Sensitivity of Parasitological Tests in Imported Plasmodium vivax Malaria in Adults and Impact of Chemoprophylaxis and Attack Type

Authors


Abstract

Background

Plasmodium vivax is the second most common species among cases of imported malaria diagnosed in Europe. The objective of this study is to describe the sensitivity of the parasitological tests in imported P. vivax malaria, and the impact of chemoprophylaxis and attack type (primary infection or relapse).

Methods

A retrospective study included the imported vivax malaria cases admitted in a French military hospital between 2001 and 2013. The reference diagnosis method was microscopy corrected by polymerase chain reaction (PCR). Thin and thick blood films examination, quantitative buffy coat (QBC) test, and a rapid diagnostic test (RDT) had been systematically performed. PCR had been carried out for ambiguous profiles.

Results

Eighty-nine cases recorded from 78 patients were included, 65 of them having recently traveled to French Guyana. Forty-two patients had properly followed chemoprophylaxis. Forty-six cases were primary infections while 43 were relapses. The sensitivity was 91% for the thin blood smear, 96% for the concentration techniques (Giemsa thick blood smear and QBC test), and 76% for the RDT. The combination of the three conventional tools has an imperfect sensitivity, both for the positive diagnosis of malaria (96%) and for the diagnosis of vivax species (80%). In 4% of the cases, the positive diagnosis was established only by the PCR. The species identification was established in 20% by the PCR. The sensibility of thin blood smear and of RDT decreased significantly with full compliance of chemoprophylaxis or primary infection, whereas the decrease of sensibility of concentration techniques was not significant.

Conclusions

This study illustrates the difficulties encountered in vivax malaria diagnosis, especially in patients who properly followed chemoprophylaxis or with primary infection due to a lower parasitemia. It underlines the lack of sensitivity of RDT for P. vivax and emphasizes the need for systematically combining various diagnosis methods.

Imported malaria is defined as an infection acquired in a malaria-endemic area but diagnosed in a nonendemic country after development of clinical signs. Plasmodium vivax is the second most common species among cases of imported malaria diagnosed in Europe.[1] In metropolitan France, the estimated number of imported malaria cases was 3,560 in 2011 among which 7.4% were connected to P. vivax.[2] The extension of the exposure areas, the report of severe cases, and the decline of chloroquine susceptibility in Indonesia, Peru, and Oceania justify the renewed medical interest for this species.[3]

The clinical presentation is nonspecific and can arise several months after the return from the exposure zone, because of relapse due to activation of dormant intrahepatic parasitic stages.[4] Prompt and accurate diagnosis of the disease is essential for targeting patients requiring antimalarial treatment and for differential diagnosis with Plasmodium falciparum. Plasmodium vivax and Plasmodium ovale require a specific therapy for radical cure, combining a blood schizonticide as chloroquine and primaquine to eradicate liver hypnozoites.[5]

While several studies have reported epidemiological, clinical, and biological features of imported vivax malaria,[1, 4, 6-11] there are few data about accuracy of parasitological tests in routine practice. Conventional techniques (microscopy and rapid diagnostic test, RDT) remain the first-line tools but difficulties have been reported.[12, 13] Molecular techniques are considered as the new gold standard but they are not yet available in all laboratories.[14] Similarly, the impact of chemoprophylaxis or of attack access (ie, primary infection or relapse) remains unknown. The purpose of this study was to assess the sensitivity of conventional procedures for the diagnosis of imported vivax malaria, and the impact of chemoprophylaxis and attack type.

Patients and Methods

Study Design

This is a descriptive monocentric retrospective study. All patients admitted at the Bégin Military Hospital, Saint-Mandé, between January 1, 2001 and March 1, 2013, for an attack of P. vivax after staying in an endemic country were included. A case of P. vivax malaria was characterized by a signature rising fever and a positive thin blood smear with at least one typical form of P. vivax or a positive P. vivax polymerase chain reaction (PCR). The epidemiological data, the presumed exposure area, the regular use of antimalarial chemoprophylaxis, and the attack type (primary infection or relapse) have been notified from the medical file, while the parasitological tests at the admission have been extracted from the data processing system of the laboratory. The epidemiological, clinical, biological, and therapeutic characteristics of most of these patients have been described in a previous retrospective study.[8]

Parasitological Tests

The laboratory systematically carried out four biological tests:

