6
Acta Tropica 125 (2013) 214–219 Contents lists available at SciVerse ScienceDirect Acta Tropica jo ur nal homep age : www.elsevier.com/locate /actatropica Field evaluation of rapid diagnostic tests for malaria in Yaounde, Cameroon Rachida Tahar a,b,1 , Collins Sayang b,c , Vincent Ngane Foumane b , Georges Soula c , Roger Moyou-Somo d , Jean Delmont c , Leonardo K. Basco a,b,,2 a Institut de Recherche pour le Développement (IRD), Unité Mixte de Recherche 198, Faculté de Médecine La Timone, Université de la Méditerranée, 13385 Marseille, France b Laboratoire de Recherche sur le Paludisme, Organisation de Coordination pour la lutte contre les Endémies en Afrique Centrale (OCEAC), B. P. 288, Yaoundé, Cameroon c Centre de Formation et de Recherche en Médecine et Santé Tropicale, Faculté de Médecine Nord, Boulevard Dramard, 13015 Marseille, France d Department of Infectious Diseases, Faculty of Medicine and Biomedical Sciences, University of Yaoundé I, Yaoundé, Cameroon a r t i c l e i n f o Article history: Received 6 December 2011 Received in revised form 29 September 2012 Accepted 9 October 2012 Available online 17 October 2012 Keywords: Malaria Plasmodium falciparum Plasmodium malariae Drug resistance Diagnosis Atovaquone a b s t r a c t Rapid diagnostic tests (RDTs) are affordable, alternative diagnostic tools. The present study aimed to evaluate RDTs available in Cameroon and compare their characteristics to follow the parasitological response of patients for 28 days. Malaria diagnosis was assessed in 179 febrile patients using conventional microscopy as the reference method. Parascreen detects both Plasmodium falciparum-specific histidine- rich protein 2 (Pf HRP-2) and Pan-specific plasmodial lactate dehydrogenase (pLDH) in all four human Plasmodium spp. Diaspot is based on the detection of Pf HRP-2. OptiMAL-IT (pLDH specific for P. falciparum and pLDH for all four human Plasmodium spp.) was assessed for comparison. The reliability of RDTs was evaluated by calculating the sensitivity, specificity, positive predictive value, negative predictive value, false-positive rate, false-negative rate, and likelihood ratio. The clinical outcome of 18 children treated with atovaquone–proguanil and followed for 28 days was evaluated using microscopy and RDTs. Of 179 samples, 133 (74.3%) were pure P. falciparum-positive smears, 4 (2.2%) pure P. malariae-positive smears, and 42 (23.5%) negative smears. Parascreen and Diaspot had high sensitivity (>92%) and positive predictive values (>94%). The specificities for Parascreen and Diaspot were 81.0% and 90.5%, respectively. The false-positive rates and the false-negative rates were 19.0% and 4.5% for Parascreen and 9.5% and 8.3% for Diaspot, respectively. Most false-negatives occurred in samples with low parasitaemia (<500 asexual parasites/L). The performance of RDTs was better at higher parasitaemia (>500 asexual parasites/L). Four pure P. malariae were only detected by the pan-Plasmodium bands of Parascreen and OptiMAL-IT. In blood samples from patients treated and followed-up for 28 days, HRP2-based RDTs remained positive in most samples until Day 28. Despite negative smears, OptiMAL-IT remained positive in several patients until Day 7 but was negative in all patients from Day 14 onwards. RDTs can improve the management of febrile patients. The validity, ease of use, and cost of HRP2-based tests were comparable. However, one of the current weaknesses of the RDT-based strategy using the tests available in Cameroon is inadequate sensitivity for low parasitaemia. In some cases, RDT results may require correct interpretation based on clinical history, clinical examination, and microscopic diagnosis. © 2012 Elsevier B.V. All rights reserved. Abbreviations: ACT, artemisinin-based combination therapy; HRP2, histidine- rich protein 2; pLDH, plasmodial lactate dehydrogenase; RDT, rapid diagnostic test; WBC, white blood cell. Corresponding author at: Institut de Recherche Biomédicale des Armées (IRBA), Boulevard Charles Livon, Parc du Pharo, 13262 Marseille cedex 07, France. Tel.: +33 491150149; fax: +33 491150164. E-mail address: [email protected] (L.K. Basco). 1 Present address: Unité Mixte de Recherche 216, Institut de Recherche pour le Développement (IRD) Université Paris Descartes, Laboratoire de Parasitologie, Faculté de Pharmacie, 4 avenue de l’Observatoire, 75270 Paris cedex 06, France. 2 Present address: Laboratoire de Parasitologie, Institut de Recherche Biomédicale des Armées (IRBA), Institut de Médecine Tropicale du Service de Santé des Armées (IMTSSA), Allée du médecin colonel Eugène Jamot, Parc du Pharo, BP60109, 13262 Marseille, France. 1. Introduction The attainment of the global goal to reduce morbidity and mortality due to malaria depends on prompt diagnosis and early treatment of patients (World Health Organization [WHO], 2000a). At present, a large majority of patients in Africa are diagnosed to have malaria on the basis of clinical signs and symp- toms. Presumptive clinical diagnosis is highly unreliable since the differential diagnosis of malaria includes many febrile ill- nesses of viral and bacterial origin. Moreover, as the majority of sub-Saharan African countries, including Cameroon, resort to artemisinin-based combination therapy (ACT), which are much more expensive than synthetic compounds like chloroquine and sulfadoxine–pyrimethamine and limited in quantity due to the necessity to extract artemisinin from Artemisia annua, a reliable 0001-706X/$ see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.actatropica.2012.10.002

Field evaluation of rapid diagnostic tests for malaria in Yaounde, Cameroon

Embed Size (px)

Citation preview

F

RRa

b

c

d

a

ARR2AA

KMPPDDA

rW

BT

lF

d(M

0h

Acta Tropica 125 (2013) 214– 219

Contents lists available at SciVerse ScienceDirect

Acta Tropica

jo ur nal homep age : www.elsev ier .com/ locate /ac ta t ropica

ield evaluation of rapid diagnostic tests for malaria in Yaounde, Cameroon

achida Tahara,b,1, Collins Sayangb,c, Vincent Ngane Foumaneb, Georges Soulac,oger Moyou-Somod, Jean Delmontc, Leonardo K. Bascoa,b,∗,2

