Along with many other aspects of P. vivax
infection, the pathophysiological aspects of vivax malaria have been largely neglected. The present study of the pulmonary physiological evidence demonstrates a reduction in the baseline VC
component of gas transfer in vivax malaria comparable to that found in uncomplicated falciparum malaria and a significant and prolonged impairment of alveolar-capillary gas transfer developing after commencement of treatment. This reduction was more apparent than that occurring in uncomplicated falciparum malaria but was less apparent than that found in severe falciparum malaria [34
]. These findings occurred despite a lower initial parasite burden in vivax malaria.
The reduction in VC volume indicates a reduction in the volume of hemoglobin-containing red cells within the pulmonary capillaries at the time of presentation. The most likely explanation is the encroachment on the hemoglobin-containing vascular compartment by sequestered leukocytes, parasitized red cells, or both. Although sequestration of leukocytes within the pulmonary microvasculature may well contribute to the subsequent impairment of DM function occurring in response to antimalarial treatment, it is unlikely to be the predominant cause of the impairment of baseline VC volume in either vivax or uncomplicated falciparum malaria because adherent leukocytes should also cause a parallel decrease in DM function, but this was not found with either Plasmodium species on presentation. In contrast, cytoadherence of red cells to the pulmonary vascular endothelium in which intraerythrocytic hemoglobin has been replaced by mature stages of P. vivax would be expected to reduce VC volume without a significant parallel decrease in DM function.
We, therefore, believe that sequestration of P. vivax
–infected red cells may contribute to reduced pretreatment VC
volume. In the absence of sequestration, peripheral parasitemia would be expected to be proportional to organ-specific dysfunction. The absence of any relationship between the level of P. vivax
parasitemia and either gas transfer or VC
volume supports the hypothesis that, as in the case of P. falciparum,
mature stages of P. vivax
–infected red cells sequester within the pulmonary microvasculature. Although all stages of P. vivax
circulate in vivax malaria, their distribution is not uniform. Mature schizonts in blood films are nearly always less common than the trophozoites from which they arise [31
]. Because of this, Field and Shute argued >50 years ago that in P. vivax
infections “an appreciable proportion ... complete their segmentation in the internal capillaries” [31
]. Animal models of malaria also support a role for nonfalciparum Plasmodium
species sequestering in the lung. Red cells parasitized with schizonts of P. vinckei
and P. yoelii
preferentially sequester within the pulmonary microvasculature [40
]. Human autopsy studies in vivax malaria are rare, and the pulmonary findings are not definitive. Among 3 early 20th-century autopsy reports of vivax malaria cases (none of which showed lung injury before death or completely excluded coinfection with falciparum malaria), one described parasites in alveolar capillaries [9
], one reported their absence in the pulmonary microvasculature (after quinine treatment) [7
], and one did not comment on their presence or absence [8
Although red cells infected with trophozoites of P. vivax
do not appear to cytoadhere to the endothelial cell ligands, CD36 or intracellular adhesion molecule–1 [41
], recent in vitro data have demonstrated cytoadherence of P. vivax
–infected red cells to CSA [4
], an endothelial cell ligand known to mediate sequestration of P. falciparum
within the placental microvasculature. The other human endothelial cells known to express CSA are those in the lung and brain [42
]. We speculate that cytoadherence of P. vivax
–infected red cells to CSA or another ligand on the pulmonary endothelium may be a mechanism for any sequestration within the pulmonary microvasculature.
Other mechanisms of microvascular obstruction known to occur in falciparum malaria are unlikely to explain the reduction in VC
volume. In contrast to the reduced red cell deformability in falciparum malaria, red cells in vivax malaria have increased deformability [44
], which cannot explain reduced VC
volume. Rosetting, the adherence of nonparasitized red cells to parasitized red cells, is known to occur in vivax malaria [41
] and could contribute to reduced pulmonary blood flow. However, reduced flow and other potential causes of reduced VC
volume such as vasoconstriction and pulmonary hypertension have not previously been found to cause reduced VC
volume without also causing reduced DM
]. Pulmonary congestion from heart failure causes normal or increased VC
volume and reduced DM
]. The pattern of uncoupling that we report in vivax and falciparum malaria [34
] has not, to our knowledge, otherwise been described.
The progressive impairment of DM
function was statistically significant after treatment of vivax malaria but not after treatment of uncomplicated falciparum malaria. The reduction in DM
function represents alveolar-capillary dysfunction, with potential causes including endothelial injury, alterations in endothelial permeability causing interstitial and alveolar edema, and intravascular sequestration of leukocytes. An inflammatory cause is supported by autopsy studies in vivax malaria showing increased alveolar-capillary monocytes [8
] and isotopic studies showing increased pulmonary phagocytic cell activity 1–2 days after the commencement of treatment for vivax malaria [27
]. This likely reflects a posttreatment intravascular inflammatory response to the death of parasites or reperfusion and is consistent with the fact that all clinical cases of ARDS in vivax malaria that have included a treatment history [10
] have occurred after the start of treatment. If vivax parasites do preferentially sequester in the lung, this may target the posttreatment inflammatory response to the lungs rather than to other organs, with both phenomena potentially explaining why vivax malaria causes lung injury.
