Sulfadoxine-Pyrimethamine–Based Combinations for Malaria: A Randomised Blinded Trial to Compare Efficacy, Safety and Selection of Resistance in Malawi

Abstract

BackgroundIn Malawi, there has been a return of Plasmodium falciparum sensitivity to chloroquine (CQ) since sulfadoxine-pyrimethamine (SP) replaced CQ as first line treatment for uncomplicated malaria. When used for prophylaxis, Amodiaquine (AQ) was associated with agranulocytosis but is considered safe for treatment and is increasingly being used in Africa. Here we compare the efficacy, safety and selection of resistance using SP or CQ+SP or artesunate (ART)+SP or AQ+SP for the treatment of uncomplicated falciparum malaria.Methodology and Findings455 children aged 1–5 years were recruited into a double-blinded randomised trial comparing SP to the three combination therapies. Using intention to treat analysis with missing outcomes treated as successes, and without adjustment to distinguish recrudescence from new infections, the day 28 adequate clinical and parasitological response (ACPR) rate for SP was 25%, inferior to each of the three combination therapies (p<0.001). AQ+SP had an ACPR rate of 97%, higher than CQ+SP (81%) and ART+SP (70%), p<0.001. Nineteen children developed a neutropenia of ≤0.5×103 cells/µl by day 14, more commonly after AQ+SP (p = 0.03). The mutation pfcrt 76T, associated with CQ resistance, was detected in none of the pre-treatment or post-treatment parasites. The prevalence of the pfmdr1 86Y mutation was higher after treatment with AQ+SP than after SP, p = 0.002.ConclusionsThe combination AQ+SP was highly efficacious, despite the low efficacy of SP alone; however, we found evidence that AQ may exert selective pressure for resistance associated mutations many weeks after treatment. This study confirms the return of CQ sensitivity in Malawi and importantly, shows no evidence of the re-emergence of pfcrt 76T after treatment with CQ or AQ. Given the safety record of AQ when used as a prophylaxis, our observations of marked falls in neutrophil counts in the AQ+SP group requires further scrutiny.Trial RegistrationControlled-Trials.com ISRCTN22075368

Adjusted Risk Difference Estimation: An Assessment of Convergence Problems with Application to Malaria Efficacy Studies

“…These were used to compare the efficacy rates in two treatment groups ("active" and control) with adjustment for up to two covariates. Two patterns of correlation between the covariates were examined; no/negligible correlations and correlations similar to those found by Bell et al [2]. All simulations used a sample size 200 (100 participants randomized to each group), again as used by Bell et al [2].…”

“…Two patterns of correlation between the covariates were examined; no/negligible correlations and correlations similar to those found by Bell et al [2]. All simulations used a sample size 200 (100 participants randomized to each group), again as used by Bell et al [2]. Five thousand (5,000) datasets were simulated for each scenario considered.…”

“…Datasets were simulated based on the methods and results of the randomized control trial described in the previous section Bell et al [2]. These were used to compare the efficacy rates in two treatment groups ("active" and control) with adjustment for up to two covariates.…”

“…Risk differences are becoming more widely reported in RCTs, especially in malaria efficacy studies [2][3][4]. These are now frequently used when the objective is to compare treatment effect between treatments as commonly used alternatives such as odds ratios and risk ratios can be difficult for general medical researchers to interpret [5].…”

“…This research was motivated by a malaria efficacy study that was conducted in Malawi between 2003 and 2005; full details of the methods and findings of this study are available in Bell DJ et al [2] but to summarise briefly:…”