Institut de Recherche pour le Développement (IRD), Unité Mixte de Recherche 198, Faculté de Médecine La Timone, Université de la Méditerranée, 13385 Marseille, FranceLaboratoire de Recherche sur le Paludisme, Organisation de Coordination pour la lutte contre les Endémies en Afrique Centrale (OCEAC), B. P. 288, Yaoundé, CameroonCentre de Formation et de Recherche en Médecine et Santé Tropicale, Faculté de Médecine Nord, Boulevard Dramard, 13015 Marseille, FranceDepartment of Infectious Diseases, Faculty of Medicine and Biomedical Sciences, University of Yaoundé I, Yaoundé, Cameroon

r t i c l e i n f o

rticle history:eceived 6 December 2011eceived in revised form9 September 2012ccepted 9 October 2012vailable online 17 October 2012

eywords:alaria

lasmodium falciparumlasmodium malariaerug resistanceiagnosistovaquone

a b s t r a c t

Rapid diagnostic tests (RDTs) are affordable, alternative diagnostic tools. The present study aimed toevaluate RDTs available in Cameroon and compare their characteristics to follow the parasitologicalresponse of patients for 28 days. Malaria diagnosis was assessed in 179 febrile patients using conventionalmicroscopy as the reference method. Parascreen detects both Plasmodium falciparum-specific histidine-rich protein 2 (Pf HRP-2) and Pan-specific plasmodial lactate dehydrogenase (pLDH) in all four humanPlasmodium spp. Diaspot is based on the detection of Pf HRP-2. OptiMAL-IT (pLDH specific for P. falciparumand pLDH for all four human Plasmodium spp.) was assessed for comparison. The reliability of RDTs wasevaluated by calculating the sensitivity, specificity, positive predictive value, negative predictive value,false-positive rate, false-negative rate, and likelihood ratio. The clinical outcome of 18 children treatedwith atovaquone–proguanil and followed for 28 days was evaluated using microscopy and RDTs. Of179 samples, 133 (74.3%) were pure P. falciparum-positive smears, 4 (2.2%) pure P. malariae-positivesmears, and 42 (23.5%) negative smears. Parascreen and Diaspot had high sensitivity (>92%) and positivepredictive values (>94%). The specificities for Parascreen and Diaspot were 81.0% and 90.5%, respectively.The false-positive rates and the false-negative rates were 19.0% and 4.5% for Parascreen and 9.5% and 8.3%for Diaspot, respectively. Most false-negatives occurred in samples with low parasitaemia (<500 asexualparasites/�L). The performance of RDTs was better at higher parasitaemia (>500 asexual parasites/�L).Four pure P. malariae were only detected by the pan-Plasmodium bands of Parascreen and OptiMAL-IT. Inblood samples from patients treated and followed-up for 28 days, HRP2-based RDTs remained positive

in most samples until Day 28. Despite negative smears, OptiMAL-IT remained positive in several patientsuntil Day 7 but was negative in all patients from Day 14 onwards. RDTs can improve the management offebrile patients. The validity, ease of use, and cost of HRP2-based tests were comparable. However, oneof the current weaknesses of the RDT-based strategy using the tests available in Cameroon is inadequatesensitivity for low parasitaemia. In some cases, RDT results may require correct interpretation based on

amin

clinical history, clinical ex

Abbreviations: ACT, artemisinin-based combination therapy; HRP2, histidine-ich protein 2; pLDH, plasmodial lactate dehydrogenase; RDT, rapid diagnostic test;

BC, white blood cell.∗ Corresponding author at: Institut de Recherche Biomédicale des Armées (IRBA),oulevard Charles Livon, Parc du Pharo, 13262 Marseille cedex 07, France.el.: +33 491150149; fax: +33 491150164.

E-mail address: [email protected] (L.K. Basco).1 Present address: Unité Mixte de Recherche 216, Institut de Recherche pour

e Développement (IRD) – Université Paris Descartes, Laboratoire de Parasitologie,aculté de Pharmacie, 4 avenue de l’Observatoire, 75270 Paris cedex 06, France.2 Present address: Laboratoire de Parasitologie, Institut de Recherche Biomédicalees Armées (IRBA), Institut de Médecine Tropicale du Service de Santé des ArméesIMTSSA), Allée du médecin colonel Eugène Jamot, Parc du Pharo, BP60109, 13262

arseille, France.

001-706X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.actatropica.2012.10.002

ation, and microscopic diagnosis.© 2012 Elsevier B.V. All rights reserved.

1. Introduction

The attainment of the global goal to reduce morbidity andmortality due to malaria depends on prompt diagnosis andearly treatment of patients (World Health Organization [WHO],2000a). At present, a large majority of patients in Africa arediagnosed to have malaria on the basis of clinical signs and symp-toms. Presumptive clinical diagnosis is highly unreliable sincethe differential diagnosis of malaria includes many febrile ill-nesses of viral and bacterial origin. Moreover, as the majorityof sub-Saharan African countries, including Cameroon, resort to

artemisinin-based combination therapy (ACT), which are muchmore expensive than synthetic compounds like chloroquine andsulfadoxine–pyrimethamine and limited in quantity due to thenecessity to extract artemisinin from Artemisia annua, a reliable

ropica

aa(

tiwtpwsldirmPorpRsi

tYiclcso

2

2

2

wawcpps82wasftwa1sm

2f

aTt

R. Tahar et al. / Acta T

nd rapid laboratory diagnosis of malaria will improve the costnd effectiveness of the management of malaria-infected patientsSayang et al., 2009).

