Randomized trial of primaquine hypnozoitocidal efficacy when administered with artemisinin-combined blood schizontocides for radical cure of Plasmodium vivax in Indonesia

Plasmodium vivax threatens nearly three billion people and sickens tens of millions annually [1]. Improper or delayed treatment may lead to severe and fatal outcomes associated with severe anemia, severe thrombocytopenia, respiratory distress, renal failure, hepatic dysfunction, coma, or shock [2, 3]. Unlike, Plasmodium falciparum, P. vivax places latent stages called hypnozoites in the liver after inoculation of mosquito-borne sporozoites. Hypnozoites later awaken and provoke renewed clinical attacks, each causing debilitating illness with risk of threatening complications and onward transmission.

In tropical Southeast Asia, P. vivax relapse behaviors resemble those in the Chesson strain from New Guinea [4, 5]. “Chesson-like” strains relapse quickly, repeatedly, and at approximately two-month intervals in almost all patients. In Thailand and Indonesia, relapse occurred among nearly 80 % of patients within two months [6, 7]. Among others treated for acute falciparum malaria, 51 % suffered P. vivax attacks within two months [6]. Relapse likely accounts for most attacks of vivax malaria where Chesson-like P. vivax occurs [8, 9]. Therapeutic success against hypnozoites may thus be appreciated as an important clinical and public health objective.

Radical cure of vivax malaria includes blood schizontocidal and hypnozoitocidal therapies. In 1952 the United States licensed the 8-aminoquinoline primaquine combined with the 4-aminoquinoline chloroquine for radical cure of P. vivax malaria. Primaquine remains the only option against relapse, although another 8-aminoquinoline, the investigational drug tafenoquine, approaches licensure [10]. Newer blood schizontocidal therapies must be combined with primaquine in radical cure because resistance to chloroquine by P. vivax often occurs, especially in Southeast Asia [11]. Resistant strains of P. vivax dominate across the Indonesian archipelago, and chloroquine was abandoned in favor of artemisinin-combined therapies (ACT) therapies over a decade ago [12].

Authoritative recommendations for the use of ACTs for treatment of acute P. vivax malaria cite good evidence of efficacy [13, 14]. However, no evidence yet attests to primaquine safety and efficacy when co-administered with most of those therapeutic options. The importance of such evidence has been highlighted by historic examples of drug-drug interactions (DDI) profoundly impacting both the safety and efficacy of 8-aminoquinolines [15]. Frontline therapy for acute P. vivax in Indonesia is the ACT dihydroartemisinin-piperaquine (DHA-PP) [12]. We reported good safety and efficacy of primaquine against relapse when administered 25 days after DHA-PP therapy, a delay imposed by the unexamined possibility of DDI [7]. Later studies affirmed no DDI issues with co-administered DHA-PP and primaquine [16]. The current study aimed to evaluate the safety and efficacy of primaquine administered concurrently with DHA-PP. We also evaluated primaquine co-administered with another ACT, artesunate-pyronaridine (AS-PYR) given recent evidence indicating a lack of significant pharmacokinetic interaction with primaquine [17]. Both of these blood schizontocidal therapies have very good efficacy against asexual blood stages of P. vivax [18–21].

The trial included a third treatment arm, artesunate [22] followed by primaquine. That arm was not intended as therapy for practice, but was an independent experimental arm testing the hypothesis that primaquine may require concurrent administration of partner blood schizontocides for good efficacy [15]. A trial in the 1950s demonstrated primaquine failure against relapse when administered after quinine rather than concurrently [23].

We recruited Indonesian soldiers exposed to malaria in eastern Indonesia after returning to a site in Central Java free of malaria transmission [24]. In addition to the distinct advantage of classifying recurrent malaria as relapses with a high degree of certainty, the infections were by the chloroquine-resistant and primaquine-tolerant Chesson-like P. vivax strains dominating eastern Indonesia [25].

This randomized trial of 0.5 mg/kg PQ daily for 14 days concurrent with AS-PYR or DHA-PP for radical cure of vivax malaria demonstrated good safety, tolerability, and efficacy for these options. All subjects cleared parasitemia by day 3 and none recurred before day 39, affirming good efficacy of AS-PYR and DHA-PP against the asexual blood stage parasites [18–21]. Initially high rates of gametocytemia rapidly cleared (<72 hr). Relapse occurred among 9/60 (15 %) and 7/60 (12 %) subjects assigned AS-PYR or DHA-PP, respectively. Incidence density was applied as the preferred mathematical treatment to derive estimates of therapeutic efficacy involving prolonged periods of follow-up [30–32]. We estimated the efficacy of radical cure by AS-PYR + PQ and DHA-PP + PQ at 94 % and 95 %, respectively (Table 3), each with good tolerability.

