-1, and (D) CTLA-4 expressions on CD4⁺ T cells were derived. (E) CTLA-4 expression on CD25^hiFOXP3⁻ cells was gated from B. The estimated effect of albendazole treatment is shown for the time points 9 and 21 mo after the start of treatment for percentages of CD25^hiFOXP3⁺ (F), PD-1⁺ (G), and CTLA-4⁺ (H) of CD4⁺ T cells and CTLA-4⁺ of CD4⁺FOXP3^– cells (I). Estimates and β (beta) were obtained by linear mixed models; 95% CIs and overall P values over time (P_time) are indicated. As an indication of magnitude of change, the actual percentage of CTLA-4⁺ of CD4⁺FOXP3^– cells in placebo and albendazole groups is shown at 9 mo (J).