Although
P. ovale spp. was identified by Stevens in 1922 [
6], it has received limited attention in medical research. In 2010, two sympatric species
, P. ovale curtisi and
P. ovale wallikeri, were confirmed through sequence analysis [
7]. However, few studies have described their epidemiological differences and have not reported consistent results. A recent study of imported cases in the United Kingdom reported that
P. o. wallikeri had a shorter latency than
P. o. curtisi [
13], concurrent with another study on imported cases in Henan Province, China [
14]. However, another study of imported cases in Jiangsu Province, China, did not report a significant difference in the latency period [
15]. In this study, the difference in the latency period of two sympatric species was slightly short of statistical significance (p = 0.070), suggesting that
P. ovale wallikeri has a shorter latency periods; however, larger studies are needed to confirm these results at the statistical significance threshold. In the study, the longest latency period of
P. ovale spp. was 1299 d, which was noted in a case of a co-infection with
P. o. curtisi and
P. falciparum. These data are concurrent with the longest latency periods reported by Nolder et al. (1083 days) [
13] and Zhou et al. (1265 days) [
14]. The long latency periods of
P. o. curtisi are worthy of attention, probably resulting in residual foci. It would be challenging to eliminate malaria in
P. ovale spp.-endemic areas. Moreover, the concept of a latency period in
P. ovale spp. has been challenged owing to limited experimental and clinical data supporting the hypnozoite model
. Richter et al. reported that evidence was not sufficient to unequivocally demonstrate that
P. ovale spp. hypnozoites are found in the human host [
17]. Another review reported only 18 cases of relapse for
P. ovale spp. in nearly 100 years [
18]. However, a recent study provides direct evidence, using molecular analyses, of the reemergence of
P. ovale curtisi strains, concurrent with the currently accepted relapse theory. Interestingly, relapse of
P. ovale wallikeri infections was not noted in this study [
19]. In this study, co-infections of
P. ovale spp.
/ P. falciparum were noted, wherein two clinical attacks were triggered (the first attack due to
P. falciparum, and the second due to
P. ovale spp.). At the primary
P. falciparum attack, patients were administered artemisinin-based drugs without primaquine, the only effective drug against dormant liver parasites [
20]. The second clinical attack due to
P. ovale spp. occurred at a different interval. The remaining
P. ovale spp. infections were treated with primaquine after diagnosis and no parasites reemerged in the blood samples (after the final diagnostic finding of the Malaria Diagnostic Reference Laboratory of Anhui Province feedback,
P. ovale spp. were treated with primaquine to prevent relapse). Our findings potentially support the existence of hypnozoites of
P. ovale spp. Moreover, the six cases of co-infections with two clinical episodes constitute the most important findings of this study. Identification of such co-infections is difficult in an endemic transmission setting. As
P. ovale spp. infections are no endemic in China, if the present patient does not travel abroad again, it should be easy to diagnose co-infections. Furthermore, long-term observation would be required. In contrast, in endemic areas, especially in the high transmission setting, numerous cases of malaria were reported. In this context, if the interval between two clinical attacks was too long, it would be difficult to link these clinical events together, probably being considered independent. These six cases strongly suggest that the potential of
P. ovale spp. to co-infect with other malarial species has been underestimated because cases as such were difficult to come across in cross-sectional surveys. Data from two recent studies provide evidence regarding this perspective [
21,
22]. It is worth noting that 71.0%
P. ovale curtisi and 80.4%
P. ovale wallikeri are associated with a previous history of malaria, suggesting that more co-infections, presenting with two clinical attacks, may not be reported. Because our surveillance system only recorded the second clinical attack due to
P. ovale spp. within the country, and not abroad, the proportion of
P. ovale spp. may have increased. Nonetheless, more evidence is required. Mehlotra et al.reported that the higher overall prevalence of
P. falciparum and
P. vivax in a human population is associated with fewer mixed infections than expected [
23], using point-prevalence data from 35 other studies. Based on our results, a potential explanation is that a moderate proportion of co-infections involving hypnozoites was undetected.
McKenzie and Bossert et al. reported that infection by one
Plasmodium species does not reduce the susceptibility to infection by other species [
24]. Therefore, an increase in the rate of co-infections suggests that the prevalence of
P. ovale spp. has been underestimated. In this study, we examined the returnees from four African countries to estimate the incidence rates of
P. ovale spp. and
P. falciparum. Although it is only an estimate, with
P. falciparum as a reference, the data suggest that the incidence of
P. ovale spp. was higher than expected and a growing body of evidence supports this hypothesis [
25‐
27]. Considering the geographic distribution, our findings suggest that the two sympatric species,
P. o. curtisi and
P. o. wallikeri, have been circulating simultaneously in numerous Africa countries, concurrent with a previous study [
28].
Owing to their similarities in morphology and life cycle,
P. ovale spp. is easily and frequently misdiagnosed as a
P. vivax infection [
29,
30]. In this study, only 20.00% of
P. ovale spp. cases (23/115) returned an accurate species identification on microscopic examination; 32.17% (37/115) of cases were misdiagnosed as
P. vivax infections and 6.96% (8/115) of cases were misdiagnosed as
P. falciparum infections. For the remaining 40.87% (47/115) of cases, species identification was not attempted, and only the results of parasite identification (positive or negative findings) were obtained. The success rates of species identification were significantly different between
P. ovale spp. (20.00%) and
P. falciparum (86.77%), Probably because two rapid diagnostic tests (Pf/pan, by Wondfo and by ACCESSBIO) highly sensitive to
P. falciparum, but insensitive to
P. ovale spp. [
31,
32], have been used to provide parasite-based diagnosis in Anhui province since 2013. Misdiagnosis of
P. ovale spp. infections as
P. vivax or
P. falciparum infections may have led to inappropriate case management measures and treatment regimens. Therefore, more sensitive point-of-care detection methods for
P. ovale spp. need to be developed and introduced in non-endemic areas.
Our study has several limitations. First, it was a case-based, retrospective study, and was subject to recall bias. Second, some differences were simply beyond the threshold for statistical significance, requiring larger studies to confirm them. Lastly, the data on the use of prophylaxis are limited, thus potentially affecting the latency period of P. ovale spp.