Tropical Medicine and International Health volume no pp 304–308 april 2002 Comparison of three antigen detection methods for diagnosis and therapeutic monitoring of malaria: a field study from southern Vietnam Nguyen Mai Huong1, Timothy M E Davis2,3, Sean Hewitt2, Nguyen Van Huong1, Tran Thi Uyen1, Doan Hanh Nhan1 and Le Dinh Cong1 National Institute of Malariology, Parasitology and Entomology, Hanoi, Vietnam Vietnam–Australia Malaria Control Project, Ministry of Health, Hanoi, Vietnam Department of Medicine, University of Western Australia, Fremantle Hospital, Fremantle, Australia Summary To compare the sensitivity, specificity and post-treatment persistence of three commonly used rapid antigen detection methods M E T H O D We studied 252 Vietnamese patients aged from to 60 years, 157 with falciparum and 95 with vivax malaria and 160 healthy volunteers An initial blood sample was taken for microscopy, and OptiMALÒ, immunochromatographic test (ICT) malaria P.f./P.v.Ò and Paracheck-PfÒ tests Patients with falciparum malaria were treated with an artesunate-based combination regimen and those with vivax malaria received chloroquine Eighty-seven patients with falciparum malaria who were initially positive for one of the antigen tests and who remained blood smear-negative underwent follow-up testing over 28 days Ò R E S U L T S Paracheck-Pf was the most sensitive test for Plasmodium falciparum (95.8% vs 82.6% for Ò ICT malaria P.f./P.v and 49.7% for OptiMALÒ) Specificities were all 100% For vivax malaria, OptiMALÒ performed better than ICT malaria P.f./P.v.Ò (sensitivities 73.7% and 20.0%, respectively), with 100% specificity in both cases All tests had low sensitivities (£ 75.0%) at parasitaemias < 1000/ll regardless of malaria species During follow-up, Paracheck-PfÒ remained positive in the greatest proportion of patients, especially at higher parasitaemias (> 10 000/ll) Residual OptiMALÒ positivity occurred only in a relatively small proportion of patients (< 10%) with parasitaemias > 10 000/ll during the first weeks after treatment Ò C O N C L U S I O N S Although microscopy remains the gold standard for malaria diagnosis, Paracheck-Pf may prove a useful adjunctive test in uncomplicated falciparum malaria in southern Vietnam OptiMALÒ had the lowest sensitivity for P falciparum but it might have a use in the diagnosis of vivax malaria and perhaps to monitor efficacy of treatment for falciparum malaria where microscopy is unavailable OBJECTIVES keywords falciparum malaria, vivax malaria, antigen detection, diagnosis, ICT malaria P.f./P.v.Ò, Paracheck-PfÒ, OptiMALÒ correspondence Timothy M E Davis, Department of Medicine, University of Western Australia, Fremantle Hospital, PO Box 480, Fremantle, Western Australia 6959, Australia Fax: +00 618 9431 2977; E-mail: tdavis@cyllene.uwa.edu.au Introduction Although examination of a blood smear by microscopy remains the gold standard for malaria diagnosis, antigen detection methods have been developed for situations in which reliable microscopy may not be available In the case 304 of Plasmodium falciparum, these new rapid methods are based on detection of histidine-rich protein-2 (HRP-2; ICT malaria P.f.Ò, ParaSight-FÒ, Paracheck-PfÒ) or parasite lactate dehydrogenase (pLDH; OptiMALÒ) Species-specific pLDH isoforms have been used to develop a test for Plasmodium vivax (OptiMALỊ) Plasmodium vivax can ª 2002 Blackwell Science Ltd Tropical Medicine and International Health volume no pp 304–308 april 2002 N M Huong et al Antigen detection in malaria diagnosis also be detected through positive antibodies against ‘panmalarial’ antigen in the absence of those against HRP-2 (ICT Malaria P.f./P.v.Ò) The sensitivity and specificity of each of these tests have been assessed in a range of clinical situations As might be expected from their common antigenic target, the performance characteristics of ICT malaria P.f.