First case of Plasmodium knowlesi infection in a Japanese traveller returning from Malaysia

The first naturally acquired zoonotic infection with Plasmodium knowlesi in a human was reported in 1965 [1]; thereafter, no such infections were reported for almost 40 years. In 2004, however, Singh et al.[2] reported that natural P. knowlesi infections in humans were common parasite species in Malaysia. Cox-Singh et al.[3] subsequently described that P. knowlesi infections in humans were observed throughout wide areas of Southeast Asia, including Thailand, Myanmar, the Philippines, Singapore and Indonesia. Additionally, 12 cases of P. knowlesi infection in travellers have been reported from non-malaria endemic countries thus far [1, 4‐14]. Here, this is the first confirmed case of P. knowlesi infection in a Japanese traveller.

In September 2012, a previously healthy 35-year-old Japanese man presented to the travel clinic in National Center for Global Health and Medicine (NCGM), Tokyo with a two-day history of daily fever, mild headache, and mild arthralgia. He had visited Malaysia for entomological and botanical field investigations over a two-month period and had stayed at Temengor (four weeks), Johor (two weeks), and Kuala Lumpur (two weeks). While in Temengor, he stayed in a tent located near a forest and had not used any malaria prevention measures, such as bed nets, mosquito repellents or chemoprophylaxis. During his stay, he was bitten by mosquitoes and saw some wild monkeys. He had no health problems and was in a good physical condition until he experienced a sudden high fever (39.0°C axillary temperature) the day after his return to Japan. He had fever spikes of >38.0°C in a 24-hour period. On the third day of his illness, he was admitted to NCGM. On admission his body temperature was 37.0°C, with blood pressure of 111/80 mmHg, pulse rate of 115 beats per minute, respiration rate of 16 per minute, and oxygen saturation of 99% (room air). Physical examination revealed mild splenomegaly and left upper abdominal pain. Laboratory investigations showed thrombocytopenia (47 × 10³ cells/mm³, reference range 150–350 cells/mm³), elevated liver enzymes (serum aspartate aminotransferase 49 U/L, reference range 13–33 U/L; alanine aminotransferase 41 U/L, reference range 8–42 U/L; alkaline phosphatase 428 U/L, reference range 115–359 U/L; and total bilirubin 1.1 mg/dl, reference range 0.3-1.2 mg/dl), a high C-reactive protein level (11.59 mg/dl, reference range 0–0.3 mg/dl), no anaemia (haemoglobin 17.3 g/dl, reference range 13.5-17.0 g/dl), and a normal leukocyte count (3,860 cells/mm³, reference range 3,500-8,500/mm³). The patient’s free blood glucose levels and renal function were normal. Although rapid diagnostic tests (BinaxNOW Malaria®; Alere, Inc, Scarborough, USA) for the Plasmodium falciparum histidine-rich protein 2 and pan-malarial aldolase were negative, malaria parasites were identified in the patient’s Giemsa-stained thin blood smear from which a parasitaemia of 0.2% (10,120 parasites/μl) was calculated. The morphological features of the parasites comprised band forms, which are similar to those characteristics of Plasmodium malariae, and heavily pigmented schizonts inside non-enlarged erythrocytes (Figure 1). To make a definitive diagnosis, two partial gene regions (the small subunit rRNA from nuclear DNA and cytochrome b from mitochondrial DNA) from five malaria parasite species were tried to be amplified by nested PCR, using respective primer sets for P. falciparum, Plasmodium vivax, Plasmodium ovale, P. malariae and P. knowlesi. The positive result was obtained by the primer sets of P. knowlesi and P. vivax in the amplification of the partial small subunit rRNA gene. However, in the amplification of the partial cytochrome b gene, the positive result was obtained by the P. knowlesi primer sets (Figure 2). The sequences of the PCR products were determined using an ABI 3130xl Genetic Analyzer (Applied Biosystems, CA, USA) after TA cloning. From this sequence analysis, the patient was confirmed that even the PCR product amplified by the P. vivax primer sets for the partial small subunit rRNA gene was proved to be the DNA of P. knowlesi. He was treated with mefloquine (1,500 mg base), then was repeated the microscopic screening for parasites every 12 hours for confirming parasite disappeared. The clinical course of the mefloquine treatment in this patient was as follows: the initial band-form and schizont-stage parasites seen in the Giemsa-stained blood smears disappeared six hours after administration of the drug and, gradually, the shape of the ring-form parasites became more irregular over time (Figure 1). The parasites disappeared completely from the blood 40 hours later, and the patient’s fever resolved 28 hours after mefloquine administration. The patient was discharged from hospital on the seventh day post-admission, without any complications. No relapses or any other health problems were observed over a five-month period.

