HUMAN AFRICAN TRYPANOSOMIASIS (HAT): THE DISEASE AND ITS DIAGNOSIS

Sleeping sickness occurs in sub-Saharan Africa and has two clinical manifestations (Odiit et al. Reference Odiit, Kansiime and Enyaru1997; Kennedy, Reference Kennedy2004). A chronic illness caused by T. b. gambiense can last for years and mainly affects individuals in Central and West Africa. An acute form of HAT, caused by T. b. rhodesiense, can kill victims within months and occurs mainly in East and southern Africa. In both cases, the disease occurs in rural and remote areas where the healthcare system is underdeveloped or non-existent thus impairing rapid and proper diagnosis. According to a recent survey reported by the WHO (Simarro et al. Reference Simarro, Jannin and Cattand2008), the most affected countries to date are the Democratic Republic of Congo, Angola, Central African Republic, Southern Sudan, Chad and Uganda. Here, in total 11,868 new cases of sleeping sickness were recorded in 2006.

T. b. gambiense infections represent more than 90% of the total number of reported cases. Individuals infected with this parasite can remain asymptomatic for months or even years. At this first (haemolymphatic) stage of the disease the trypanosomes multiply in the tissues and body fluids. The patient typically experiences symptoms such as headache, fever and joint pains. As these symptoms are non-specific to HAT, the patient often remains undiagnosed at this stage of the disease, which decreases the chances of treatment success when diagnosed later. The disease progresses into the second (meningoencephalitic) stage, when trypanosomes spread into the central nervous system by crossing the blood-brain barrier. Second-stage patients experience various symptoms including confusion, sensory sleep disturbances, endocrine dysfunction, mental changes, poor coordination and, if not treated, lapse into a coma and eventual death.

T. b. rhodesiense infections represent less than 10% of all reported cases, with only 486 infections recorded by the WHO in 2006 (Simarro et al. Reference Simarro, Jannin and Cattand2008). The disease normally develops rapidly into the second stage with a high chance of lethal outcome. Also in case of T. b. rhodesiense HAT, chronic forms of the disease have been reported (MacLean et al. Reference MacLean, Chisi, Odiit, Gibson, Ferris, Picozzi and Sternberg2004).

Treatment of sleeping sickness is complicated and depends on (1) the T. brucei subspecies involved, and (2) on the stage of the disease (Docampo and Moreno, Reference Docampo and Moreno2003; Pepin and Milord, Reference Pepin and Milford1994). While first-stage infections by T. b. gambiense are treated with pentamidine (Sanderson et al. Reference Sanderson, Dogruel, Rodgers, De Koning and Thomas2009) or DFMO (difluromethylornithine) (Heby et al. Reference Heby, Persson and Rentala2007), only the latter can be used to treat the encephalitic stage of the disease. In contrast, early stage T. b. rhodesiense infections can be treated with suramin but not DFMO (Iten et al. Reference Iten, Matovu, Brun and Kaminsky1995), while here treatment of second-stage T. rhodesiense HAT relies on melarsoprol (Balasegaram et al. Reference Balasegaram, Young, Chappuis, Priotto, Raguenaud and Checchi2009). This arsenic derivative has high toxicity and the adverse effects of its use include the development of encephalopathy in 5–10% of patients (Pepin and Milord, Reference Pepin and Milford1994). In this treatment setting it is clear that an accurate diagnosis of HAT is essential for proper drug treatment. While lack of drug treatment leads to certain death, avoiding unnecessary treatment can reduce costs of HAT control programmes, and will save the patient from severe treatment-associated side effects.

Diagnostic methods used for sleeping sickness confirmation are cumbersome and inaccurate due to inadequate sensitivity and specificity of currently available tests. In rural Africa, due to a lack of the hospital infrastructure, diagnosis of T. gambiense HAT is conducted by various control programmes that rely on active case detection through mass population screening, followed by diagnostic confirmation and staging. In practice, some individuals might be subjected to 6 consecutive tests in order to confirm their infection status and before being treated or dismissed without treatment. Each diagnostic method (described below) has its shortcomings and only cumulative analysis of all the results allows final diagnosis and treatment. In contrast, T. b. rhodesiense diagnosis tends to take place in rural health centres with patients self-reporting or being referred by rural health workers.