Ò, ParaSight-FÒ and Paracheck-PfÒ are similar Although the overall sensitivity and specificity of each of these three tests is high (usually >90%), their sensitivity falls off at parasite densities < 350/ll (Lema et al 1999; Ricci et al 2000; Singh et al 2000; Proux et al 2001) In the case of OptiMALÒ for falciparum malaria, results of field studies appear more variable with an overall sensitivity of between 60% (Fryauff et al 2000) and 96% (Moody et al 2000) Nevertheless, some of this variability may also be explained by the relatively poor performance of OptiMALÒ at low parasitaemias (< 500/ll) (Iqbal et al 1999; Fryauff et al 2000; Ricci et al 2000) The ICT malaria P.f./P.v.Ò has a reported sensitivity of between 45% (Cho et al 2001) and 95% (Tjitra et al 1999) for vivax malaria, but its ability to detect a parasitaemia at < 1500/ll has been questioned (Singh et al 2000) The sensitivity of OptiMALÒ for P vivax is between 70% (Fryauff et al 2000; Cho et al 2001) and 96% (Moody et al 2000) but there have been no reports of its relationship with parasitaemia It has been suggested that the disappearance of pLDH that accompanies parasite clearance might be used as a measure of therapeutic response (Palmer et al 1999; Moody et al 2000) By contrast, HRP-2 can persist for several weeks after successful treatment (Shiff et al 1993) Nevertheless, studies of OptiMALÒ in this context have not always taken into consideration the ability of gametocytes to produce detectable levels of pLDH (Oduola et al 1997), especially when regimens that not include artemisinin or primaquine have been used (Tjitra & Anstey 2001) Because of the need for further studies (WHO 2000) and as antigen detection methods have not been evaluated formally in Vietnam, we compared the sensitivity, specificity and post-treatment persistence of OptiMALÒ, ICT Malaria P.f./P.v.Ò and Paracheck-PfÒ in Vietnamese patients and healthy volunteers To limit the effect of gametocytaemia, the patients with falciparum malaria received artesunate as initial therapy Subjects and methods Subjects We studied 252 Vietnamese patients with slide-positive malaria and 160 asymptomatic, aparasitaemic control subjects The study was conducted between July and ª 2002 Blackwell Science Ltd September 2000 in Phuoc Long district, Binh Phuoc province (147 patients and 60 controls) and in Lac district, Dac Lac province (105 patients and 100 controls) The study districts in Binh Phuoc and Dac Lac provinces are in highly endemic areas of southern Vietnam containing dense jungle and rubber plantations Both areas are well known for multidrug resistant falciparum malaria Patients aged 4–60 years with fever or a recent history of fever were eligible for recruitment Pregnant patients, and those with severe malaria or concomitant illness were excluded, as were those who had been treated for malaria in the previous weeks All controls were also aged between and 60 years and were from the same areas as the patients Control subjects who tested positive by any of the three antigen detection methods were questioned regarding febrile episodes in the previous weeks and excluded if they reported that they had received treatment for malaria during this period All patients and control subjects gave informed consent to study procedures that were approved by the Research Committee of the National Institute for Malariology, Parasitology and Entomology and by the Ethical Committee of the Ministry for Health, Vietnam Methods An initial fingerprick blood sample was taken from each patient and control for thick and thin smears and OptiMALÒ (DioMed AG, Switzerland), ICT malaria P.f./P.v.