  • A thin blood smear was examined for at least 20 minutes (200 fields, objective 100×) by an experienced microscopist. A positive diagnosis of malaria was confirmed with the presence of several Plasmodium in erythrocytic forms. A positive P. vivax diagnosis was defined with at least one typical form of P. vivax.
  • Two concentration techniques: a Giemsa thick blood film was examined for at least 5 minutes (100 fields, objective 100×) and a quantitative buffy coat (QBC) Malaria Test was also examined for at least 5 minutes.[6] A positive result was defined by at least one visible trophozoite.
  • RDT: the ICT Malaria Pf/Pv (Binax Inc., Portland, ME, USA) immunochromatographic test was used from January 2001 to June 2006. This two-band test is based on the detection of both the histidine-rich protein (HRP)2-specific from P. falciparum and the pan-specific aldolase of the genus Plasmodium. From June 2006, this RDT was replaced by the Core Malaria Pan/Pv/Pf (Core Diagnostics Ltd., Birmingham, UK). This three-band test is based on the detection of the HRP2, the pan-specific Plasmodium lactate dehydrogenate (pLDH), and the specific P. vivax LDH (PvLDH). The RDT was performed according to the manufacturer's instructions.

Based on a biologist's decision, the PCR was performed as second intention in the following cases: negative result for the thin blood films with positivity of one concentration technique or RDT, no evidence of typical form in thin blood films because of low parasitemia, doubt regarding biparasitism, or ambiguity with P. ovale for patients returning from Africa. PCR was also performed in case of a strong clinical suspicion despite the negativity of the other four techniques. The PCR protocol consisted in four SYBR Green real-time PCR specific for each Plasmodium species (P. falciparum, P. vivax, P. ovale, and P. malariae) and was performed on the Lightcycler instrument (Roche Diagnostics, Meylan, France). Sequences used for the design of primers were 18S RNA genes of the four species (molecular target previously described[15]).

Statistical Analysis

Quantitative variables were expressed as median and interquartile range, while qualitative variables were measured by absolute numbers and by percentages. The chi-square test and Fisher's exact test were used to compare data in different groups. A p value <0.05 was considered statistically significant.

Results

Population of Study

Eighty-nine malaria cases recorded from 78 patients (4 women and 74 men) have been included. The median age was 28 (interquartile range 25–37) with extremes ranging from 17 to 58 years. Most of them were military personnel (85%). The other characteristics of the patients and the cases are reported in Table 1. The majority of the patients had stayed in French Guyana (65/78, 83%). Forty-two patients (55%) had properly followed chemoprophylaxis. The prescribed medications for chemoprophylaxis were doxycycline (62/66, 94%), chloroquine–proguanil combination (2/66, 4%), mefloquine (1/66, 1%), and atovaquone–proguanil combination (1/66, 1%). Forty-six cases were primary infections (52%), while 43 were relapses (48%). No patient was admitted in intensive care unit and the outcome was successful in all cases.

Table 1. Characteristics of patients with Plasmodium vivax malaria
 Patients (n = 78)Attacks (n = 89)
Provenance  
South America66 (85%)73 (82%)
Asia6 (8%)8 (9%)
Africa5 (6%)7 (8%)
Not specified1 (1%)1 (1%)
Observance of chemoprophylaxis (CP)  
Properly followed CP42 (55%)43 (48%)
With omissions or prematurely stopped CP18 (23%)25 (28%)
No CP10 (13%)14 (16%)
Not specified8 (9%)7 (8%)
Clinical form  
Primary infection 46 (52%)
Relapse 43 (48%)

Sensitivity of Parasitological Tests

The different parasitological tests in positive malaria diagnosis and in vivax species diagnosis are presented in Table 2. The thin blood smear was positive for 81 cases (91%). Parasitemia was successfully quantified in 65 cases (73%) with a median density of 7,407 parasites/μL (interquartile range 3,802–16,164). For the other cases, parasitemia was <200 parasites/μL. In 73% of the cases, the P. vivax species identification was successfully established. A case of biparasitism (P. vivax and P. falciparum), confirmed by PCR, was evidenced in a patient returning from Comoros. The concentration techniques were positive for 85 cases (96%).

Table 2. Addition of parasitological methods, used alone or in combination in positive diagnosis of malaria and diagnosis of vivax species (n = 89)
 Positive diagnosis of malaria, n (%)Positive diagnosis of P. vivax malaria, n (%)
  1. QBC = quantitative buffy coat; PCR = polymerase chain reaction.

  2. a

    One of the tests was a two-band test.

  3. b

    A PCR was performed only in 22 cases.

  4. c

    Species identification established by the positivity of the PvLDH band, with the Core Malaria Pan/Pv/Pf.

