Performance of Rapid Diagnostic Tests for Plasmodium ovale Malaria in Japanese Travellers.

BACKGROUND: Rapid diagnostic tests (RDTs) are used widely in the diagnosis of malaria. Although the effectiveness of RDTs for malaria has been described in many previous studies, the low performance of RDT particularly for Plasmodium ovale malaria in traveller has rarely been reported.
METHODS: This was a retrospective cohort study conducted on Japanese travellers diagnosed with malaria at the National Center for Global Health and Medicine between January 2004 and June 2013. The diagnosis of malaria was confirmed by microscopic examination, RDT, and polymerase chain reaction in all patients. The RDTs used in our study were Binax NOW Malaria (Binax Inc., Scarborough, Maine, USA) (BN) and SD Malaria Antigen Pf/Pan (Standard Diagnostics Inc., Korea) (SDMA). We compared the sensitivity of the RDTs to P. ovale malaria and Plasmodium vivax malaria.
RESULTS: A total of 153 cases of malaria were observed, 113 of which were found among Japanese travellers. Nine patients with P. ovale malaria and 17 patients with P. vivax malaria undergoing RDTs were evaluated. The overall sensitivity of RDTs for P. ovale malaria and P. vivax malaria was 22.2% and 94.1%, respectively (P < 0.001). The sensitivity of SDMA for P. ovale malaria and P. vivax malaria was 50% and 100%, respectively. The sensitivity of BN for P. vivax malaria was 90.0%, but it was ineffective in detecting the cases of P. ovale malaria.
CONCLUSIONS: The sensitivity of RDTs was not high enough to diagnose P. ovale malaria in our study. In order not to overlook P. ovale malaria, therefore, microscopic examination is indispensable.

Introduction

Malaria is a serious and sometimes fatal disease. Early and accurate detection of malaria infection is important to reduce its associated morbidity and mortality [1]. The gold standard for the detection of malaria infection is microscopic examination, and the most reliable method is genetic examination by polymerase chain reaction (PCR) or deoxyribonucleic acid (DNA) sequencing [2]. Rapid diagnostic tests (RDTs) for malaria are widely used as a substitute for microscopic examination because of their good accessibility. RDTs detect different target antigens such as histidine rich protein 2 (HRP2) [3-5] or Plasmodium falciparum-specific lactate dehydrogenase (PfLDH) [5-7] for Plasmodium falciparum, Plasmodium vivax-specific lactate dehydrogenase (Pv-pLDH) [8], and aldolase [3, 4, 6] or plasmodium lactate dehydrogenase (pLDH) [5, 7, 9, 10] for P. vivax, Plasmodium ovale and Plasmodium malariae.

Although the usefulness of RDTs for P. falciparum is well recognized [11, 12], RDTs are often ineffective in detecting other malaria species. Moreover, very few reports have described the performance of RDT in travellers particularly regarding P. ovale malaria. For returned travellers in non-endemic settings, it is sometimes difficult to distinguish P. vivax malaria from P. ovale malaria without PCR because the clinical and microscopic features are very similar. We investigated the performance of RDTs for the detection of P. ovale malaria in Japanese travellers by comparing their performance with that of RDTs for P. vivax malaria.

Methods

This study was a retrospective cohort study on Japanese travellers diagnosed with malaria at the National Center for Global Health and Medicine (NCGM) between January 2004 and June 2013. Malaria cases were diagnosed by microscopic examination using initial blood samples at the NCGM Research Institute. Parasite concentration (parasitemia) was calculated by counting the number of parasites per 200,000 red blood cells in a thin blood smear. The RDTs used in our study were Binax NOW Malaria (Binax Inc., Scarborough, Maine, USA) (BN), which detects histidine-rich protein for the identification of falciparum malaria and plasmodium aldolase for the identification of non-falciparum malaria, and SD Malaria Antigen Pf/Pan (Standard Diagnostics Inc., Korea) (SDMA), which detects plasmodium lactic acid dehydrogenase (pLDH) for the identification of all four human malaria parasites. BN could not be obtained in our institute between July 2008 and May 2010, which made it necessary to use SDMA during the period instead of BN. RDTs were performed once for each sample. Concerning the line intensity reading, a test line strong enough to read was deemed positive. Microscopic examination and RDTs were immediately conducted after taking blood tests, and RDTs were conducted in accordance with the manufacturer’s instructions.