Ò (AMRAD, NSW, Australia) and Paracheck-Pf Ò (Orchid Biomedical System, Goa, India) tests All tests used were within expiry date Three experienced technicians performed antigen testing according to the manufacturer’s recommendations, each taking responsibility for a single test type Test results were recorded without reference to one another and then results were cross-checked by each technician in turn Blood smears were stained with Giemsa and examined by a skilled microscopist, without reference to the results of antigen testing Parasite densities expressed as number per micro litre whole blood were determined by counting the number of asexual forms per 1000 white blood cells and assuming a total white cell count of 8000 · 106/l Patients identified as positive by any test for vivax malaria were treated with chloroquine and discharged Patients diagnosed with falciparum malaria and asexual parasite density > 1000/ll whole blood were admitted to hospital and randomized to receive artesunate combined with either chloroquine or sulphadoxine/pyrimethamine in conventional doses Those with parasitaemias < 1000/ll were treated according to national guidelines with 305 Tropical Medicine and International Health volume no pp 304–308 april 2002 N M Huong et al Antigen detection in malaria diagnosis sequential artesunate and mefloquine All treatments were supervised and patients were observed for at least 60 after dosing Any patients vomiting during this period were retreated but excluded from the study All admitted patients were kept in hospital for at least days All patients in Binh Phuoc, whether admitted to hospital or not, and all admitted patients in Dac Lac were asked to return for follow-up on days 7, 14, 21 and 28 or if they became symptomatic On these occasions patients gave a blood sample, underwent antigen testing and a clinical assessment Those who failed to attend were contacted by local health workers in an attempt to provide as complete a follow-up as possible Results Overall, Paracheck-PfÒ was the best test for detecting P falciparum [sensitivity (95% confidence intervals) 95.8 (92–99%)], followed by ICT malaria P.f./P.v.Ò [82.6 (77–89%)] and OptiMALÒ [49.7 (42–58%)] After excluding six control subjects (two from Binh Phuoc and from Dac Lac) who tested positive for any of the tests, the specificities were all 100% For diagnosis of vivax malaria, the OptiMALÒ test performed better than ICT malaria P.f./ P.v.Ò [sensitivities 73.7 (65–83%) and 20.0 (12–28%), respectively], with 100% specificity in both cases The results of sensitivity testing according to parasitaemia and by study site are shown in the Table It should be noted that, because of the requirement for fever in patients from a high transmission area with a degree of semi-immunity, there were no cases in which the parasite density was < 100/ll There were lower sensitivities for each of the three tests at Dac Lac compared with Binh Phuoc All three tests performed suboptimally for P falciparum infections at parasitaemias < 1000/ll (overall sensitivities £ 75.0%), but both Paracheck-PfÒ and ICT malaria P.f./P.v.Ò had good sensitivity above this cut-off OptiMALÒ did not perform well in any situation except in the highest parasitaemia subgroup at Binh Phuoc, although it appeared to be a good test for P vivax at parasitaemias > 1000/ll In the case of ICT malaria P.f./P.v.Ò, adequate sensitivity was seen only at the highest parasitaemias The persistence of positive antigen test results in 87 patients who did not represent with P falciparum asexual parasitaemia is shown in Fig There was a stepwise increase in the proportion of patients with positive tests at days post-treatment and beyond as the baseline parasitaemia increased for all three tests Consistent with its high diagnostic sensitivity, Paracheck-Pf Ò was the test with the greatest proportion of positive patients for any time-point and parasite density In the case of OptiMALÒ, at parasitaemias > 10 000/ll the test remained positive in a relatively small proportion of patients during the first weeks after treatment Serial results for the ICT malaria P.f./P.v.Ò were intermediate between those of the other two tests Discussion Our data show that Paracheck-PfÒ is the antigen testing device of choice for uncomplicated falciparum malaria in endemic areas of southern Vietnam This result is consistent with a Thai study (Proux et al 2001) in which the sensitivity of Paracheck-PfÒ for P falciparum was 92.3% (compared with 95.8% in the present study) and 84.6% for ICT malaria P.f./P.v.