Microscopic examination of Giemsa-stained blood smears ishe gold standard for malaria diagnosis. The procedure is labour-ntensive and requires at least 1 h (usually more in a health centre

ith numerous suspected cases of malaria) from blood collec-ion to results. Moreover, many health facilities in Africa do notossess a microscope, and even if there is a microscope in goodorking condition, well-trained and experienced microscopists are

carce. High-quality Giemsa stain and immersion oil may also beacking in peripheral health centres. Within this context, rapidiagnostic tests (RDTs) are affordable alternative diagnostic tools

n the absence of a microscope (WHO, 2000b; Long, 2009). RDTsequire a drop of fingerpricked capillary blood (5–20 �L) to detectalaria-specific proteins, some of which are specific to the genus

lasmodium (pan plasmodial lactate dehydrogenase, pLDH), whilethers are specifically expressed by a given species (e.g. histidine-ich protein 2 [HRP2] or pf pLDH by Plasmodium falciparum andv pLDH by P. vivax) by immunochromatographic methods. SomeDTs detect P. falciparum, with or without the other Plasmodiumpecies, and non-P. falciparum (i.e. P. vivax, P. ovale, and/or P. malar-ae).

Our previous study showed the effectiveness of the RDT-basedreatment with ACT over the presumptive treatment approach inaounde (Sayang et al., 2009). This survey also demonstrated the

mportance of the characteristics of RDTs. The present study wasonducted with the aim to evaluate two RDTs available throughocal commercial suppliers and one RDT imported from Europe. Weompared the diagnostic characteristics of these RDTs in a dispen-ary and compared them to follow-up the parasitological responsef patients for 28 days.

. Patients, materials and methods

.1. Patients

.1.1. Group 1, patients enrolled for RDT onlyAll febrile adults (no age limits) and children (or history of fever

ithin the past 48 h) attending the Nlongkak Catholic mission-ry dispensary, Yaounde, for presumptive uncomplicated malaria,ere enrolled in the study in 2008–2009 after a written informed

onsent. Giemsa-stained blood smear was prepared from finger-ricked capillary blood and examined under light microscope. Thearasite density was expressed as the number of asexual para-ites per �L of blood, assuming white blood cell (WBC) density of000/�L. In case of low parasitaemia (i.e. <10 asexual parasites per00 WBC), parasites were counted against 500 WBCs. A blood smearas considered to be negative in the absence of asexual parasites

gainst 500 WBCs. All smears were read by two experienced micro-copists, and the final parasite densities were the mean calculatedrom two counts. A third microscopist examined and determinedhe parasite density in case of >15% discordant results. Patientsith a positive blood smear were treated with a 3-day regimen of

rtesunate–amodiaquine (artesunate, 4 mg/kg/day + amodiaquine,0 mg/kg/day) as the first-line and artemether–lumefantrine asecond-line treatment, according to the current Cameroonian anti-alarial treatment guidelines.

.1.2. Group 2, patients enrolled for RDT evaluation for clinicalollow-up

Eighteen children aged less than 5 years were enrolled aftern informed written consent of the parents or legal guardians.he inclusion criteria as recommended by the 2003 WHO pro-ocol were as follows: rectal temperature ≥ 38.0 ◦C, P. falciparum

125 (2013) 214– 219 215

parasitaemia between 2000 and 200,000 asexual parasites/�L, easyaccess to the dispensary, absence of malnutrition and danger signsof severe and complicated falciparum malaria, and absence ofhistory of allergic reactions to antimalarial drugs (WHO, 2003).These 18 patients were treated with oral atovaquone–proguanil(atovaquone, 20 mg/kg/day, +proguanil, 8 mg/kg/day, once dailyfor 3 days) and followed-up for 28 days. Each dose was admin-istered under medical supervision. Fingerpricked capillary bloodwas obtained on Days 0, 2, 3, 7, 14, 21, and 28 for blood exam-ination and RDTs. The clinical outcomes were classified as earlytreatment failure, late clinical failure, late parasitological failure,or adequate clinical and parasitological response, according tothe 2003 WHO definitions (WHO, 2003). Patients who failed torespond to atovaquone–proguanil treatment were treated withartemether–lumefantrine. This study was reviewed and approvedby the Cameroonian national ethics committee and the Cameroo-nian Ministry of Public Health.

2.2. RDT

Two RDTs that are commercially available in Yaoundé,Cameroon during the study period were assessed. DiaSpot P. falcip-arum rapid test device (batch numbers MAL6090104, MAL6110020,MAL7050030; Acumen Diagnostics Inc., Livermore, CA) is a chro-matographic immunoassay for the qualitative detection of the P.falciparum HRP2 in whole blood within 15 min. This RDT does notdetect P. vivax, P. malariae, or P. ovale. Parascreen (batch no. 101098;Zephyr Biomedicals, Goa, India; distributed by International Dis-pensary Association Foundation, Amsterdam, The Netherlands)is an immunoassay that detects P. falciparum-specific HRP2 andPlasmodium genus-specific (P. falciparum, P. vivax, P. ovale, and P.malariae) pLDH within 15 min. In the presence of P. falciparum,either the ‘Pf’ HRP2 line alone (usually in the presence of low para-sitaemias) or two bands on the ‘Pf’ and ‘Pan’ lines may appear. Thelatter case indicates P. falciparum or mixed infection. If P. falciparumis absent but one of the three other Plasmodium species (or mixedinfections without P. falciparum) is present, the ‘Pan’ band is visible.

In addition to these two RDT devices that are avail-able in Cameroon, OptiMAL-IT (batch no. 46110.23.01; DiaMedAG, Cressier sur Morat, Switzerland) was ordered in Europefor comparison, mostly in patients who were treated withatovaquone–proguanil and followed for 28 days. This RDT was alsoassessed in blood samples with low parasitaemia ≤500 asexualparasites/�L of blood. OptiMAL-IT is an immunochromatographictest that detects P. falciparum-specific LDH (‘Pf’ band) and Plasmo-dium sp. LDH (‘P’ band for all 4 human Plasmodium spp. LDH species)using monoclonal antibodies within 20 min.

All three RDTs were performed according to the manufactur-ers’ instructions. Two- to three-page users’ guides provided by themanufacturers in French were distributed to the health person-nel, and test procedures were explained and demonstrated for eachRDT during a 1-h training session. The ease of application of eachRDT was evaluated during a feedback session. The interpretation ofresults was controlled by one or two experienced researchers.