During HAT screening, individuals are first examined for clinical signs such as enlarged cervical lymph nodes, oedema, splenomegaly, hepatomegaly, fever, pruritus and/or weight loss. These clinical signs are in general insufficient for specific diagnosis of sleeping sickness. After clinical examination, in T. b. gambiense-endemic areas individuals are screened using the Card Agglutination Test for Trypanosomiasis (CATT), applied on whole blood. Developed in 1978, CATT is an antibody-detecting test that is widely used for screening of the population affected by T. b. gambiense (Magnus et al. Reference Magnus, Vervoort and Van Meirvenne1978). No similar test is available to diagnose T. b. rhodesiense infections in Eastern Africa. Next, CATT-seropositive individuals undergo further parasitological examination that involves microscopic analysis of a wet smear of lymph node aspirate or mHCT (micro-Haematocrit Centrifugation Technique). If parasites are found, the individuals are subjected to a stage determination test, performed after a lumbar puncture and analysis of cerebrospinal fluid (CSF). According to the guidelines of WHO, second-stage sleeping sickness is defined by the presence in the CSF of (1) parasites, (2) increase in white blood cell levels, or (3) increased total protein concentration (Lejon et al. Reference Lejon, Reiber, Legros, Djé, Magnus, Wouters, Sindic and Büscher2003; Chappuis et al. Reference Chappuis, Loutan, Simarro, Lejon and Büscher2005), albeit this parameter is used less frequently, as it requires specific reagents often unavailable in the rural health centres (Chappuis et al. Reference Chappuis, Loutan, Simarro, Lejon and Büscher2005). The detection of parasites in the CSF classifies an individual as a second-stage patient, irrespective of any other parameter. With respect to white blood cell counts, the WHO considers an individual with a count higher than 5 cells/μl as a second-stage patient (Miezan et al. Reference Miezan, Meda, Doua, Yapo and Baltz1998). However, some African countries with high prevalence of HAT, such as Angola, use a 20 cells/μl as cut-off before second-stage drug treatment is initiated (Stanghellini and Josenando, Reference Stanghellini and Josenando2001; Chappuis et al. Reference Chappuis, Loutan, Simarro, Lejon and Büscher2005).

While the introduction of T. b. gambiense CATT for mass population screening has certainly facilitated the recent successes of anti-trypanosomiasis campaigns, most difficulties arise when CATT seropositives are not confirmed by a parasitological examination (i.e. the direct observation of parasites). These individuals are classified as ‘serological suspects’ and may undergo up to 5 additional tests in order to confirm their status. In this confirmation procedure, the suspect patients are tested once again by CATT but now using 1/4 diluted sera rather than whole blood. Individuals who score negative here are then sent home without treatment, while those that still show a positive CATT score are further examined using parasitological tests such as mHCT, QBC (Quantitative Buffy Coat) and/or mAECT (mini-Anion Exchange Centrifugation Technique) (Woo, Reference Woo1970; Levine et al. Reference Levine, Wardlaw and Patton1989; Gutierrez et al. Reference Gutierrez, Corbera, Doreste and Büscher2004). Interpretation of CATT results is problematic when the score remains positive with the use of 1/4 diluted serum, while no parasitological confirmation can be obtained. Individuals falling within this category are normally tested once again using CATT. Serological suspects who score negative below 1/8 serum dilution will be sent home to be called later for follow-up. The WHO recommended in this case a follow-up period of two years. It is obvious that this raises some serious problems, as HAT-suspects are difficult to re-examine due to non-existent sustainable health care infrastructure, lack of transport and population movements.

As CATT is based on anti-parasite antibody screening, as opposed to parasite or parasite-antigen screening, the number of screened individuals who fall within the problematic ‘false positive’ group can be high. Indeed, one study in Ivory Coast demonstrated that only one individual from 77 CATT-seropositives was initially parasite negative, developed HAT during a two-year follow up (Simarro et al. Reference Simarro, Ruiz, Franco and Josenando1999; Garcia et al. Reference Garcia, Jamonneau, Magnus, Laveissière, Lejon, N'Guessan, N'Dri, Van Meirvenne and Büscher2000). Similarly, researchers in Angola found that only half of serologically suspected individuals who were initially parasitologically negative were diagnosed with HAT during the follow-up (Simarro et al. Reference Simarro, Ruiz, Franco and Josenando1999). The phenomenon of seropositivity in parasitologically negative individuals is usually attributed to the inadequate specificity of CATT. One factor involved in this might be the cross-reactivity of antibodies induced by exposure to non-pathogenic trypanosome species. For example, individuals who were bitten by the tsetse flies carrying T. b. brucei or T. congolense developed antibody responses that cross-reacted with antigens from human pathogenic T. b. gambiense and T. b. rhodesiense (Magnus et al. Reference Magnus, Vervoort and Van Meirvenne1978; Chappuis et al. Reference Chappuis, Loutan, Simarro, Lejon and Büscher2005). In addition, sera from HAT patients and suspects shows a wide spectrum of antigen recognition related to the fact that those individuals might also carry other diseases such as malaria, tuberculosis, HIV-AIDS and helminth infections. Several of these infections have been characterized as the cause for non-specific B cell proliferation, and as such are doomed to have a negative impact on CATT specificity due to non-specific cross-reactivity.