Ò (compared to our 82.6%) As is the case in Thailand (Proux et al 2001), Paracheck-PfÒ is significantly cheaper than ICT malaria P.f./P.v.Ò in Vietnam The relatively high sensitivity of Paracheck-PfÒ appears to come at the cost of persistence of a positive result in successfully treated patients, a finding that may limit its usefulness in areas in which malaria transmission rates are very high Indeed, if a patient has a history of treated malaria within at least weeks of presentation, the result should be interpreted with caution Six of our healthy control subjects tested positive for at least one of Table Sensitivity of the three antigen detection methods for patient subgroups classified by study site and baseline parasitaemia Plasmodium falciparum Plasmodium vivax Study site Parasite density (/ll) Number Paracheck-PfÒ ICT malaria P.f./P.v.Ò OptiMALÒ Number ICT malaria P.f./P.v.Ò OptiMALÒ Dac Lac £ 1000 1000–£ 10 000 10 000–£ 100 000 11 41 13 55 100 100 27 88 100 29 69 29 38 100 45 100 100 Binh Phuoc £ 1000 1000–£ 10 000 10 000–£ 100 000 17 41 34 88 100 100 88 100 100 29 69 91 10 36 14 89 10 100 100 306 ª 2002 Blackwell Science Ltd Tropical Medicine and International Health volume no pp 304–308 april 2002 N M Huong et al Antigen detection in malaria diagnosis Parasitaemia ≤1000/µl 100 Parasitaemia 1000 to ≤10,000/µl Parasitaemia 10,000 to ≤100,000/µl 90 Figure Percentages of patients with persistence of a positive test result for OptiMALÒ (d——-d), ICT malaria P.f./P.v.Ò (s- - - - -s) and Paracheck-Pf Ò (m— - —m) tests subdivided by presenting parasitaemia Numbers in each subgroup are shown in parentheses adjacent to each line There were no patients with baseline parasitaemia < 1000/ll who were both positive for OptiMALÒ and had full followup data Percentage positive 80 70 60 (56) (5) 40 (28) (4) 30 20 (25) 10 (17) (21) 0 10 the three tests used and, in each case, there was a history consistent with recently treated malaria The currently available OptiMALÒ test did not perform well enough in our study to be recommended as a diagnostic aid in Vietnam In fact, its overall sensitivity (49.7%) was the lowest so far reported in the literature (60%; Fryauff et al 2000), and a number of patients whose parasitaemias were easily detectable by microscopy (>40 000/ll) tested negative The reason for this result is unclear It may relate to the quality of the batch used in the present study Alternatively, strain-specific differences in the antigen–antibody interaction used in the test may be to blame, as the reported sensitivity of OptiMALÒ has varied significantly between different geographical locations Nevertheless, OptiMALÒ is a better test than ICT malaria P.f./P.v.Ò for diagnosing vivax malaria and, as reported by others (Palmer et al 1999; Moody et al 2000), perhaps also in assessing the response to treatment in falciparum malaria cases who have an initially positive OptiMALÒ result None of our subjects remained positive for OptiMALÒ beyond 14 days after treatment The performance of all of the tests fell away at relatively low parasitaemias, consistent with the findings of others (Fryauff et al 2000; Iqbal et al 1999; Lema et al 1999; Ricci et al 2000; Singh et al 2000; Proux et al 2001) This can be potentially dangerous, as to miss the diagnosis of malaria in a one-off assessment of an ambulant, febrile patient may mean that complications develop because appropriate treatment is not instituted ª 2002 Blackwell Science Ltd (26) 50 20 30 10 20 30 10 20 30 Time (days) The assessment of a negative result in this situation will be clearly influenced by the clinical features and, given the differences between our two study sites, perhaps also local test performance characteristics Depending on the availability of microscopy, consideration should be given to bringing the patient back for review including repeat antigen testing Patients who present with established complications need to be managed in an inpatient facility, and repeated tests (whether blood smears or antigen detection) are usually more easily