Effect of adherence to primaquine on the risk of Plasmodium vivax recurrence: a WorldWide Antimalarial Resistance Network systematic review and individual patient data meta-analysis.
Mehdipour P., Rajasekhar M., Dini S., Zaloumis S., Abreha T., Adam I., Awab GR., Baird JK., Brasil LW., Chu CS., Cui L., Daher A., do Socorro M Gomes M., Gonzalez-Ceron L., Hwang J., Karunajeewa H., Lacerda MVG., Ladeia-Andrade S., Leslie T., Ley B., Lidia K., Llanos-Cuentas A., Longley RJ., Monteiro WM., Pereira DB., Rijal KR., Saravu K., Sutanto I., Taylor WRJ., Thanh PV., Thriemer K., Vieira JLF., White NJ., Zuluaga-Idarraga LM., Guerin PJ., Price RN., Simpson JA., Commons RJ., WWARN Vivax Adherence Study Group None.
BACKGROUND: Imperfect adherence is a major barrier to effective primaquine radical cure of Plasmodium vivax. This study investigated the effect of reduced adherence on the risk of P. vivax recurrence. METHODS: Efficacy studies of patients with uncomplicated P. vivax malaria, including a treatment arm with daily primaquine, published between January 1999 and March 2020 were identified. Individual patient data from eligible studies were pooled using standardized methodology. Adherence to primaquine was inferred from i) the percentage of supervised doses and ii) the total mg/kg dose received compared to the target total mg/kg dose per protocol. The effect of adherence to primaquine on the incidence of P. vivax recurrence between days 7 and 90 was investigated by Cox regression analysis. RESULTS: Of 82 eligible studies, 32 were available including 6917 patients from 18 countries. For adherence assessed by percentage of supervised primaquine, 2790 patients (40.3%) had poor adherence (≤ 50%) and 4127 (59.7%) had complete adherence. The risk of recurrence by day 90 was 14.0% [95% confidence interval: 12.1-16.1] in patients with poor adherence compared to 5.8% [5.0-6.7] following full adherence; p = 0.014. After controlling for age, sex, baseline parasitaemia, and total primaquine dose per protocol, the rate of the first recurrence was higher following poor adherence compared to patients with full adherence (adjusted hazard ratio (AHR) = 2.3 [1.8-2.9]). When adherence was quantified by total mg/kg dose received among 3706 patients, 347 (9.4%) had poor adherence, 88 (2.4%) had moderate adherence, and 3271 (88.2%) had complete adherence to treatment. The risks of recurrence by day 90 were 8.2% [4.3-15.2] in patients with poor adherence and 4.9% [4.1-5.8] in patients with full adherence; p