Our finding of significant progressive alveolar-capillary dysfunction after treatment of vivax malaria but not uncomplicated falciparum malaria, despite a lower initial parasitemia, is consistent with a greater inflammatory response to a given parasite burden in P. vivax,
compared with that in P. falciparum.
This is also supported by the subclinical changes in oxygen saturation, which improved after treatment in uncomplicated falciparum malaria but decreased after treatment in those with vivax malaria. These organ-specific findings are consistent with previous studies comparing the systemic inflammatory response between species, with higher plasma levels of the fever-inducing cytokine TNF-α in vivax than in falciparum malaria in some [32
] but not all studies.
As in our previous studies of severe malaria [34
], confounding factors are unlikely to explain the gas transfer findings. Fever alone is an unlikely factor, because gas transfer worsened after the subjects became afebrile, and overall diffusing capacity is not greatly influenced by temperature [29
]. Differential effects of fever on DM
function and VC
volume are reported to be negligible [29
]. Anemia, which is worse in patients with vivax malaria, is taken into account with a correction factor applied in the derivation of DLCO
]. Spirometric results were normal in vivax malaria and cannot explain the reduction in gas transfer. There remains some uncertainty relating to the assumption implied in the derivation of θ
, the reaction rate for CO with red cells. Although we cannot exclude an alteration in θ
due to changes in internal red cell viscosity, membrane physicochemical properties, red cell shape, and alterations in hemoglobin in vivax malaria, θ
does not appear to be significantly altered in red cells infected with P. falciparum
when controlled for hemoglobin concentration (B. Russell, Menzies School of Public Health Research, Darwin, personal communication). There was no difference in the proportion of patients with vivax and falciparum malaria with gas transfer studies of adequate quality, and this is unlikely to account for the differences in gas transfer between the groups.
Chloroquine has anti-inflammatory effects. These may have attenuated inflammatory changes after treatment of vivax malaria, and our estimates of reduced DM
function may be conservative. Because chloroquine-resistant vivax malaria spreads [2
], it is possible that use of drugs without anti-inflammatory effects may lead to an increase in the incidence of acute lung injury in vivax malaria. Supervised high doses of primaquine can result in modest but clinically insignificant levels of methemoglobinemia. The lower primaquine dose administered in the present study and the known poor compliance with prolonged courses make it unlikely that primaquine-induced methemoglobinemia would have significantly confounded gas transfer findings. Methemoglobinemia should not affect DM
function, although it may possibly reduce VC
volume. However, VC
volume was impaired before primaquine administration and had normalized by day 7, by which time any drug-induced methemoglobin would have had its effect.
As was previously found in nonimmune travelers [27
], cough also occurred in the majority of adults with vivax malaria living in a malaria-endemic area. Vivax malaria is likely to be underrecognized as a cause of fever and cough in areas of endemicity and, as with P. falciparum
], may be mistaken for acute respiratory infection [27
]. Studies in Indonesian children have shown that cough and chest crackles are at least as common in vivax as in falciparum malaria [49
]. The respiratory manifestations of vivax malaria may have implications for the syndromic treatment of fever, cough, and fast breathing in young children. In areas where P. falciparum
is endemic, the World Health Organization has recommended treatment for both malaria and pneumonia [50
]. Prospective studies are necessary to determine the magnitude of overlap in clinical presentations between vivax malaria and pneumonia and whether these recommendations need to be extended to vivax malaria–endemic areas.
In conclusion, the membranous and vascular components of gas transfer are uncoupled in vivax malaria, both at presentation and after treatment. The reduction in VC volume on presentation without impairment of DM function suggests that, before treatment, the reduced VC volume of hemoglobin-containing cells may be due to the sequestration of parasitized red cells more than leukocytes within the pulmonary vasculature. The return of VC volume to normal after treatment is consistent with the clearance of sequestered parasitized cells. The progressive and significant decrease in DM function after treatment suggests prolonged alveolar-capillary injury, which we hypothesize is due to an inflammatory response to parasite killing or reperfusion. If vivax parasites do preferentially sequester in the lung, this may target this posttreatment inflammatory response to the lungs rather than other organs, with both phenomena potentially explaining why vivax malaria causes lung injury but rarely injury to other organs. The posttreatment decrease in DM function with treatment was statistically significant in vivax malaria but not in uncomplicated falciparum malaria and occurred despite a lower mean parasitemia. This is consistent with the inflammatory response to the toxins of P. vivax being greater than that to the toxins of P. falciparum.