Drug-related changes occurred among subjects. The significant elevation of AST/ALT in a minority of subjects receiving AS-PYR (5/60) has been reported in other studies [33]. QTc elongation appeared in subjects taking AS-PYR and DHA-PP, but not AS + PQ (Fig. 4). These effects are known with those compounds and were not exacerbated by primaquine (daily 0.24 mg/kg dosing) in DDI studies in healthy volunteers [16, 17]. The data in the current trial are consistent with those observations. Consumption of a carbohydrate snack immediately prior to each primaquine dose very likely explains the relatively very low rates of vomiting despite high dose primaquine [27].

The AS + PQ treatment in the current trial tested the hypothesis that PQ requires concurrent blood schizontocidal therapy for good efficacy against hypnozoites. In 1955, Alving et al. [23] demonstrated inferior efficacy of PQ administered after quinine therapy (15/19 relapsed) relative to concurrent administration of same therapies at the same doses (2/19 relapsed) among those randomized groups. We administered AS first, then commenced PQ after a two day washout; 10/60 subjects relapsed, but no more frequently than among the other two treatment arms (P > 0.5). A higher total dose of PQ in the current trial (7 mg/kg) compared to Alving et al. [23] (3.5 mg/kg) likely explains the good efficacy of PQ despite otherwise similar treatment (artesunate and quinine are each efficacious against asexual blood stages and eliminated within 24 hr of dosing). PQ at the higher dose may thus be administered after treating the acute attack with assurance of good efficacy. This offers a therapeutic option when concurrent administration is either inconvenient or contraindicated. Primaquine at the 7 mg/kg total dose may thus commence at any point after therapy of acute malaria, or presumptively without an acute attack (as with post-travel presumptive anti-relapse therapy) with assurance of good efficacy.

PQ is prescribed at a total dose of either 3.5 or 7.0 mg/kg over 14 days, depending upon geographic origin of infection [34]. The New Guinea Chesson strain responded to daily 7.0 mg/kg, but not 3.5 mg/kg [35]. Studies confirm the tolerance of P. vivax strains from tropical Southeast Asia and New Guinea to the lower dose [25, 35, 36]. Despite applying the higher dose under direct supervision, 14 % (26/180) of our subjects relapsed. Possible explanations include hypnozoite resistance to PQ [37], and cytochrome P-450 2D6 (CYP2D6) polymorphisms impeding conversion of PQ to active metabolites [38, 39]. We are now examining CYP2D6 polymorphisms among the therapeutic failures in this trial and anticipate reporting these findings in the near future. Uncertainty regarding the basis of relapse despite high dose PQ therapy is a limitation of this study.

In addition to the primary incidence density endpoint analysis, Table 3 also shows crude estimates of efficacy as expressed by proportion of subjects remaining free of recurrent parasitemia after six months. This endpoint has been used in trials of hypnozoitocides and meta-analyses of them. Those trials were conducted where reinfection among treatment groups could not be excluded, or where variable risks and timing of natural relapse was not controlled in trial design. In other words, rather than report efficacy against relapse, those trials simply compare the effect of hypnozoitocidal therapy upon rates of recurrence from any source of bloodstream infection. In the trial reported here, by virtue of near certainty regarding the absence of confounding by recrudescence (wholly effective blood schizontocides) or reinfection (absence of risk at the study site), along with a relapse control arm (risk and rate of relapse without primaquine therapy), we were able to report therapeutic efficacy of primaquine against relapse using incidence density risk analysis.

This trial provided robust estimates of PQ efficacy against relapse when combined with modern therapies. The findings of good safety, tolerability and efficacy in this trial reassures, but does not address the serious problem of PQ effectiveness. PQ often fails because it is: 1) unavailable or prohibited [40]; 2) rarely prescribed [41]; 3) withheld from patients who may be pregnant, <1 yr of age, or deficient in G6PD [34]; 4) not offered to patients of unknown G6PD status [42]; 5) not completed by the patient [43]; or 6) disabled by CYP2D6 polymorphisms [37, 38, 44]. The sum of these potential pitfalls deeply erodes PQ effectiveness in routine practice in the often-impoverished endemic areas of the tropics. Good efficacy is the starting point for then addressing a range of barriers to the good effectiveness that realizes maximum clinical and public health impacts.

Patients with P. vivax receiving AS-PYR + PQ or DHA-PP + PQ enjoyed good tolerability, safety and efficacy of these regimens. AS-PYR and DHA-PP may thus be considered validated options to chloroquine for radical cure of P. vivax with PQ. Further, PQ at 7 mg/kg total dose does not require concurrent administration of blood schizontocide for good efficacy. PQ may thus follow therapy with untried blood schizontocides for either therapy or chemoprophylaxis with the expectation of good efficacy.