Guidance for Using Tafenoquine for Prevention and Antirelapse Therapy for Malaria — United States, 2019

Julia C. Haston, MD; Jimee Hwang, MD; Kathrine R. Tan, MD


Morbidity and Mortality Weekly Report. 2019;68(42):1062-1068. 

In This Article

Rationale and Evidence

A total of 269 articles were identified. After excluding 232 during title review and 29 during abstract review, the eight remaining articles were reviewed fully and included in the analysis: five articles related to prophylaxis[11–15] and three to antirelapse;[16–18] among these eight articles, seven included additional information for safety.[11–13,15–18] All five studies cited by the FDA label were captured. An additional four peer-reviewed articles addressing the in vivo activity of hypnozoites of P. ovale and use of primaquine for P. ovale were reviewed;[6,19–21] however, because they did not assess tafenoquine use, they were not included in the tafenoquine review.


Three of the five articles included were randomized controlled trials (RCTs), one was a reanalysis of data from an RCT, and one was a randomized human challenge study[11–15] (Table 1). Two RCTs compared tafenoquine (200 mg for 3 days, then weekly thereafter for up to 6 months) to placebo; both found a protective efficacy of 86% (95% confidence intervals [CIs] = 73–93 and 76–92).[11,12] Although not powered to detect statistical differences in efficacy, one of these RCTs described the efficacy between tafenoquine (protective efficacy = 86%, 95% CI = 76–92) and mefloquine (protective efficacy = 86%, 95% CI = 72–93).[12] The third RCT compared tafenoquine to mefloquine and observed no cases of malaria in either arm.[13] The data from this study were reanalyzed in a separate study, using an estimation of attack rate, and found protective efficacy to be 100% (95% CI = 93–100).[14] These studies suggest comparable efficacy between tafenoquine and mefloquine, the current standard of care. Finally, the randomized human challenge study demonstrated 100% efficacy (95% CI = 40–100) of tafenoquine against the blood stage of P. falciparum in healthy volunteers compared with placebo.[15]

Antirelapse Therapy

One phase 2b randomized dose-selection trial and two phase 3 RCTs examined the efficacy of tafenoquine in the prevention of relapse in patients with confirmed P. vivax malaria at the labeled recommended regimen (Table 1).[16–18] Among these studies, tafenoquine was found to prevent relapse in 62%–89% of cases with a single 300 mg dose. In the large phase 2 dose-response study, efficacy of 300 mg and 600 mg were similar, and significantly higher than that of chloroquine alone (300 mg dose: 89.2%, 95% CI = 77–95, log-rank test p-value <0.001; 600 mg dose: 91.9%, 95% CI = 80–97, p<0.001; chloroquine: 36.5%, 95% CI = 23–52).[16]

P. ovale Efficacy

Tafenoquine is not labeled for use in P. ovale. Because P. ovale is relatively rare, accounting for fewer than 5% of malaria cases globally,[19] it was not evaluated in the tafenoquine studies. Based on the biologic similarity of the hypnozoites of ovale and vivax, a CDC expert committee previously recommended the use of primaquine off-label for antirelapse therapy of P. ovale.[6] With similar in vivo response of P. ovale to primaquine to that of P. vivax,[20,21] CDC subject matter experts are extrapolating the use of tafenoquine to P. ovale.


Seven of the eight reviewed studies provided safety outcomes; four reported safety outcomes at the prophylaxis dose and three at the antirelapse therapy dose (Table 2).[11–13,15–18] Common adverse events included abdominal pain, constipation, diarrhea, vertigo, dizziness, sleep disturbances, and headache. Two studies described a nonsignificant increase in methemoglobin.[11,13] Another reported asymptomatic decreases in hemoglobin, which resolved without intervention.[18] One study described vortex keratopathy (a condition characterized by changes in the corneal epithelium resulting in a whorl pattern) in approximately 90% of patients receiving tafenoquine prophylaxis; the condition did not affect visual acuity and resolved within 1 year following drug discontinuation.[13] Of note, persons with G6PD deficiency were excluded because 8-aminoquinolines can cause hemolytic anemia in these persons.