2.3. Data interpretation and analysis

The gold standard was microscopic examination of Giemsa-stained thick blood smears. All RDTs assessed in the presentstudy have an internal control that ensures that the testis valid. Parascreen RDT was interpreted as positive if ‘Pf’(HRP2) or ‘Pan’ band (LDH) (or both) was visible. Likewise,

the result of OptiMAL-IT was positive if ‘Pf’ and/or ‘P’ band(s)were visible. Diaspot was interpreted as positive when asingle band was visible. The performance of RDTs, includ-ing 95% confidence intervals, was assessed by calculating the

216 R. Tahar et al. / Acta Tropica

Table 1Comparison of results obtained with rapid diagnostic tests (RDT).

RDT Microscopy

Positive P.falciparum

Positive pure P.malariae

Negative

DiaspotPositive 122 0 4Negative 11 4 38

ParascreenPositive 127 4 (‘Pan’ positive, ‘Pf’

negative)8

Negative 6 0 34OptiMAL-IT

Positive 15 1(‘Pan’ positive, ‘Pf’negative)

1

Negative 9 0 13

Asf

stspnpmIoufn

3

3

pfsf9prwn

lms4

TP

av

ll 179 blood samples were tested in parallel with Diaspot and Parascreen. Of theseamples, 39 were tested with OptiMAL-IT in samples with low parasitaemia. ‘Pf’, P.alciparum; ‘Pan’, Plasmodium spp.

ensitivity, specificity, positive predictive value, negative predic-ive value, false-positive rate (i.e. the number of false-positive

[positive test in a person without malaria] as a percentage of allatients without malaria according to the reference method), false-egative rate (i.e. the number of false-negatives [negative test in aerson who does have malaria] as a percentage of all patients withalaria), and likelihood ratio, i.e. sensitivity/(1 − specificity), using

nStat (GraphPad Software, Inc., San Diego, CA). Logarithmic valuesf parasitaemia were compared between groups by the Student’snpaired t-test. The coefficient of determination (r2) was calculatedrom the parametric Pearson correlation coefficient. Statistical sig-ificance was fixed at 0.05.

. Results

.1. Comparison of three RDTs in patient group 1

A total of 179 blood samples were collected to assess theerformance of the RDTs. There were 133 (74.3%) pure P.

alciparum-positive smears, 4 (2.2%) pure P. malariae-positivemears, and 42 (23.5%) negative smears (Table 1). Among P.alciparum-positive patients, the geometric mean parasitaemia was420 asexual parasites/�L of blood (range, 33–230,000 asexualarasites/�L). The parasitaemia of four pure P. malariae samplesanged from 1130 to 7650 asexual parasites/�L. P. ovale and P. vivaxere not detected by microscopy in the present study. P. vivax hasever been reported from Cameroon.

Parascreen and Diaspot were highly sensitive and had simi-

ar positive predictive values (94.1–96.8%) (Table 2, excluding 4 P.

alariae samples). Eleven false-negatives with Diaspot occurred inamples with 33, 45, 48, 61, 126 (+gametocytes), 155, 212, 250,30, 3890, and 5870 asexual parasites/�L; five of these samples

able 2erformance of rapid diagnostic tests (RDT) for P. falciparum malaria in Yaoundé, Camero

RDT Na Sensitivity Specificity PPV

(95% CI) (95% CI) (95% CI)

Diaspot 175 91.7(85.7–95.8)

90.5(77.4–97.4)

96.8(92.1–99.1)

Parascreen 175 95.5(90.4–98.3)

81.0(65.9–91.4)

94.1(88.6–97.4)

OptiMAL-IT 38 62.5(40.6–81.2)

92.9(66.1–99.8)

93.8(69.8–99.8)

a For Parascreen and OptiMAL-IT, only the results of the P. falciparum-specific bands assessed in samples with low parasitaemia, which explains its low sensitivity and high falsalue; NPV, negative predictive value; 95% CI, 95% confidence interval.

125 (2013) 214– 219

yielded a positive ‘P’ band and a negative ‘Pan’ band with Parascreen(33, 126 [presence of gametocytes], 250, 3890, and 5870 asexualparasites/�L), while six showed negative ‘P’ and ‘Pan’ bands withParascreen (45, 48, 61, 155, 212, 430 asexual parasites/�L). Threefalse-positives were observed with both Diaspot and Parascreenin the same samples. Five additional false-positives were obtainedwith Parascreen (positive ‘P’ band and negative ‘Pan’ band), butnot with Diaspot. One of these five patients with a false-positiveresult started self-medication with artesunate–amodiaquine 24 hbefore consultation. Another patient had a negative smear for asex-ual parasites but carried P. falciparum gametocytes (Diaspot andOptiMAL-IT negative). For the remaining patients, the false-positiveresult was unexplained. Three patients carried P. falciparum game-tocytes and asexual parasites: (i) 668 asexual parasites/�L, Diaspotpositive, Parascreen ‘Pan’ and ‘Pf’ bands positive, OptiMAL-IT notdone; (ii) 126 asexual parasites/�L, Diaspot negative, Parascreen‘Pan’ band negative, Parascreen ‘Pf’ band positive, OptiMAL-IT neg-ative; and (iii) 92 asexual parasites/�L, Diaspot positive, Parascreen‘Pan’ band negative, Parascreen ‘Pf’ band positive, and OptiMAL-ITnegative.

The ‘Pan’ band and the ‘P’ band of Parascreen RDT were dis-cordant, i.e. negative ‘Pan’ band and positive ‘P’ band, in 8 of42 smear-negative samples (8 false-positives) and 30 of 133P. falciparum-positive samples. Blood samples with discordantresults had a lower parasitaemia (geometric mean, 783 asex-ual parasites/�L; range, 33–54,000 asexual parasites/�L; P < 0.05),compared to P. falciparum-positive samples with concordant pos-itive ‘Pan’ and ‘P’ band results (geometric mean, 28,100 asexualparasites/�L; range, 668–230,000 asexual parasites/�L). Four pureP. malariae were only detected by the ‘Pan’ band of Parascreen.Therefore, Parascreen (also OptiMAL-IT in one sample tested) cor-rectly detected the samples infected with P. malariae with a positive‘Pan’ band and a negative ‘Pf’ band, while Diaspot did not, andcannot detect, P. malariae.

