Primaquine Review

The ability of Plasmodium vivax (P.vivax) to lie dormant for long periods of time has important implications for malaria control programs and the elimination agenda. The only anti-malarial drug currently widely available with hypnozoitocidal activity is primaquine.

Despite over 60 years of use, primaquine’s clinical efficacy is poorly understood. Current dosing regimens vary with region according to the prevailing epidemiology (Footnote 1,2) however, this is often based on weak evidence with a general absence of data regarding the risks and benefits of different dosing schedules.

The number and frequency of P. vivax relapses varies according to a number of factors. This heterogeneous epidemiology confounds interpretation of antirelapse clinical efficacy studies making comparison of different treatment regimens extremely difficult.

This systematic review of the literature highlights the diversity of methodologies that have been applied to antirelapse clinical trials and summarises the evidence of primaquine efficacy for different regimens across a range of epidemiological and geographical locations.

• There is marked heterogeneity in the design of studies assessing relapse prevention (radical cure) by primaquine in P. vivax malaria.

• The risk of recurrent malaria varies considerably according to primaquine dosing, partner  drug, duration of follow up and the geographic location of the study.

• As relapse may occur many weeks after the initial infection, if a slowly eliminated anti-malarial drug is given in treatment, studies of relapse prevention should have long follow up (>8 weeks).

• Recurrence rates following treatment with very low dose primaquine (total dose <2.5mg/kg) show no significant benefit compared to control patients receiving no antirelapse medication.

• Low dose primaquine regimens (total dose >2.5mg/kg and <5.0 mg/kg) resulted in an adequate response (<10% recurrent infections) in 32% (26/82) of studies and were only better than a control arm in 50% (6/12) of studies.

• High dose primaquine regimens (total dose >5.0 mg/kg) have high efficacy (OR=0.03 [95%CI: 0.01-0.13]) compared with patients not receiving primaquine. Few high dose Primaquine studies followed patients for more than 2 months.

• High dose regimens administered intermittently over a prolonged period of time (more than 28 days) have significantly lower effectiveness than regimens administered over shorter time courses.

A Pubmed search was conducted for all relevant primaquine clinical trials and data extracted systematically. The total dose of primaquine was defined as very low (≤2.5 mg/kg), low (>2.5mg/kg-<5.0 mg/kg) and high dose (≥ 5.0 mg/kg). Details of the review methodology are presented in the review profile file.

Download the full library of reference and extracted data. A full analysis is presented in John GK, Douglas NM, von Seidlein L et al. Primaquine radical cure of Plasmodium vivax: a critical review of the literature. Malaria Journal 2012; 11:280 doi: 10.1186/1475-2875-11-280.

A total of 87 clinical trials were analysed, of which 43 (49%) were conducted on or after 2000. Map 1 displays the study sites of all antirelapse studies in the review. Maps 2-4 display the studies split by dosing regimen. Dosing regimens varied greatly, with very low total dose used in 44 (28%) of 156 treatment arms, low dose in 82 (53%) and high dose in 28 (18%) regimens.

Map 1. Study sites of primaquine antirelapse studies; One study (yellow), two studies (orange), three or more (red). Click on map to enlarge.

Map 1 displays the study sites of all antirelapse studies in the review. Maps 2-4 display the studies split by dosing regimen. Dosing regimens varied greatly, with very low total dose used in 44 (28%) of 156 treatment arms, low dose in 82 (53%) and high dose in 28 (18%) regimens.

Scatter plots 1-4 highlight the wide range of primaquine efficacy presented by published trials. Both very low dose and low dose primaquine were vulnerable to high rates of recurrence of P. vivax. In contrastthe use of high dose primaquine was associated more often with acceptable efficacy, although most studies were of short duration (Footnote 2)

Figure 1. Scatter plot shows the risk of recurrence at the end of the study following low dose primaquine.

 A total of 18 (21%) studies included patients recruited into control arms in which no primaquine was prescribed, enabling direct comparison of efficacy controlling for background epidemiology. The forest plots demonstrate the efficacy of the three main dose categories of primaquine. Low dose primaquine regimens were significantly better than control regimens in 50% (6/12) of studies assessed (OR=0.14 [95% CI 0.06-0.35], p<0.001). Two studies demonstrate the high efficacy of high dose primaquine when compared with a control arm (OR=0.03 [95%CI: 0.01-0.13]; p<0.0001).  

Figure 2. Forest plot of the efficacy of low dose primaquine in studies with a control arm. Click on plot to enlarge.

Forest plots 1-4 show all three dosing regimens and the effectiveness of directly observed treatment in primaquine therapy.

The investigation of different primaquine regimens is crucial if we are to develop robust primaquine regimens with known efficacy that can be applied safely in malaria endemic settings. The lack of standardised methodologies and interpretation of antirelapse clinical trials is a major impediment to the optimal use of primaquine and novel hypnozontocidal agents under development. WWARN proposes a set of key components of antirelapse clinical trials along with a series of research priorities, to stimulate further debate on the subject.

View the abstract: John GK, Douglas NM, von Seidlein Let al. Primaquine radical cure of Plasmodium vivax: a critical review of the literature. Malaria Journal 2012; 11:280 doi:10.1186/1475-2875-11-280.

¹Guidelines for the treatment of malaria. Geneva: World Health Organization; 2010. http://www.who.int/malaria/publications/atoz/9789241547925/en/index.html

²Hill DR et al. Primaquine: report from CDC expert meeting on malaria chemoprophylaxis I. PubMed 2006 September; 75(3):402-15.