Blood smear81 (91)65 (73)
Concentration techniques (thick smear and QBC test)85 (96)0 (0)
Rapid diagnostic test (RDT) alone68 (76)NAa
PCR alone22 (100)b22 (100)b
Smear + concentration techniques85 (96)65 (73)
Smear + RDT81 (91)71 (80)c
Smear + concentration techniques + RDT85 (96)71 (80)
Smear + concentration techniques + RDT + PCR89 (100)89 (100)

There was no significant difference between ICT Malaria Pf/Pv (18/21 positive, indicating a sensitivity of 86%) and Core Malaria Pv/Pf (50/68 positive, indicating a sensitivity of 74%) (p = 0.38). The sensitivity of RDT was statistically lower when parasitemia was under 200 parasites/μL (p < 0.001).

The best combination for the positive diagnosis of malaria has associated the thin blood smear and the concentration techniques (96%), while the best combination for the diagnosis of vivax species has associated the thin blood smear and the RDT (80%). The combination of the three tools has sensitivity <100%, both for the positive diagnosis of malaria (96%) and for the diagnosis of vivax species (80%).

In 22 cases, a PCR was performed. The result was positive for P. vivax in 100% of cases. The PCR allowed the diagnosis of vivax species in 20% of the cases (18/89). The use of PCR determined positive malaria diagnosis in 4% of the cases (4/89) where the first-line techniques had negative results.

Impact of Chemoprophylaxis on Sensitivity of Tests

The impact of chemoprophylaxis on the sensitivity of parasitological tests is presented in Table 3. In seven cases, the chemoprophylaxis regimen was not noted in the medical file.

Table 3. Impact of adherence to chemoprophylaxis (CP) on the sensitivity of parasitological tests and on the parasitemia (n = 82)
 Full compliance (n = 43)Omissions or prematurely stopped CP (n = 25)No CP (n = 14)p Value
  1. RDT = rapid diagnostic test.

Positive thin blood smear (%)35 (81%)25 (100%)14 (100%)0.02
Parasitemia < 200 parasites/μL (%)19 (44%)5 (20%)0 (0%) 0.002
Positive concentration techniques (%)39 (91%)25 (100%)14 (100%)0.29
Positive RDT (%)26 (60%)22 (88%)14 (100%) 0.002

The sensitivity of thin blood smear decreased from 100% for patients with bad compliance or absence of chemoprophylaxis to 81% for patients with full compliance to chemoprophylaxis (p = 0.02). A blood parasite density with <200 parasites/μL was significantly associated with properly followed chemoprophylaxis (p = 0.002).

The sensitivity of concentration techniques decreased from 100% for patients with bad compliance or absence of chemoprophylaxis to 91% for patients with full compliance to chemoprophylaxis but the difference was not statistically significant (p = 0.29).

The quality of adherence to chemoprophylaxis impacted the sensitivity of the RDT (p = 0.002). This last rate changed from 100% in the absence of chemoprophylaxis to 88% in the case of inconsistent chemoprophylaxis and to 60% when chemoprophylaxis was followed correctly.

Recourse to the PCR matched to 16 cases with a properly followed chemoprophylaxis, to 2 cases with omissions or prematurely stopped chemoprophylaxis, and to 2 cases with absence of chemoprophylaxis. For the last two cases requiring PCR testing, the quality of observance was not noted in the medical file.

Impact of Attack Type on Sensitivity of Tests

The impact of attack type on the sensitivity of parasitological tests is presented in Table 4. The sensitivity of thin blood smear increased from 83% in primary infection to 100% in relapse (p = 0.006). A blood parasite density with <200 parasites/μL was significantly associated with primary infection (p = 0.004).

Table 4. Impact of attack type on the sensitivity of parasitological tests and on the parasitemia (n = 89)
 Primary infection (n = 46)Relapse (n = 43)p Value
  1. RDT = rapid diagnostic test.

Positive thin blood smear (%)38 (83%)43 (100%)0.006
Parasitemia < 200 parasites/μL (%)19 (41%)6 (14%)0.004
Positive concentration techniques (%)41 (89%)43 (100%)0.06
Positive RDT (%)27 (59%)41 (95%)<0.001

Similarly, the difference of sensitivity for RDT was statistically significant between primary infection and relapse (p < 0.001). The increase in sensitivity for concentration techniques between primary infection and relapse remains nonsignificant (p = 0.06). Recourse to the PCR matched to 17 cases of primary infection (77%) and to 5 cases of relapse (23%).

Discussion

In this study, the predominance of cases arising in French Guyana is explained by the specific military role of the hospital and the resurgence of vivax cases among French military population in this overseas department, in relation with reinforced measures to combat illegal gold mining and with military exercises in the rainforest. Plasmodium vivax is the most frequently retrieved species in imported malaria in the French army since 1999.[16]

This work illustrates the difficulties faced by the biologists for imported P. vivax malaria diagnosis. False negatives are associated with delay in receiving adequate treatment. A Canadian study showed that the diagnosis was missed at presentation in 51% of cases of imported vivax malaria, and nonreference laboratories did correctly identify P. vivax in only 26% of cases.[17] These difficulties are often partially due to the low parasitemia.[4, 18]

In contrast to others series of imported P. vivax malaria,[4, 6] the originality of our study is the reference diagnosis method based on microscopy corrected by PCR, which was performed in ambiguous profiles.