OptiMAL-IT was assessed in 26 random samples (11 P.falciparum-positive samples [8000–118,000 asexual parasites/�L],1 pure P. malariae-positive sample, and 14 smear-negative sam-ples) and 13 samples with low parasitaemia (range, 38–430 asexualparasites/�L). OptiMAL-IT yielded 9 false negative results and 1false positive with P. falciparum samples. False negatives occurredin samples with low parasitaemias (<500 asexual parasites/�L): 38,45, 48, 61, 92 (with P. falciparum gametocytes), 126 (with P. falcip-arum gametocytes), 155, 212, and 430 asexual parasites/�L. Thethreshold parasite density for OptiMAL-IT-positive results couldnot be determined, as the RDT was positive in few samples withlow parasitaemias (75, 92, 205, 260 asexual parasites/�L). For all 38P. falciparum-positive samples, ‘Pf’ and ‘P’ bands yielded the sameresults, i.e. either both positive or both negative.

The sensitivity of each of the assessed RDTs in relation to P.falciparum parasite density is summarised in Table 3. The sensi-tivity was low (<67%) for blood samples with low parasitaemia(<200 asexual parasites/�L and <500 asexual parasites/�L) with

on.

NPV False-positiverate

False-negativerate

Likelihoodrate

(95% CI)

77.6(63.4–88.2)

9.5 8.3 9.6

85.0(70.2–94.3)

19.0 4.5 5.0

59.1(36.3–79.3)

7.1 37.5 8.8

re presented here. Four additional samples were pure P. malariae. OptiMAL-IT wase-negative rate in the present study. N, number of samples; PPV, positive predictive

R. Tahar et al. / Acta Tropica

Table 3Plasmodium falciparum parasite density and sensitivity of rapid diagnostic tests formalaria.

Parasite density(asexualparasites/�L)

Na Sensitivity

Diaspot Parascreen OptiMAL-IT

<200 12 (9) 50.0 66.7 22.2<500 18 (13) 50.0 66.7 30.8500–5000 27 (0) 96.3 100 –>5000 88 (11) 98.9 100 100

a

op

aapP

3e

ca1pbAspwou

hrpws‘wiO

Fdibonh

N, number of blood samples tested with Diaspot and Parascreen (the numbersf samples tested with OptiMAL-IT in parentheses). Samples with <500 asexualarasites/�L include those with <200 asexual parasites/�L.

ll 3 RDTs. Parascreen and Diaspot were highly sensitive, i.e. 100%nd 96.3%, respectively, for blood samples with >500 asexualarasites/�L. The performance of OptiMAL-IT, as well as that ofarascreen and Diaspot, was better at higher parasitaemia.

.2. RDTs as a complementary tool to evaluate therapeuticfficacy

Diaspot, Parascreen, and OptiMAL-IT were used as aomplementary tool to monitor the therapeutic efficacy oftovaquone–proguanil during the 28-day follow-up period in8 patients (Fig. 1). On Day 0 (before treatment), P. falciparumarasitaemia ranged from 3840 to 192,000 asexual parasites/�L oflood (geometric mean parasitaemia, 40,200 asexual parasites/�L).ll three RDTs were positive before treatment. For unknown rea-ons, the ‘Pan’ band, but not the ‘Pf’ band, of Parascreen in 2atients (parasitaemia, 3840 and 120,000 asexual parasites/�L)as negative before treatment, and in all subsequent samples

btained from these 2 patients, the ‘Pan’ band remained negativentil Day 28.

On Day 3, 15 of 18 (83%) patients had a negative smear, and 3ad low parasitaemia (65–866 asexual parasites/�L). OptiMAL-ITemained positive in 10 patients (8 smear negative and 2 smearositive), but was negative in 8 patients, including one patientith 866 asexual parasites/�L. On Day 7, all patients had a negative

mear. Diaspot remained positive in all but one patient, ParascreenP’ band was positive in all patients, and Parascreen ‘Pan’ band

as negative in all patients. On Day 7, OptiMAL-IT remained pos-tive in 5 smear-negative patients. From Day 14, only Diaspot andptiMAL-IT were evaluated. All 18 patients had negative smears on

ig. 1. Number of patients who were positive with microscopy and different rapidiagnostic tests before (Day 0) and after treatment with atovaquone–proguanil dur-

ng a 28-day follow-up period. Black bars, microscopy; hatched bars, Diaspot; whitears, OptiMAL-IT (Pf band); double hatched bars, Parascreen (Pan band). The Pf bandf Parascreen was positive in all patients on Day 0, Day 3, and Day 7. Parascreen wasot evaluated on Day 14, Day 21, and Day 28, except on Day 28 in one patient whoad a recrudescent malaria.

125 (2013) 214– 219 217

Day 14 and Day 21, and 17 had negative smears on Day 28. Diaspotwas still positive in 17 of 18 patients on Day 14, 10 of 18 patientson Day 21, and 7 of 17 patients on Day 28. With OptiMAL-IT, all 5patients who had a positive result on Day 7 became negative on Day14 and Day 21. One patient had a late parasitological failure on Day28, with a parasite density of 2420 asexual parasites/�L. Diaspot,Parascreen, and OptiMAL-IT were positive for this patient.

4. Discussion

This survey assessed the comparative validity of RDTs availablein Cameroon. Results lay down the needs for a procurement chainanalysis of RDT and its implementation in the field. The presentstudy was conducted in low malaria transmission settings wherea majority of patients infected with malaria parasites are symp-tomatic due to a relatively low level of acquired immunity and bothadults and children attend health facilities with malaria-associatedsymptoms throughout the year. Entomological infection rate variesbetween 0 and 33 infective mosquito bites per year and per person,depending on the district (Fondjo et al., 1992; Manga et al., 1992,1993). In urban malaria, asymptomatic carriage is rare, requiringspecific antimalarial treatment when blood smear is positive. More-over, it is known that the majority (about 95%) of malaria infectionsare due to P. falciparum. Within this context, RDTs with a highsensitivity and low false-negative rate are required. The abilityto distinguish between P. falciparum and non-falciparum species,with either a combination of HRP2 and pLDH, as in the case ofParascreen, or a combination of pan-Plasmodium and P. falciparum-specific pLDH, is less important, as all cases of malarial infectionsare treated with ACT in Cameroon.