Most P. knowlesi infections are reported in the rural jungle areas of the Malaysia peninsular and in Borneo (Sabah [17] and Sarawak [2] province) where both of the infectious hosts (macaque and leaf monkeys) and vectors (Anopheles mosquitoes) predominantly exist. The patient in this case was probably infected with P. knowlesi in the Temengor jungle area as he had not entered any other areas of jungle for four weeks before returning to Japan. Interestingly, the incubation period in this patient was probably over four weeks, which is longer than the nine to 12 day incubation period reported for P. knowlesi[18]. This finding is consistent with a case report for a different traveller who also experienced a long incubation period (>17 days) [13]. Clinical data for P. knowlesi infection in travellers are limited; therefore, more data are needed to determine the correct incubation period for P. knowlesi.

No cases of treatment failure have been reported in travellers treated with anti-malarial drugs such as quinine [4], doxycycline [4], mefloquine [8], atovaquone/proguanil [5, 10, 14], and artemether/lumefantrine [12]. In the present case, mefloquine enabled the patient to recover without any complications. Chloroquine has been usually used for uncomplicated P. knowlesi infection [19]; however, in recent years, a prospective comparative study suggested that early intravenous artemisinin treatment improved the prognosis in severe cases of P. knowlesi infection [17]. The major risk factors of increasing severity are associated with a high parasitaemia [17, 20] and thrombocytopenia [20]. The risk factors of severity in P. knowlesi infection should be evaluated appropriately, and if necessary, should be considered the intravenous artemisinin treatment as other malaria parasite species [21].

Microscopic examination of Giemsa-stained blood smears is the gold standard for differentiating Plasmodium species; however, differentiating P. knowlesi from other malaria species is very difficult because P. knowlesi band forms and schizonts of mature trophozoites in P. knowlesi are similar to those of P. malariae, while the ring forms of P. knowlesi early trophozoites are similar to those of P. falciparum[2, 17, 18]. This patient was initially suspected of contracting P. malariae after observing his Giemsa-stained thin blood smear (Figure 1). Although PCR is required to obtain a definitive diagnosis, the spurious amplification of a P. vivax gene using P. knowlesi-specific primers has been reported [22]. In the present study, on the other hand, a P. knowlesi gene was amplified using P. vivax-specific primers [16]. These spurious amplifications were observed when the target gene of PCR was the small subunit rRNA gene [22] because of the similarity between the DNA sequences of P. knowlesi and P. vivax. Thus, a caution is needed for using this gene to differentiate malaria species by PCR. Indeed, some cases of P. knowlesi infection in travellers have been initially diagnosed as P. ovale[4] or P. vivax[13] by nested PCR; however, these were later confirmed as P. knowlesi by DNA sequencing. Therefore, DNA sequencing is indispensable for final confirmation of P. knowlesi infection.

There is limited access to DNA sequencing procedures for identifying P. knowlesi in endemic areas. Hence, accurate reporting of clinical and epidemiological data should contribute to better understanding of the clinical features of P. knowlesi infection, as well as the true incidence of the disease. Close monitoring of febrile travellers returning from P. knowlesi-endemic areas and complete travel histories (including exposure to areas with wild monkeys) should prompt suspicion of P. knowlesi infection [17] and a move towards further diagnostic tests. Additionally, greater awareness of the risk of the emerging P. knowlesi problem in travellers is necessary among health care workers in non-endemic countries. Thus, reporting accurately diagnosed cases of P. knowlesi infection in travellers would be important for assessing the true nature of this emerging human infection.

Oral informed consent was obtained from the patient for publication of this case report and any accompanying images after explanation of the report objectives.