Microscopy (thin and thick blood films) is the reference method recommended by the WHO but it requires training and experience. Blood smear examination allows positive diagnosis, species identification, and quantification of parasitemia. Its detection threshold is 200 parasitized red blood cells/μL.[19] In our series collected in an experienced laboratory, eight cases (10%) were not identified by blood smear. Low blood parasite density made it difficult or impossible to assess species identification in 16 other cases. Moreover, an examiner with little experience could miss a coinfection due to several plasmodial species or confuse P. vivax with P. ovale in patients returning from Africa.[18] Concentration techniques (thick blood smear and QBC Malaria Test) provide a lower threshold than blood smear from 10 to 20 parasites/μL but they do not allow a species identification. In our series, they provided a malaria diagnosis for four more patients (5%), leading to a sensitivity of 96%. Regarding RDTs, they did not detect any additional cases compared with the blood smear. On the contrary, considering the microscopy corrected by PCR as the reference method, the overall sensitivity of RDTs in this study was low (76%), in accordance with data reported in nonendemic setting in collection of stored samples: 66% to 87.5%.[20-22] Indeed, the main disadvantage of RDTs is their lack of sensitivity for the detection of non-falciparum species, whether they are based on aldolase or pLDH detection.[23] Of the 67 detecting P. vivax RDTs evaluated by the WHO, only 17 of them have a higher than 90% sensitivity for parasitemia up to 200 parasites/μL.[24] RDT detecting pLDH has been reported to be more efficient than the ones detecting aldolase,[24, 25] which is the reason for change of tests in our laboratory. More recently, the Clearview® malaria pLDH RDT has been reported to have an overall sensitivity of 90.9% for P. vivax detection, including samples with low parasitemia.[26] RDT implementation is fast and simple and the interpretation is easy. They may be of help to nonexperienced laboratories. However, these laboratories must be aware that a RDT alone is insufficient to exclude vivax malaria diagnosis.[23]

Molecular techniques allow the detection of some unidentified cases by conventional diagnostic techniques,[17, 27, 28] as demonstrated by four cases (5%) only detected by PCR in our series. The main contribution of PCR resides in its high negative predictive value, owing to an excellent sensitivity with a very low detection threshold around 0.01 parasitized erythrocyte/μL.[28] PCR is also interesting to redirect or confirm species identification: in 16 cases, blood smears failed to deliver a definitive species identification owing to a low level of parasitemia. On the other hand, molecular biology presents numerous limitations such as significant cost, lack of standardization, and unavailability in emergency,[28] explaining that this technique cannot be used as a first-intention diagnostic test.

To our knowledge, this is the first study that evaluated the impact of chemoprophylaxis and of attack type on the sensitivity of parasitological tests in imported vivax malaria.

The majority of patients who acquired malaria took no chemoprophylaxis, used an inappropriate regimen, or were not compliant.[1] However, some patients can develop malaria despite a properly followed chemoprophylaxis. The involved species are mostly P. vivax and P. ovale, suggesting that hypnozoites are not sensitive to prophylaxis regimens usually prescribed.[6] The latency period between return and clinical attack is longer in travelers who had been adherent to their chemoprophylaxis.[4, 29] Similarly, the parasite density is lower in patients who were taking chemoprophylaxis than in those who were not.[4] Adherence to chemoprophylaxis increases the difficulty of laboratory diagnosis in reducing the sensitivity of parasitological tools. A notion of properly followed chemoprophylaxis should evoke a possible false negative with conventional techniques and can be an argument to perform a PCR in case of an evocative clinical context. On the other hand, all conventional tools have a sensitivity of 100% in the absence of chemoprophylaxis.

The primary infection is associated with a lower parasitemia than observed in relapse. This result can be explained by the fact that patients with a relapse have taken no chemoprophylaxis for several weeks or months, whereas the prescribed medications have short half-lives. A false negative is possible in primary infection, even by combining microscopic techniques and RDT. In contrast, the microscopic techniques seem sufficient to exclude the diagnosis in cases of suspected relapse. However, caution is required and this result must be confirmed with further studies.

In conclusion, this work illustrates the difficulties encountered in vivax malaria diagnosis, especially in patients with chemoprophylaxis or primary infection due to a lower parasitemia. It underlines the lack of sensitivity of RDT for P. vivax and emphasizes the need for systematically combining various diagnosis tools.

Declaration of Interests

The authors state that they have no conflicts of interest.