The high positive predictive values of RDT (Diaspot andParascreen) reflect their sensitivity and partly depend on malariaprevalence. In Yaoundé, where malaria prevalence among febrilechildren and adults is relatively high, these RDTs proved to beuseful screening tests. Based on the likelihood rate, which esti-mates the probability for a patient with a positive test is likelyto be malaria-infected, as compared to a patient with a negativetest, HRP2-based Diaspot showed the best performance, followedby Parascreen. The results obtained with OptiMAL-IT cannot becompared with those of Diaspot and Parascreen due to a muchsmaller sample tested with OptiMAL-IT. Moreover, in the presentstudy, OptiMAL-IT was preferentially assessed in samples knownto have low parasitaemias. Nonetheless, the present study confirmsthe higher sensitivity of HRP2-based RDTs for low parasitaemias,as compared with pLDH-based detection, as supported by theobservation of more discordance between ‘Pf’ and ‘Pan’ bands onParascreen in samples with low parasite densities. These resultsare in agreement with previous studies (Hopkins et al., 2007, 2008;Rakotonirina et al., 2008; Ashley et al., 2009).

According to the manufacturer, Diaspot yields a 100% sensitivity,100% specificity, 100% positive predictive value, 0% false-positiverate, and 0% false-negative rate (n = 332 samples). As with Diaspot,the manufacturer of Parascreen claims 100% sensitivity, 100% speci-ficity, 100% positive predictive value, 0% false-positive rate, and 0%false-negative rate (n = 226, of which 16 P. falciparum-positive bymicroscopy). Neither of the manufacturers mentions the minimaldensity required for a positive test. In the present study, the perfor-mance of these RDTs was less than what has been claimed by themanufacturers.

In the present study, few samples were tested with OptiMAL-IT. This RDT was considered to be of less importance because

of the following reasons: (i) Parascreen incorporates both P.falciparum-specific HRP2 and Plasmodium genus-specific pan-pLDH, (ii) OptiMAL-IT is more expensive than other RDTs and (iii)is not available locally, and (iv) many other studies have evaluated

2 ropica

OaRettiAanadfsc

mnhawaud5woMsw2dasbwst

umRtnpd2espSsennfbh

tbtbrwt

18 R. Tahar et al. / Acta T

ptiMAL-IT elsewhere in Africa and Madagascar, both in earliernd recent years (Cooke et al., 1999; Mankhambo et al., 2002;akotonirina et al., 2008; Valéa et al., 2009; WHO, 2009; Ansaht al., 2010). Only viable parasites are thought to yield a posi-ive OptiMAL-IT test. The manufacturer claims that a successfulreatment yields a negative test within 2–4 days and that a pos-tive test 5–7 days after treatment may indicate resistant parasite.lthough the number of patients followed with both microscopynd OptiMAL-IT is limited, data presented in the present study doot support the use of either HRP2-based or pLDH-based RDTs as

complementary tool to evaluate therapeutic response within 7ays post-treatment. At present, RDTs are not recommended forollow-up of treatment. Microscopic examination of blood smears istill indispensable for evaluation of therapeutic efficacy and correctlassification of clinical and parasitological outcomes.

One of the major confounding factors in the field is self-edication prior to medical consultation, which may lead to a

egative smear and a positive HRP2-based RDT. The present studyas confirmed the persistence of HRP2 antigen for at least 1 monthnd detection of pLDH up to 7 days after a successful treatmentith highly effective drug combination. False-positive RDTs in such

context may result in overprescription of antimalarial drugsnless a complete and accurate history of previous antimalarialrug consumption is recorded. In our experience, approximately0% of febrile patients in Yaoundé resort to self-medicationith “traditional medicine,” quinine, sulfadoxine–pyrimethamine,

r, more recently, artesunate–amodiaquine before consultation.any of these self-administered drugs, in particular quinine and

ulfadoxine–pyrimethamine, are purchased from informal outlets,hich are known to supply low-quality or counterfeit drugs (Basco,

004). Such drugs may influence microscopy and RDT results,epending on the dose and drug quality. False-positive RDT maylso result from a misdiagnosis by microscopy if RDT is more sen-itive than microscopy (Bell et al., 2005). Diagnosis by PCR maye able to correct such a discrepancy. However, the present studyas designed to compare RDT results to microscopy as the gold

tandard. Moreover, PCR diagnosis is currently not an appropriateechnology in the field in most sub-Saharan African countries.

False-negative RDTs have a more serious impact on individ-al patients as they may not receive adequate treatment unlessicroscopy is performed. The major known cause of false-negative

DTs is low parasitaemia. Higher sensitive RDTs will be requiredo detect malaria parasites in these cases. Another cause of false-egative HRP2-based RDTs can be non-expression of the HRP2rotein and therefore non-detection by HRP2-based RDTs, asescribed in the South American Amazon region (Gamboa et al.,010). HRP2 sequence variation was previously thought to influ-nce RDT sensitivity but was not fully supported by subsequenttudies (Baker et al., 2005, 2010). The extent of hrp2 gene polymor-hism in Cameroonian P. falciparum isolates is currently unknown.o far, hrp2 gene sequence of only 4 Cameroonian P. falciparumtrains has been determined (Baker et al., 2005, 2010), but thexistence of hrp2- and/or hrp3-deleted P. falciparum strains hasever been reported in this country. Another possible cause of false-egative HRP2-based RDT results is the prozone effect (defined as

alse-negative immunological tests due to high antigen [or anti-ody] concentrations) that may be observed in blood samples withyperparasitaemia (Gillet et al., 2009, 2011).

All three RDTs assessed in the present study are comparable inerms of ease of application. Diaspot has a single well into whichoth the blood sample and lysis buffer are placed. Parascreen haswo wells, one for the blood sample and the other for lysis buffer,

ut once they are placed correctly, there is no further handlingequired. Errors with the use of these wells are unlikely to occurith Parascreen after a short training session. As for OptiMAL-IT,

here are two wells; blood sample and lysis buffer are mixed in the

125 (2013) 214– 219

first well and the stick is transferred to the second well containinglysis buffer after 10 min. Therefore, OptiMAL-IT requires an extrahandling step. Diaspot was purchased for 1.14 euros per cassette atthe time of the study. Parascreen cost was 1.17 euros (29.25 eurosor 38.00 USD for 25 tests) per cassette without shipping and cus-toms fees. OptiMAL-IT costs 3 times more (about 3.00 euros percassette), without taking into consideration shipping and customsfees. Based on the results of our studies, the Cameroonian Ministryof Public Health plans to include a RDT as part of the managementof febrile illnesses. However, more studies on RDTs are necessaryto select the most appropriate RDT and assess its usefulness indifferent epidemiological situations in the country.