A final problem occurring with the CATT assay, is related to the fact that it uses a single laboratory-adapted strain of T. b. gambiense (LiTat 1.3) parasites for agglutination. While the LiTat 1.3 VSG variant is believed to be expressed in a subset of parasites at least once in every infected-individual during most West-African HAT infections, this is not always the case. For example the LiTat1.3 variant is apparently missing or not being expressed during some HAT infections in Cameroon (Asonganyi et al. Reference Asonganyi, Bedifeh, Ade and Ngu1994; Dukes et al. Reference Dukes, Gibson, Gashumba, Hudson, Bromidge, Kaukus, Asonganyi and Magnus1992) and regions of Uganda (Enyaru et al. Reference Enyaru, Matovu, Akol, Sebikali, Kyambadde, Schmidt, Brun, Kaminsky, Ogwal and Kansiime1998). Hence, in an effort to broaden the geographic user range of the CATT, the assay has been converted to an antibody-based latex agglutination test that combines the detection of LiTat 1.3, 1.5 and 1.6 VSG variants by antibody agglutination (Truc et al. Reference Truc, Lejon, Magnus, Jamonneau, Nangouma, Verloo, Penchenier and Büscher2002; Penchenier et al. Reference Penchenier, Grébaut, Njokou, Eboo Eyenga and Büscher2003).

One remark to be made about the evaluation of specificity of CATT, or the alternative LATEX/CATT, relates to the reported specificity values of over 90% (Jamonneau et al. Reference Jamonneau, Truc, Garcia, Magnus and Büscher2000) that seem to contradict the high rate of false positive CATT cases reported above. These specificity values are highly biased by inclusion of uninfected individuals, and caution has to be taken when interpreting these values. For example, in the study referred to above, results were obtained by screening of a cohort of 425 parasitologically negative patients, without a single confirmed HAT case. CATT in this case resulted in 33 ‘false-positive’ scores. Hence, as specificity is calculated as 100×True Negatives/(True Negatives+False Positives) (Dukes et al. Reference Dukes, Rickman, Killick-Kendrick, Kakoma, Wurapa, de Raadt and Morrow1984), the value obtained was 92·5%. In a similar study that included 59 confirmed HAT cases and 2019 parasitologically negative individuals, and used same calculation method (Penchenier et al. Reference Penchenier, Grébaut, Njokou, Eboo Eyenga and Büscher2003), a CATT specificity of 85·4% was reported. Here CATT-positive results could only be confirmed in 1 of 6 patients, i.e. 352 individuals obtained a CATT-positive score while only 59 were found to be parasitologically positive HAT patients.

RECENT DEVELOPMENTS IN PARASITE AND PARASITE-ANTIGEN DETECTION

A new test that would be able to replace the CATT should have greater specificity, hence avoiding false positive scores in patients with unrelated parasitic or other infections. It is obvious that even if such an assay were to be developed, the major limitation of an antibody-based detection test is the fact that it would be useless in the follow-up of treated patients in whom anti-trypanosome antibodies persist in circulation for a significant period (Paquet et al. Reference Paquet, Ancelle, Gastellu-Etchegorry, Castilla and Harndt1992). An alternative approach avoiding the problem of the persistence of anti-trypanosome antibodies would be a parasite antigen detection test. This would also have an improved specificity, but it might be impaired by the low concentration of antigens available in circulation, and hence suffer an impaired sensitivity.