arranged if the diagnosis is not confirmed initially In addition, antimalarial therapy is often given empirically to such patients on clinical grounds Our results add to the evidence that antigen testing can prove a valuable adjunct to clinical assessment of the patient and blood film microscopy in certain circumstances, but their sensitivity indicates that they should not yet be regarded as a first-line diagnostic test As supported by the conclusions of a recent review (WHO 2000), careful examination of thick and thin blood smears by a trained microscopist should remain the goal of malaria control programmes in countries such as Vietnam Acknowledgements We would like to thank Ms Cath Barker and Dr Le Thi Nga of the Vietnam-Australia Malaria Control Project for their valuable assistance This study received financial support from AusAID (the Australian Agency for International Development) 307 Tropical Medicine and International Health volume no pp 304–308 april 2002 N M Huong et al Antigen detection in malaria diagnosis References Cho D, Kim KH, Park SC et al (2001) Evaluation of rapid immunocapture assays for diagnosis of Plasmodium vivax in Korea Parasitology Research 87, 445–448 Fryauff DJ, Purnomo, Sutamihardja MA et al (2000) Performance of the OptiMAL assay for detection and identification of malaria infections in asymptomatic residents of Irian Jaya, Indonesia American Journal of Tropical Medicine and Hygiene 63, 139–145 Iqbal J, Sher A, Hira PR & Al Owaish R (1999) Comparison of the OptiMAL test with PCR for diagnosis of malaria in immigrants Journal of Clinical Microbiology 37, 3644–3646 Lema OE, Carter JY, Nagelkerke N et al (1999) Comparison of five methods of malaria detection in the outpatient setting American Journal of Tropical Medicine and Hygiene 60, 177–182 Moody A, Hunt Cooke A, Gabbett E & Chiodini P (2000) Performance of the OptiMAL malaria antigen capture dipstick for malaria diagnosis and treatment monitoring at the Hospital for Tropical Diseases, London British Journal of Haematology 109, 891–894 Oduola AM, Omitowoju GO, Sowunmi A et al (1997) Plasmodium falciparum: evaluation of lactate dehydrogenase in monitoring therapeutic responses to standard antimalarial drugs in Nigeria Experimental Parasitology 87, 283–289 Palmer CJ, Validum L, Lindo J et al (1999) Field evaluation of the OptiMAL rapid malaria diagnostic test during anti-malarial therapy in Guyana Transactions of the Royal Society of Tropical Medicine and Hygiene 93, 517–518 308 Proux S, Hkirijareon L, Ngamngonkiri C, McConnell S & Nosten F (2001) Paracheck-Pf: a new, inexpensive and reliable rapid test for P falciparum malaria Tropical Medicine and International Health 6, 99–101 Ricci L, Viani I, Piccolo G et al (2000) Evaluation of OptiMAL Assay test to detect imported malaria in Italy New Microbiology 23, 391–398 Shiff CJ, Premji Z & Minjas JN (1993) The rapid manual ParaSight-F test A new diagnostic tool for Plasmodium falciparum infection Transactions of the Royal Society of Tropical Medicine and Hygiene 87, 646–648 Singh N, Saxena A & Valecha N (2000) Field evaluation of the ICT malaria P.f/P.v immunochromatographic test for diagnosis of Plasmodium falciparum and P.vivax infection in forest villages of Chhindwara, central India Tropical Medicine and International Health 5, 765–770 Tjitra E & Anstey NM (2001) Will the high rates of post-treatment sexual stage parasitaemia seen in malaria-endemic areas make the OptiMAL antigen test unreliable in predicting malaria treatment outcome? British Journal of Haematology 113, 255–257 Tjitra E, Suprianto S, Dyer M, Currie BJ & Anstey NM (1999) Field evaluation of the ICT malaria P.f/P.v immunochromatographic test for detection of Plasmodium falciparum and Plasmodium vivax patients with a presumptive clinical diagnosis of malaria in eastern Indonesia Journal of Clinical Microbiology 37, 2412–2417 WHO (2000) New Perspectives Malaria Diagnosis Report of a Joint WHO/USAID Informal Consultation, 25–27 October 1999 WHO, Geneva ª 2002 Blackwell Science Ltd