Severe Vivax Malaria: Newly Recognised or Rediscovered?

Stephen Rogerson and Richard Carter discuss two new studies that challenge current dogma by suggesting that vivax malaria can cause severe disease.

M alignant tertian" and "benign tertian" are terms that have long been used for two of the major diseases we recognise as malaria. The former is generally considered to be synonymous with Plasmodium falciparum and the latter with P. vivax infection. As the names "malignant" and "benign" suggest, the current dogma is that P. falciparum can be severe and life-threatening while P. vivax tends to be mild. However, two studies published in this issue of PLoS Medicine challenge this dogma.

The Two New Studies
Ric Price and colleagues, in a prospective study based in Timika, southern Papua in Indonesia [1], and Blaise Genton and colleagues, in a prospective study in the Wosera region of Papua New Guinea [2], report similar rates and outcomes of severe malaria due to P. vivax or P. falciparum. The two studies had different settings, and the cultural and ethnic characteristics of the patient populations were also different. Price and colleagues collected data from all patients attending the outpatient and inpatient departments of the only hospital in the region, using systematic data forms and computerised hospital records. In contrast, Genton and colleagues investigated patients presenting at two rural health facilities.
In both settings, clinical and severe disease was most common in young children, with P. vivax cases peaking at an earlier age than those of P. falciparum. In Timika, in children under five years old, around 30% of cases of either P. falciparum or P. vivax were classed as severe, and around 80% of such cases were accompanied by severe anaemia (haemoglobin less than 5 g/dl). The remainder presented with respiratory distress, impaired consciousness (i.e., cerebral malaria), or overlapping syndromes. In the Wosera, in children under five years attending the health centre, around 9% of vivax and 12% of falciparum infections were classified as severe malaria. However, respiratory distress was defined less stringently in the Wosera (more than 40 breaths/minute from two to 60 months) than in Timika (more than 50 breaths/minute in this age group). When cough and diarrhoea (which are poorly specific for malaria) were used as exclusions in the Wosera study, severe malaria rates in children under five years fell to 7% and 4% of P. falciparum and P. vivax infections, respectively. In the Wosera, anaemia occurred in about 20% and 40% of severe cases of vivax and falciparum, respectively, while respiratory distress occurred in 60% and 40% of severe cases of vivax and falciparum, respectively. Neurologic symptoms were present in 25% of severe cases of either species.
The cases of cerebral malaria due to P. vivax reported in both studies, occurring in all age groups, are intriguing, because this malaria complication has rarely been reported previously in association with P. vivax infection. In P. falciparum, cerebral malaria is primarily attributed to sequestration of infected erythrocytes in cerebral vessels. As P. vivax does not sequester, coma must arise by other means-perhaps of systemic metabolic origin-in vivax malaria. This aetiology may also underlie some cerebral malaria presentations in children.
Both studies, inevitably, have limitations. First, co-morbidities, including concomitant bacterial or viral infections, which could have decreased the malaria-attributable fraction of disease [3], were not actively investigated. Second, microscopy was used for parasite detection and speciation, which routinely leads to Ric Price and colleagues present data from southern Papua, Indonesia, suggesting that malaria resulting from infection with Plasmodium vivax is associated with substantial morbidity and mortality.
Research in Translation discusses health interventions in the context of translation from basic to clinical research, or from clinical evidence to practice. marked underestimation of mixed infections in particular [4]. Some "severe vivax" cases may actually have been mixed infections.

Severe Vivax Malaria
Despite these limitations, a striking feature of the two studies is the overall comparable incidence of severe disease in P. vivax and P. falciparum infections in each setting. There were differences in the prevalence of the components of severe disease in the two locations and a notable disparity in the overall rates of severe disease. Thus, in Timika, anaemia defined severe malaria more often than in Wosera, where anaemia was especially infrequent in severe vivax. This difference may be attributable to the fact that in Timika, both P. vivax and P. falciparum were commonly resistant to chloroquine, the first-line malaria treatment (with sulfadoxine-pyrimethamine in the case of P. falciparum), for much of the study period. Because persistent blood infections with malaria are likely to increase anaemia, drug treatment failure could have contributed to the high anaemia rates in Timika. In Wosera, where chloroquine resistance is common in P. falciparum but rare in P. vivax, severe anaemia defined 40% of severe falciparum cases, but only 20% of severe vivax cases. Thus the Wosera data are also consistent with parasite drug resistance as a factor in anaemia. If these conjectures are correct, the recent adoption of dihydroartemisinin-piperaquine treatment in Papua [5] may lead to a significant decrease in severe malarial anaemia.
Other factors, such as host genetics, could also have influenced the differences in severe disease manifestation between study sites. It has been shown, for example, that children in Papua New Guinea with the red cell genetic polymorphism "South East Asian ovalocytosis", a trait common in this population, are almost completely protected from cerebral malaria [6]. In Timika, indigenous Papuans were much more likely to be severely anaemic than were immigrants. Might host genetics have contributed to the high anaemia rates reported in the Timika study?

Interactions between Vivax and Falciparum
There is a view that malarial disease severity in New Guinea may be lower overall than in Africa. For example, case fatality rates for P. falciparum among hospitalised children in New Guinea ranged from 1.6% to 3.5% [7,8] versus 2% to 9% in recent African studies [9][10][11]. This difference raises the question of whether P. vivax, which is common in New Guinea but rare in Africa, may actually ameliorate the severity of P. falciparum where the two species are co-endemic.
Maitland and colleagues have suggested that the earlier age peak for P. vivax disease may protect against the later-acquired P. falciparum through species-transcending immunity [12]. Studies have shown that symptoms of P. vivax or P. falciparum infection are, indeed, significantly reduced by recent previous P. vivax infection (but not P. falciparum infection) [13]. In contrast, the two new studies in PLoS Medicine report that mixed species infections had worse outcomes than infections of either species alone [1,2]. If previous vivax infection does protect against severe P. falciparum infection, further investigations are needed so that policies for P. vivax control can balance reduction in severe vivax with possible loss of vivax-associated protection against severe P. falciparum infection.