When parasites were detected microscopically, the species was confirmed by nested PCR. The small subunit rRNA (SSU rRNA) from the nuclear DNA of the four human malaria parasite species were amplified by nested PCR using respective primer sets for P. falciparum, P. vivax, P. ovale and P. malariae. The nested PCR condition and the species specific primers followed Kimura’s method [13]. To detect P. knowlesi DNA, the SSU rRNA from nuclear DNA and cytochrome b from mitochondrial DNA of P. knowlesi were amplified by nested PCR. The sequences of the PCR products were determined using an ABI 3130xl Genetic Analyzer (Applied Biosystems, CA, USA) after TA cloning. P. knowlesi specific primer set, the condition of the nested PCR and DNA sequencing were the same as those used in our previous study [14].

Statistical analysis was performed by Stata ver.11 (Stata Corp, Texas, USA). The chi-square and Fisher’s exact probability tests were used to compare the characteristics of P. ovale malaria and P. vivax malaria for dichotomous variables, and the Student’s t-test was used to compare continuous variables. A difference of P < 0.05 was considered significant.

Results

Microscopic examination was performed for diagnosis in all of the malaria cases. A total of 153 cases of malaria were observed between January 2004 and June 2013, 113 of which were found among Japanese travellers. Of the latter, there were 78 cases of P. falciparum malaria (including 1 case of co-infection with P. malariae malaria and 1 case of co-infection with P. vivax malaria confirmed by PCR), 22 cases of P. vivax malaria (including 1 case of co-infected with P. falciparum malaria confirmed by PCR), 11 cases of P. ovale malaria, 3 cases of P. malariae malaria, and 1 case of P. knowlesi malaria. RDTs were performed in 67, 21, 10, 2, 1 and 2 cases among the mono-infection cases of P. falciparum malaria, P. vivax malaria, P. ovale malaria, P. malariae malaria, P. knowlesi malaria and co-infection cases, respectively. The cases confirmed by PCR were 11, 19, 10, 3 and 1 cases in P. falciparum malaria, P. vivax malaria, P. ovale malaria, P. malariae malaria and P. knowlesi malaria, respectively (Table 1).

Table 1.

The characteristics of 153 malaria cases from microscopic analysis

		Japanese Travellers	RDTs	PCR
Total	153	113	103	42
Mono-Infections
P. falciparum		76	67	9
P. vivax		21	21	18
P. ovale		11	10	10
P. malariae		2	2	2
P. knowlesi		1	1	1
Co-Infections
P. falciparum + P. malariae		1	1	1
P. falciparum + P. vivax		1	1	1

We focused on a comparison of P. ovale malaria mono-infection cases and P. vivax malaria mono-infection cases. There were 10 cases of P. ovale malaria (mean age, 27.8 years; male-to-female ratio, 2.3:1) and 18 cases of P. vivax (mean age, 29.8 years; male-to-female ratio, 2.6:1). These patients constituted the final study population. The results of RDTs were evaluated for 9 patients with P. ovale malaria and 17 patients with P. vivax malaria (excluding 1 co-infected case).

Among the 18 cases of P. vivax malaria, 13 (72.2%) had been infected in Asia, 4 (22.2%) in South America, and 1 (5.6%) in Africa (Republic of Rwanda) (Table 2). All of the 10 cases of P. ovale malaria had been infected in Africa (Table 3). The average parasitemia in the cases of P. ovale malaria was significantly lower than that in P. vivax malaria (P = 0.002). The sensitivity of RDTs for P. ovale malaria was also significantly lower than that for P. vivax malaria (P < 0.001). The sensitivity of SDMA for P. vivax malaria and P. ovale malaria was 100% (7/7) and 50% (2/4), respectively. The sensitivity of BN for P. vivax malaria was 90.0% (9/10), but it was ineffective in detecting the cases of P. ovale malaria (0/5; Table 4). There were only 3 cases of P. malariae malaria, one of which showed co-infection with P. falciparum. The other 2 cases of P. malariae malaria were mono-infection of P. malariae. In the mono-infection cases, one case of P. malariae malaria showed a positive line detecting pLDH of SDMA while the other case showed a positive line detecting aldolase of BN.

