Primaquine for Prophylaxis of Malaria: Has the CYP Sailed?
Primaquine (PQ) is widely used for the radical cure of Plasmodium vivax malaria in combination with chloroquine (CQ) and in some countries (mostly the United States) for presumptive antirelapse therapy (PART, or terminal prophylaxis) to prevent relapse caused by P. vivax or Plasmodium ovale. Although use of PQ for primary prophylaxis of malaria is not an approved indication by the US Food and Drug Administration, the US Centers for Dlisease Control and Prevention (CDC) has recommended PQ at a dose of 30 mg daily for use as primary malaria chemoprophylaxis in areas where the risk of malaria is predominantly due to P. vivax. PQ is an appealing option for primary prophylaxis use for several reasons. In addition to being generally well-tolerated, due to its activity in the pre-erythrocytic liver stage, it offers the option of a shortened post-exposure dosing regimen of 7 days. Furthermore, when used as primary prophylaxis, PQ protects against relapse of P. vivax or P. ovale, obviating the need for PART.
The efficacy of PQ as causal prophylaxis was established in trials conducted in several populations, including Indonesian migrants, Kenyan children, and Colombian soldiers. It should be noted that there have been no controlled trials to study the efficacy of this regimen in western travelers. In the Colombian soldiers, the population most similar to European nonimmune populations, PQ was found to have an efficacy of 89% [95% confidence interval (CI), 75%–96%] at preventing all malaria, although this was higher (94%, CI 78%–99%) at preventing malaria due to Plasmodium falciparum. In an observational study comparing PQ to mefloquine and doxycycline in Israeli travelers to Ethiopia, direct measurement of prophylactic efficacy was not possible due to lack of a placebo control, but based on infection rates observed in the other arms, it is likely that efficacy exceeded 90%.
Despite these appealing characteristics, PQ also has a number of drawbacks that have led to limitations in its use. First, although the risk of hemolysis associated with glucose-6-phosphate dehydrogenase (G6PD) deficiency can be mitigated with screening, this adds to the cost and inconvenience of this regimen. Although some travelers find the requirement for daily dosing to be inconvenient, developing hepatic schizonts appears to display a time-dependent window of susceptibility to PQ, making this regimen potentially less forgiving to lapses in adherence than some others might be. Now, in addition to these there is another cause for concern with the use of PQ for primary prophylaxis.
What Has Changed?
Recently, Bennett and colleagues have reported failure of PQ radical cure in two individuals with decreased activity of the hepatic isoenzyme cytochrome P450 (CYP) 2D6 in the setting of a malaria challenge study. In that trial, 2 of 33 individuals infected with P. vivax experienced relapse, one 3 times, the other twice, despite adequate, supervised, weight-based treatment with PQ in conjunction with CQ. Investigation into the causes of PQ failure determined that both had decreased CYP 2D6 activity. To further explore these results, Pybus and colleagues recently showed that mice in which the CYP 2D family has been knocked out are not protected from Plasmodium berghei infection, even by doses of PQ twofold higher than are typically protective in that model. PQ activity was restored in transgenic mice that had the functional human 2D6 enzyme inserted in place of the mouse 2D enzyme family. Taken together, these results demonstrate that PQ efficacy in the radical cure of vivax malaria and in causal prophylaxis of rodent P. berghei is dependent on the generation of an active metabolite through the activity of the CYP 2D6 enzyme.
Cytochrome P450 2D6 is an isoenzyme that is involved in the metabolism of as many as 25% of drugs in clinical use. It is also one of the members of the cytochrome P450 enzyme family with the greatest prevalence and diversity of genetic polymorphism. Over 40 allelic variants have been described with enzymatic activity ranging from fully functional to nonfunctional. The frequency of these variants differs greatly between ethnic groups. Among populations with a European heritage, the frequency of poor metabolizer (PM) phenotype that has two nonfunctioning alleles is 5% to 10%. Although the frequency of nonfunctional alleles is lower in other ethnic groups, the frequency of alleles with reduced function is as high as 50% in most Asian populations.
Implications for Clinical Use
The finding that CYP 2D6 activity is required for PQ efficacy in radical cure of vivax malaria has troubling implications for its use as primary prophylaxis. For as many as 5% to 10% of Caucasian travelers, PQ would be expected to provide no benefit at all. It is uncertain what impact intermediate metabolizer phenotypes might have on efficacy. It is important to note that one of the subjects in the report from Bennett and colleagues fell into this category. This is worrisome given the high frequency of this phenotype in many populations. Because the decision to prescribe malaria chemoprophylaxis reflects an assessment of sufficient risk of travel-associated malaria for a given traveler or itinerary, this risk of failure is higher than what may be acceptable to many travelers or travel medicine specialists. It is intriguing to note that this rate of failure is similar to what has been observed in clinical trials of PQ primary prophylaxis.[6, 7, 12] In coming years it will be interesting to note whether observational reports of PQ failure for radical cure are also attributable to CYP 2D6 polymorphism. Although it might be tempting to look to pharmacogenetic factors for failure of PQ in multiple settings, this is likely only one of many factors, as supported by differing rates of relapse dependent on the location of acquisition of infection. There are no data on the efficacy of PQ in individuals with deficient CYP 2D6 activity for prophylaxis of falciparum malaria, although caution seems justified.
These results have implications for the use of PQ for multiple indications. While one might expect that failures of PQ due to ineffective metabolic activation might occur in both its use for radical cure and primary prophylaxis, we are principally concerned with failures of primary prophylaxis for a number of reasons. It is true that failures of radical cure in endemic populations present a greater public health burden. While exploration of possible causes of failure of radical cure in these populations is of interest for generating hypotheses, because there are no other classes of antimalarial drugs available for radical cure, there are no therapeutic options available other than PQ. This stands in contrast to the indication of primary prophylaxis, for which there are other safe and effective alternatives. Furthermore, failures of primary prophylaxis can pose a particularly significant risk because they affect a population that frequently has little to no preexisting immunity and may have little access to prompt medical care.
In light of these findings, the role of PQ in primary malaria chemoprophylaxis in P. vivax predominant areas should be reassessed. In addition, although PQ is not recommended for primary prophylaxis in areas with a significant risk of P. falciparum malaria, its use for this indication in travel medicine should also be discouraged. These considerations are unlikely to affect alternative prophylactic medications, which all, unlike PQ, have potent in vitro activity in the absence of active CYP-dependent metabolism. It is our opinion that PQ should not be considered as an option for travelers of European ancestry with travel itineraries deemed to pose sufficient risk to justify the use of chemoprophylaxis. If it is used, back-up treatment and education to seek medical care are essential, just as they are for all at-risk travelers. In the future, genotyping and “personalized medicine” may eliminate this risk, although at present CYP 2D6 testing is neither affordable nor widely available. Lastly, these findings have implications for the development of other 8-aminoquinolne antimalarials for chemoprophylaxis, posing challenges to any other drugs that may also require metabolic activation.
Opinions expressed here are solely those of the authors and do not reflect the opinion of the National Institutes of Health, the US Government, or the Bill & Melinda Gates Foundation.
Declaration of Interests
The authors state they have no conflicts of interest to declare.