Malaria Severity Depends on Context
The findings of the two new studies [1,2] should be considered from a wider biological and historical perspective, which shows that malarial disease severity is highly context dependent. For example, from the 16th through the 19th centuries in England, mortality rates associated with "marsh agues"-probably due largely to vivax malaria-gradually fell from high to negligible levels. This fall occurred either because of attenuation of P. vivax infection through adaptation of parasite and/or host, or through environmental change, such as reduced exposure to other diseases or improved nutrition [14]. Indeed, host undernutrition and co-infections, including HIV, helminth, and acute bacterial or viral infections, can all influence the outcomes in malaria [3]. Access

Box 1. Crucial Tools for P. vivax Control
Artemisinin-based combination therapies are generally effective, but longer-acting combinations may additionally prevent early relapses.
tests. Present tests are inadequately sensitive at low parasitaemia, which is common in P. vivax.
The role of intermittent preventive treatment in pregnancy or childhood against vivax is unknown. Small studies suggest that insecticide-treated bed nets decrease disease.
stages. G6PD deficiency, selected for by malaria, can result in fatal haemolysis from current agents.
New vaccines are entering human trials. If efficacious, their deployment will require major planning efforts.

Grimberg et al., 2007 [19]
The investigators show that antibodies directed against "region two" of PvDBP inhibit merozoite invasion, suggesting that vaccines targeted towards this region could reduce blood stage infection.

Barcus et al., 2007 [7]
Along with the two new papers in this issue of PLoS Medicine [1,2], Barcus and colleagues' study contributes to our understanding of the importance of P. vivax as a cause of severe clinical disease. This paper represents a significant advance in the difficult process of developing robust in vitro culture systems for P. vivax.

Ratcliff et al., 2007 [5]
Both treatments cured falciparum and vivax malaria with high efficacy, but dihydroartemisinin-piperaquine was much better than artemetherlumefantrine at preventing re-infections with P. falciparum and relapses of P. vivax during a six-week follow-up.

Singh et al., 2006 [18]
This paper shows how the Duffy binding protein interacts with its ligand, which is also relevant to understanding structure and interactions of other important adhesive proteins of malaria parasites.
to antimalarial drugs and presence of parasite drug resistance are clearly also relevant to the clinical outcome.
P. vivax infection is characterised by relapse in which dormant liver stages awaken, initiating new bouts of blood infection. Repeated malarial blood infections, from whatever source, have debilitating consequences including cachexia, spontaneous abortions, male infertility, developmental arrest, and impaired mental function [15]. All of these debilitating clinical features have been present in populations where P. vivax is known, or presumed, to have been the main prevalent malaria. However, these manifestations are also context dependent-they are influenced by circumstances such as the level of endemicity, existence of co-infections, access to treatment, and presence of parasite resistance.
It is interesting that the present reports of severe vivax malaria come from New Guinea, where drugresistant P. vivax was first reported. This drug resistance itself could be a driver of severity together with the particularly high regional malaria transmission force. Were severe vivax to be found more widely, it would have significant implications for control of the infection, especially as P. vivax invariably increases relative to P. falciparum under effective transmission reduction [16].

The Way Forward for Vivax Malaria Control
With calls for increased efforts to control malaria internationally, it will be important to ensure that P. vivax receives appropriate attention. We still lack reliable estimates of its global burden, and are only now starting to appreciate certain aspects of disease presentation of P. vivax malarial infection. The burden and severity of vivax in different settings requires further study. With the availability of the P. vivax genome (GenBank accession no. AAKM00000000), it will be possible to compare isolates associated with severe or uncomplicated disease. However, connecting genetic differences with pathogenic potential may not be straightforward [17].
Better disease control (see Box 1) might include preventive strategies (vaccines, insecticide-treated bed nets, and intermittent preventive treatment), more effective curative treatment (finding the most suitable alternative to chloroquine), and better relapse prevention (long-acting drugs such as dihydroartemisinin-piperaquine decrease early relapse compared to other artemisinin-based combination therapies [5]). Elimination of dormant parasites in the liver (hypnozoites) is especially difficult because available drugs can cause severe haemolysis in individuals with glucose-6-phosphate dehydrogenase (G6PD) deficiency. A simple point-of-care assay for G6PD deficiency, or new, safer drugs that target liver stage parasites, would be invaluable tools for eradication of hypnozoites.
Vaccines against P. vivax are in development. One target is the P. vivax Duffy binding protein (PvDBP), which mediates merozoite invasion of erythrocytes [18,19]; human vaccine studies using this target are likely to begin this year. Another target is the P. vivax gamete surface protein Pvs25, a candidate for a transmissionblocking vaccine. Indeed, this vaccine candidate is already in Phase I human trials [20]. Should either candidate progress to large-scale trials, it will be critical to examine impact on both P. vivax infection (including, potentially, increased chronicity of infection and anaemia following PvDBP vaccination) and P. falciparum infection. While New Guinea is an ideal site in which to evaluate such vaccines, evaluation must also be done in low-transmission areas where most P. vivax infections are symptomatic, and where diminished symptomatology could dangerously postpone treatment. The two reports by Price et al. and Genton et al. provide information about disease burden critical to improved decision making for the public health management of P. vivax malaria [1,2].