Table 2.

The characteristics of the cases of P. vivax malaria

No.	Year	Country	Type of RDTs	Result of RDTs (T1/T2, P.f, Pan)*	Parasitemia (%)	Result of PCR
1	2004	Banuatu	BN	+/+	0.15	Pf, Pv
2	2006	Rwanda	BN	−/+	0.18	Pv
3	2007	Indonesia	BN	−/+	0.20	Pv
4	2007	Indonesia	SDMA	−/+	ND	Pv
5	2007	PNG	BN	−/+	0.20	Pv
6	2007	Malaysia	BN	−/−	0.005	Pv
7	2007	Brasil	BN	−/+	0.79	Pv
8	2008	India	SDMA	−/+	0.06	Pv
9	2008	India	SDMA	−/+	0.40	Pv
10	2009	PNG	SDMA	−/+	0.13	Pv
11	2009	Equador	SDMA	−/+	0.15	Pv
12	2009	Thailanda, India, Nepal, Bangladesh	BN	−/+	0.06	Pv
13	2009	French Guiana	SDMA	−/+	0.34	Pv
14	2010	Sudan	SDMA	−/+	0.22	Pv
15	2010	Brasil	BN	−/+	0.10	Pv
16	2010	Indonesia	BN	−/+	0.09	Pv
17	2010	India	BN	−/+	0.33	Pv
18	2011	PNG	BN	−/+	0.02	Pv

PNG: Papua New Guinea; BN: Binax NOW® malaria; SDMA: SD malaria antigen®; ND: not described; Pf: Plasmodium falciparum; Pv: Plasmodium vivax.

*T1: the positive band of Histidine rich protein 2 and T2: the positive band of aldolase for BN, P.f: the positive band of Plasmodiun falciparum lactate dehydrogenase (LDH) and Pan: the positive band of Plasmodium LDH, for SDMA.

Table 3.

The characteristics of the cases of P. ovale malaria

No.	Year	Country	Type of RDTs	Result of RDTs (T1/T2, P.f, Pan)*	Parasitemia (%)	PCR
1	2004	Ghana	BN	−/−	0.030	Po
2	2005	Senegal	BN	−/−	0.060	Po
3	2006	Uganda	ND	ND	0.040	Po
4	2007	Uganda	BN	−/−	0.040	Po
5	2008	Uganda	SDMA	−/−	0.001	Po
6	2009	Ghana	SDMA	−/+	0.080	Po
7	2010	Uganda	SDMA	−/+	0.004	Po
8	2010	Ghana	SDMA	−/−	0.012	Po
9	2012	Uganda	BN	−/−	0.042	Po
10	2013	Uganda	BN	−/−	0.003	Po

BN: Binax NOW® malaria; SDMA: SD malaria antigen®; M: male; F: female; ND: not described, Po: Plasmodium ovale.

*T1: the positive band of Histidine rich protein 2 and T2: the positive band of aldolase for BN, P.f: the positive band of Plasmodiun falciparum lactate dehydrogenase (LDH) and Pan: the positive band of Plasmodium LDH, for SDMA.

Table 4.

The comparison between P. vivax malaria cases and P. ovale malaria cases

	P. vivax malaria	P. ovale malaria	P value
Number of patients	18	10
Age (mean ± SD)	29.8 ± 12.0	27.8 ± 7.0	0.326
Sex (male)	13	7	0.900
Areas to travel
Africa	1	10
Asia, Oceania	13	0	< 0.001
South America	4	0
Average parasitemia (%)	0.229	0.031	0.002
Positive by RDTs (%)	16/17 (94.1)	2/9 (22.2)	< 0.001
BN (%)	9/10 (90.0)	0/5 (0)	< 0.001
SDMA (%)	7/7 (100)	2/4 (50.0)	0.039

The data are presented as number of patients, unless otherwise specified.