5. Conclusion

The present study showed that Diaspot and Parascreen are use-ful laboratory tools in a dispensary in Yaounde, Cameroon. Theirperformance, ease of use, and cost were comparable. However,Parascreen has the advantage of being previously evaluated inEthiopia, Peruvian Amazon, and India, and its performance hasbeen published (Endeshaw et al., 2008, 2010; Ashton et al., 2010;Bendezu et al., 2010; Singh et al., 2010). Diaspot has not been eval-uated elsewhere due to unavailability in other countries. All threeRDTs, in particular pLDH-based RDT, performed poorly for sampleswith low (<500 asexual parasites/�L) parasite density. The resultsof the present study suggest that HRP2-based RDTs are not usefulto follow therapeutic response, while pLDH-based RDT may have alimited role beyond 7 days post-treatment. However, the generalconclusions of the present study are only applicable to the RDTsthat were available in Cameroon and assessed in the field. RDTsare important, cost-effective laboratory tools that improve patientmanagement and avoid presumptive treatment based only on clin-ical symptoms, in particular in remote areas where microscopy isnot available or is of poor quality. There are persistent technicalproblems (inadequate sensitivity for low parasitaemia, HRP2 anti-gen variation, and low [or absence of] HRP2 expression) that needto be overcome. Clinical acumen and microscopic diagnosis are stillneeded while waiting for highly sensitive RDTs.

Competing interest

The authors declare that they have no competing interest.

Acknowledgements

The authors thank the patients for their participation in thestudy and the health staff of Nlongkak Catholic missionary dis-pensary for recruiting febrile patients. This work was supported bythe European Union (Redox Antimalarial Drug Discovery, READ-UP, FP6-2004-LSH-2004-2.3.0-7, STREP no. 018602), Service deCoopération et d’Actions Culturelles of the French Embassy inYaoundé, and French Agence Nationale de la Recherche (projectRES-ATQ, ANR-08-MIE-024). CS was supported by a Ph.D. grant ofthe Agence Universitaire de la Francophonie.

References

Ansah, E.K., Narh-Bana, S., Epokor, M., Akanpigbiam, S., Quartey, A.A., Gyapong, J.,Whitty, C.J.M., 2010. Rapid testing for malaria in settings where microscopyis available and peripheral clinics where only presumptive treatment is avail-able: a randomised controlled trial in Ghana. British Medical Journal 340, c930,http://dx.doi.org/10.1136/bmj.c930.

Ashley, E.A., Touabi, M., Ahrer, M., Hutagalung, R., Htun, K., Luchavez, J., Dureza,C., Proux, S., Leimanis, M., Lwin, M.M., Koscalova, A., Comte, E., Hamade, P.,Page, A.L., Nosten, F., Guerin, P.J., 2009. Evaluation of three parasite lactatedehydrogenase-based rapid diagnostic tests for the diagnosis of falciparum andvivax malaria. Malaria Journal 8, 241.

ropica

A

B

B

B

B

B

C

E

E

F

G

G

R. Tahar et al. / Acta T

shton, R.A., Kefyalew, T., Tesfaye, G., Counihan, H., Yadeta, D., Cundill, B., Reithinger,R., Kolaczinski, J.H., 2010. Performance of three multi-species rapid diagnostictests for diagnosis of Plasmodium falciparum and Plasmodium vivax malaria inOromia Regional State, Ethiopia. Malaria Journal 9, 297.

aker, J., Ho, M.F., Pelecanos, A., Gatton, M., Chen, N., Abdullah, S., Albertini, A., Ariey,F., Barnwell, J., Bell, D., Cunningham, J., Djalle, D., Echeverry, D.F., Gamboa, D., Hii,J., Kyaw, M.P., Luchavez, J., Membi, C., Ménard, D., Murillo, C., Nhem, S., Ogutu, B.,Onyor, P., Oyibo, W., Wang, S.Q., McCarthy, J., Cheng, Q., 2010. Global sequencevariation in the histidine-rich proteins 2 and 3 of Plasmodium falciparum: impli-cations for the performance of malaria rapid diagnostic tests. Malaria Journal 9,129.

aker, J., McCarthy, J., Gatton, M., Kyle, D.E., Belizario, V., Luchavez, J., Bell, D., Cheng,Q., 2005. Genetic diversity of Plasmodium falciparum histidine-rich protein 2(PfHRP2) and its effect on the performance of PfHRP2-based rapid diagnostictests. Journal of Infectious Diseases 192, 870–877.

asco, L.K., 2004. Molecular epidemiology of malaria in Cameroon. XIX. Qualityof antimalarial drugs used for self-medication. American Journal of TropicalMedicine and Hygiene 70, 245–250.

ell, D.R., Wilson, D.W., Martin, L.B., 2005. False-positive results of a Plasmodiumfalciparum histidine-rich protein 2-detecting malaria rapid diagnostic test due tohigh sensitivity in a community with fluctuating low parasite density. AmericanJournal of Tropical Medicine and Hygiene 73, 199–203.

endezu, J., Rosas, A., Grande, T., Rodriguez, H., Llanos-Cuentas, A., Escobedo, J.,Gamboa, D., 2010. Field evaluation of a rapid diagnostic test (ParascreenTM) formalaria diagnosis in the Peruvian amazon. Malaria Journal 9, 154.

ooke, A.H., Chiodini, P.L., Doherty, T., Moody, A.H., Ries, J., Pinder, M., 1999.Comparison of a parasite lactate dehydrogenase-based immunochromato-graphic antigen detection assay (OptiMAL®) with microscopy for the detectionof malaria parasites in human blood samples. American Journal of TropicalMedicine and Hygiene 60, 173–176.