The trypanosome surface is coated with 10⁷ identical molecules of variant surface glycoprotein (VSG), regularly changing the antigenic composition during progressing infection. While VSG molecules account for 10% of the entire protein content of the parasite, making it an attractive diagnostic marker in terms of copy number, the development of a single, reliable and universal diagnostic test able to detect a common antigen poses a major obstacle due to antigenic variation. On the other hand, non-variant trypanosome surface proteins, such as the invariant surface glycoproteins ISG64 (Jackson et al. Reference Jackson, Windle and Voorheis1993), ISG65 and 75 (Ziegelbauer et al. Reference Ziegelbauer, Multhaup and Overath1992; Ziegelbauer and Overath, Reference Ziegelbauer and Overath1992) and ISG100 (Nolan et al. Reference Nolan, Jackson, Windle, Pays, Geuskens, Michel, Voorheis and Pays1997) are only expressed in relatively low copy numbers, although their invariant nature makes them attractive diagnostic candidates. Other possible antigen candidates include various intracellular proteins such as trypanosome-derived proteases including oligopeptidase B (Troeberg et al. Reference Troeberg, Pike, Morty, Berry, Coetzer and Lonsdale-Eccles1996), trypanothione (Fairlamb et al. Reference Fairlamb, Henderson and Cerami1989), and various cysteine proteases (Pamer et al. Reference Pamer, So and Davis1989; Authie et al. Reference Authié, Muteti, Mbawa, Lonsdale-Eccles, Webster and Wells1992; Troeberg et al. Reference Troeberg, Pike, Morty, Berry, Coetzer and Lonsdale-Eccles1996; Okenu et al. Reference Okenu, Opara, Nwuba and Nwagwu1999). These antigens are currently under evaluation for diagnostic screening, using confirmed HAT-positive sera collected from patients and negative controls from endemic African regions. As these assays are being developed it will be important to include samples from CATT-seropositive suspects that cannot be confirmed by parasitological analysis, in order to overcome the problems outlined above with this patient group.

In contrast to the continued lack of an antigen detection test for HAT, in malaria considerable progress has been made. For example the Rapid Test for Diagnostics of malaria (RTD) is based on the detection of the Plasmodium falciparum-specific antigens histidine-rich protein 2 (PfHRP2) or parasite specific lactate dehydrogenase (pLDH). This test is currently available in an easy to use dipstick format (Meena et al. Reference Meena, Joshi, Joshi, Sridhar, Waghdhare, Gangane and Kalantri2009).