P. vivax: Plasmodium vivax; P. ovale: Plasmodium ovale; RDT: rapid diagnostic test; BN: Binax NOW® malaria; SDMA: SD malaria antigen®

Discussion

RDTs are generally useful for diagnosing malaria infection [3-5]. However, the sensitivity of RDTs depends upon the species of malaria parasite. The highest sensitivity was observed for P. falciparum malaria (78.8–99.1%) [5, 11, 15–18], followed by P. vivax malaria (77.6–96%), P. ovale malaria (18.4–80.0%), and P. malariae malaria (21.4–47.0%) [3–5, 19, 20]. For non-falciparum malaria, RDT sensitivity was particularly low for P. ovale and P. malariae, a finding that may be attributable to the low parasitemia [3-5], the difference in targeted antigens [6, 9, 10], or the genetic variability between the infected parasites [21, 22].

Low parasitemia was associated with false-negative results from the RDTs regardless of malaria species [4, 5]. Moreover, since reinfection and semi-immune status generally cause low parasitemia [23], those influences need to be eliminated. Travellers who visiting families and relatives and expatriates living in malaria endemic areas were excluded from our study because they were at a higher risk of reinfection. The parasitemia in P. ovale malaria was consequently as low as that in immune patients and was significantly lower than that in P. vivax malaria. This finding, along with those in previous reports [23], suggests that P. ovale malaria presents with low parasitemia even in non-immune travellers, which may result in false-negative results from the RDTs.

As mentioned above, P. malariae and P. ovale malaria may be difficult to diagnose by RDTs due to the low parasitemia [7, 8, 11]. Recently, Houzé et al. [24] reported that the new RDT, Clearview, has improved sensitivity for P. malariae malaria but that the sensitivity for P. ovale malaria was not improved. Moreover, even a case of P. ovale malaria with relatively high parasitemia (21,150 parasites/μL) resulted in a false-negative RDT result in their study, implying some reason for false-negative results other than low parasitemia in P. ovale malaria. The difference in target antigens from parasites, such as HRP2 and PfLDH for P. falciparum or pLDH and aldolase for pan-malaria species, can also influence the sensitivity of RDTs [6]. Detecting P. ovale and P. vivax, two RDTs (BN for aldolase detection and SDMA for pLDH detection) were used in our study. SDMA showed relatively good sensitivity for both P. vivax and P. ovale as compared to BN, probably because pLDH-based RDTs generally perform better than aldolase-based RDTs [20]. Bigaillon et al. [6] also reported the ineffectiveness of BN in detecting P. ovale malaria, which they suggested was due to low aldolase production by P. ovale malaria [10]. Aldolase-based RDTs generally show low sensitivity not only for P. ovale malaria but also for other types of malaria [9], but low sensitivity was also observed for pLDH-based RDTs [10–12, 16, 17, 24].

Although the genetic variability of HRP2 was related to the low sensitivity of RDTs for P. falciparum malaria [21], it was reported that the genetic variability of pLDH [22] and aldolase [25, 26] did not explain the relatively poor performance of RDTs for the detection of P. falciparum, P. vivax, and P. malariae. Talman et al. [22] reported that P. ovale exhibited three different types of amino acid sequence (named O1, O2 and O3), whichmay contribute to the relatively poor detection of P. ovale. The sensitivity of RDTs is still insufficient for accurate diagnosis of P. ovale malaria regardless of the type of antigens. Therefore, microscopic examination is preferable for the definitive diagnosis of P. ovale malaria.

The sensitivity of RDTs was not high enough to accurately diagnose P. ovale malaria in Japanese travellers who were never infected with malaria previously. Thus, microscopic examination is required to ensure that P. ovale malaria is not overlooked.