ndeshaw, T., Gebre, T., Ngondi, J., Graves, P.M., Shargie, E.B., Ejigsemahu, Y., Ayele,B., Yohannes, G., Teferi, T., Messele, A., Zerihun, M., Genet, A., Mosher, A.W.,Emerson, P.M., Richards, F.O., 2008. Evaluation of light microscopy and rapiddiagnostic test for the detection of malaria under operational field conditions:a household survey in Ethiopia. Malaria Journal 7, 118.

ndeshaw, T., Graves, P.M., Shargie, E.B., Gebre, T., Ayele, B., Yohannes, G., Zerihun,M., Genet, A., Melak, B., Kebede, A., Jima, D., Tadesse, Z., Ngondi, J., Mosher, A.W.,Richards, F.O., Emerson, P.M., 2010. Comparison of Parascreen Pan/Pf, ParacheckPf and light microscopy for detection of malaria among febrile patients, North-west Ethiopia. Transactions of the Royal Society of Tropical Medicine andHygiene 104, 467–474.

ondjo, E., Robert, V., Le Goff, G., Toto, J.C., Carnevale, P., 1992. Le paludisme urbain àYaoundé (Cameroun). 2. Etude entomologique dans deux quartiers peu urban-isés. Bulletin de la Societe de Pathologie Exotique 85, 57–63.

amboa, D., Ho, M.F., Bendezu, J., Torres, K., Chiodini, P.L., Barnwell, J.W., Incar-dona, S., Perkins, M., Bell, D., McCarthy, J., Cheng, Q., 2010. A large proportion

of P. falciparum isolates in the Amazon region of Peru lack pfhrp2 and pfhrp3:implications for malaria rapid diagnostic tests. PLoS ONE 5, e8091.

illet, P., Mori, M., Van Esbroeck, M., Van den Ende, J., Jacobs, J., 2009. Assess-ment of the prozone effect in malaria rapid diagnostic tests. Malaria Journal 8,271.

125 (2013) 214– 219 219

Gillet, P., Scheirlinck, A., Stokx, J., De Weggheleire, A., Chauque, H.S., Canhanga,O.D.J.V., Tadeu, B.T., Mosse, C.D.D., Tiago, A., Mabunda, S., Bruggeman, C., Bot-tieau, E., Jacobs, J., 2011. Prozone in malaria rapid diagnostic tests: how manycases are missed? Malaria Journal 10, 166.

Hopkins, H., Bebell, L., Kambale, W., Dokomajilar, C., Rosenthal, P.J., Dorsey, G., 2008.Rapid diagnostic tests for malaria at sites of varying transmission intensity inUganda. Journal of Infectious Diseases 197, 510–518.

Hopkins, H., Kambale, W., Kamya, M.R., Staedke, S.G., Dorsey, G., Rosenthal, P.J., 2007.Comparison of HRP2- and pLDH-based rapid diagnostic tests for malaria withlongitudinal follow-up in Kampala, Uganda. The American Journal of TropicalMedicine and Hygiene 76, 1092–1097.

Long, E.G., 2009. Requirements for diagnosis of malaria at different levels of the lab-oratory network in Africa. American Journal of Clinical Pathology 131, 858–860.

Manga, L., Fondjo, E., Carnevale, P., Robert, V., 1993. Importance of low dispersion ofAnopheles gambiae (Diptera: Culicidae) on malaria transmission in hilly townsin south Cameroon. Journal of Medical Entomology 30, 936–938.

Manga, L., Robert, V., Messi, J., Desfontaine, M., Carnevale, P., 1992. Le paludismeurbain à Yaoundé, Cameroun. 1. Etude entomologique dans deux quartiers cen-traux. Memoires de la Societe Royale belge d’Entomologie 35, 155–162.

Mankhambo, L., Kanjala, M., Rudman, S., Lema, V.M., Rogerson, S.J., 2002. Evaluationof the OptiMAL rapid antigen test and species-specific PCR to detect placentalPlasmodium falciparum infection at delivery. Journal of Clinical Microbiology 40,155–158.

Rakotonirina, H., Barnadas, C., Raherijafy, R., Adrianantenaina, H., Ratsimbasoa,A., Randrianasolo, L., Jahevitra, M., Andriantsoanirina, V., Ménard, D., 2008.Accuracy and reliability of malaria diagnostic techniques for guiding febrile out-patient treatment in malaria-endemic countries. American Journal of TropicalMedicine and Hygiene 78, 217–221.

Sayang, C., Soula, G., Tahar, R., Basco, L.K., Gazin, P., Moyou-Somo, R., Delmont, J.,2009. Use of a histidine-rich protein 2-based rapid diagnostic test for malaria byhealth personnel during routine consultation of febrile outpatients in a periph-eral health facility in Yaounde, Cameroon. American Journal of Tropical Medicineand Hygiene 81, 343–347.

Singh, N., Shukla, M.M., Shukla, M.K., Mehra, R.K., Sharma, S., Bharti, P.K., Singh, M.P.,Singh, A., Gunasekar, A., 2010. Field and laboratory comparative evaluation ofrapid malaria diagnostic tests versus traditional and molecular techniques inIndia. Malaria Journal 9, 191.

Valéa, I., Tinto, H., Nikiema, M., Yamuah, L., Rouamba, N., Drabo, M., Guiguemde, R.T.,d’Alessandro, U., 2009. Performance of OptiMAL-IT compared to microscopy, formalaria detection in Burkina Faso. Tropical Medicine and International Health14, 338–340.

WHO, 2000. WHO expert committee on malaria. WHO Technical Report Series 892,twentieth report. Geneva, Switzerland, World Health Organization.

WHO, 2000b. Malaria Diagnosis. New Perspectives. World Health Organization,Geneva, Switzerland, WHO/CDS/RBM/2000.14.

WHO, 2003. Assessment and Monitoring of Antimalarial Drug Efficacy for the

Treatment of Uncomplicated Falciparum Malaria. World Health Organization,Geneva, Switzerland, WHO/HTM/RBM/2003.50.

WHO, 2009. Malaria Rapid Diagnostic Test Performance. Results of WHO ProductTesting of Malaria RDTs: Round 1 (2008). World Health Organization, Geneva,Switzerland.