Over the past decade new approaches to antigen detection have been subject of experimental study. For example, RNA-derived aptamers have been proposed (Homann et al. Reference Homann, Lorger, Engstler, Zacharias and Goringer2006) that combine high specificity, thermo-stability and simplicity of use. Aptamers are single-stranded nucleic acid probes that can be selected from a large random library. They are analogues to antibodies in their mode of action, can be coupled to biotin or gold particles and are used for development of lateral flow tests (dip sticks). For example, aptamers are already being used to detect cocaine in human blood or serum, using immobilized cocaine aptamer-linked nanoparticle aggregates onto a lateral flow device (Swensen et al. Reference Swensen, Xiao, Ferguson, Lubin, Lai, Heeger, Plaxco and Soh2009). While the use of anti-trypanosome aptamers remains to be applied in a easy-to-use diagnostic format, VSG-specific serum-stable aptamers have now been generated and show affinities in subnanomolar range, binding to the structurally conserved epitopes of the VSG (Goringer et al. Reference Goringer, Homann and Lorger2003; Lorger et al. Reference Lorger, Engstler, Homann and Göringer2003). Other novel probes with HAT diagnosis potential are trypanosome-specific Nanobodies® (Nbs). These probes are derived from naturally occurring heavy-chain camel IgG antibodies, which lack the light chains of a conventional antibody (Hamers-Casterman et al. Reference Hamers-Casterman, Atarhouch, Muyldermans, Robinson, Hamers, Songa, Bendahman and Hamers1993; Muyldermans et al. Reference Muyldermans, Baral, Retamozzo, De Baetselier, De Genst, Kinne, Leonhardt, Magez, Nguyen, Revets, Rothbauer, Stijlemans, Tillib, Wernery, Wyns, Hassanzadeh-Ghassabeh and Saerens2009). Using recombinant technology, 15 KDa single heavy-chain fragments can be generated from these camel antibodies (Muyldermans, Reference Muyldermans, Baral, Retamozzo, De Baetselier, De Genst, Kinne, Leonhardt, Magez, Nguyen, Revets, Rothbauer, Stijlemans, Tillib, Wernery, Wyns, Hassanzadeh-Ghassabeh and Saerens2009). Nbs and Nb-probes are much smaller than conventional antibodies, can access unique epitopes, are higly resistant to heat and pH changes and can be tailored for highly specific purposes (Lauwereys et al. Reference Lauwereys, Arbabi Ghahroudi, Desmyter, Kinne, Hölzer, De Genst, Wyns and Muyldermans1998; Vincke et al. Reference Vincke, Loris, Saerens, Martinez-Rodriguez, Muyldermans and Conrath2009). They have already been shown to have diagnostic and therapeutic potential for various infections and diseases (Cortez-Retamozo et al. Reference Cortez-Retamozo, Backmann, Senter, Wernery, De Baetselier, Muyldermans and Revets2004; Tijink et al. Reference Tijink, Laeremans, Budde, Stigter-van Walsum, Dreier, de Haard, Leemans and van Dongen2008; Deckers et al. Reference Deckers, Saerens, Kanobana, Conrath, Victor, Wernery, Vercruysse, Muyldermans and Dorny2009). In the context of HAT, a Nb that binds to the conserved carbohydrate epitopes of VSG was shown to be a useful tool in both parasite detection and specific anti-parasite drug targeting (Stijlemans et al. Reference Stijlemans, Conrath, Cortez-Retamozo, Van Xong, Wyns, Senter, Revets, De Baetselier, Muyldermans and Magez2004; Baral et al. Reference Baral, Magez, Stijlemans, Conrath, Vanhollebeke, Pays, Muyldermans and De Baetselier2006). In addition, several poly-reactive anti-trypanosome Nbs were generated more recently, starting from a VHH library that was constructed using lymphocytes of a T. evansi-infected camel (Saerens et al. Reference Saerens, Stijlemans, Baral, Nguyen Thi, Wernery, Magez, De Baetselier, Muyldermans and Conrath2008). Combining these Nbs might result in the future development of new HAT-diagnostic tools that could include fluorescent probes for rapid and easy parasite detection by microscopy. In this context, the introduction by Zeiss™ of an LED-based immuno-fluorescence microscopy technology that is highly suitable for field use and also relatively cheap is a major breakthrough which will act as an enabling technology for newly developed fluorescent detection probe methods (Lang et al. Reference Lang, Zeiser, Schultz, Stellmacher, Vollmer, Zabel and Goldmann2008). Under normal light microscopy conditions, the detection limits of parasitological techniques such as fresh thin blood smears and thick drop smears stained with Giemsa are 10,000 parasites/ml and 5,000 parasites/ml blood, respectively. This remains a diagnostic limiting factor as HAT patients may harbour parasite numbers below 100 parasites/ml (Chappuis et al. Reference Chappuis, Loutan, Simarro, Lejon and Büscher2005). Hence, an estimated 30% of HAT-positive individuals are missed using conventional microscopy techniques. While fluorescent microscopy techniques would not improve the theoretical detection limit, they would however make parasite detection much more easy and reliable, and therefore should reduce the number of false-negative scores.

In order to improve the detection limit of parasitological techniques, parasites in the blood can be concentrated using the mini-Anion Exchange Centrifugation Technique (m-AECT) (Kimber, Reference Kimber1984; Gutierrez et al. Reference Gutierrez, Corbera, Doreste and Büscher2004). Here, parasites are separated and concentrated from the blood based on the difference in cell surface charge at a given pH. While red and white blood cells adhere to the DEAE gel particles of the mini-column at a pH range of 7·7–8·0, trypanosomes will flow through under these pH conditions (Lanham, Reference Lanham1968). Finally, sample analysis relies on microscope screening of the column eluate, reaching an estimated detection limit of 100 parasites/ml. This detection limit is comparable to the Quantitative Buffy Coat method (QBC) (Levine et al. Reference Levine, Wardlaw and Patton1989), a technique that is also used for detection of malaria parasites. It involves staining of the centrifuged blood with acridine-orange and its examination under a UV light source. Here, acridine orange will stain the DNA of both white blood cells and parasites, but will not stain the DNA-lacking red blood cells. Acridine-orange pre-coated QBC capillaries can be loaded with blood collected from a finger-prick and subsequently centrifuged, separating white blood cells and trypanosomes from erythrocytes according to their densities. Subsequently, the QBC tube can be examined using a capillary holder and a standard light microscope equipped with the UV microscope adapter, or a fluorescence microscope.

RECENT MOLECULAR BIOLOGY DEVELOPMENTS FOR HAT DIAGNOSIS

In recent years, several unique gene targets have been identified that could result in the development of new molecular tools for the diagnosis of HAT. These targets include unique conserved 18S rRNA trypanosome sequences (Radwanska et al. Reference Radwanska, Magez, Perry-O'Keefe, Stender, Coull, Sternberg, Büscher and Hyldig-Nielsen2002 a; Deborggraeve et al. Reference Deborggraeve, Claes, Laurent, Mertens, Leclipteux, Dujardin, Herdewijn and Büscher2006), the 5.8S rRNA-internal transcribed spacer 1 and 2 (ITS1 and 2) (Njiru et al. Reference Njiru, Constantine, Guya, Crowther, Kiragu, Thompson and Dávila2005; Thekisoe et al. Reference Thekisoe, Kuboki, Nambota, Fujisaki, Sugimoto, Igarashi, Yasuda and Inoue2007), genes such as esag6 and esag7, sra or tgsgp (Kabiri et al. Reference Kabiri, Franco, Simarro, Ruiz, Sarsa and Steverding1999; Radwanska et al. Reference Radwanska, Claes, Magez, Magnus, Perez-Morga, Pays and Büscher2002 b, Reference Radwanska, Chamekh, Vanhamme, Claes, Magez, Magnus, De Baetselier, Büscher and Paysc), or the repetitive insertion mobile element (RIME) (Hide and Tilley, Reference Hide and Tilley2001; Cox et al. Reference Cox, Tilley, McOdimba, Fyfe, Eisler, Hide and Welburn2005). These targets could be subject to detection by molecular probes such as PNA (peptide nuclear acid)-probes or to detection by amplification methods such as PCR or LAMP (loop-mediated isothermal amplification). The latter technique was developed almost a decade ago (Notomi et al. Reference Notomi, Okayama, Masubuchi, Yonekawa, Watanabe, Amino and Hase2000) and was more recently shown to be capable of detection down to 1 fg trypanosome DNA, equivalent to 0·01 parasite (Thekisoe et al. Reference Thekisoe, Kuboki, Nambota, Fujisaki, Sugimoto, Igarashi, Yasuda and Inoue2007). However, while this detection limit can be reached with purified DNA as starting material, it has yet to be achieved when infected blood is used as the starting sample. LAMP allows DNA amplification without the need for heat denaturation of the double-stranded DNA into a single-stranded form, relying on autocycling strand displacement DNA synthesis by a Bst DNA polymerase. Amplification and detection can be completed in a single step yielding a stem–loop DNA structure with DNA being amplified up to 10¹⁰ times within 15 to 60 minutes. While PCR reactions require the use of a thermocycler, LAMP amplifies DNA at constant temperature, using a simple heating block, resulting in high specificity and high efficiency DNA amplification. Hence, this technique appears to be excellent for implementation in settings such as HAT screening centres, where the need for expensive and difficult to maintain equipment often poses problems. In addition, while PCR reactions require high quality purified DNA in order to be reliable, heat-treated blood can be used as a starting material for the LAMP reaction as it uses four to six primers that recognize six to eight regions on the target DNA respectively. Another advantage of LAMP over PCR-based methods is that DNA amplification can be monitored spectrophotometrically rather than requiring gel electrophoresis and may even be observed with the naked eye. An example that demonstrates the superiority of LAMP over PCR is the use of RIME as target for diagnosis of HAT (Njiru et al. Reference Njiru, Mikosza, Matovu, Enyaru, Ouma, Kibona, Thompson and Ndung'u2008). Here the authors suggest a LAMP sensitivity of up to a 100-fold better than PCR in terms of number of parasites per ml, combined with the advantage of (1) a much simplified field applicable working protocol of LAMP as compared to PCR, and (2) a high tolerance to different biological substances. The promising results from LAMP are attracting investment from the commercial sector and this is likely to be a powerful catalyst to the development and delivery of viable diagnostic tests. As an example Eiken Chemical Co. (Japan) that specialises in the development of diagnostic LAMP reactions, now provides all the components needed for amplification lyophilised in a single tube, to which the test sample can be added. LAMP has now been shown to be effective for the detection of T. brucei brucei, T. brucei gambiense, T. b. rhodesiense, T. evansi and T. congolense in blood samples, using pfra as target (Kuboki et al. Reference Kuboki, Inoue, Sakurai, Di Cello, Grab, Suzuki, Sugimoto and Igarashi2003). Here LAMP sensitivity was up to 100 times higher than that of PCR-based methods. In addition, a LAMP-assay for the specific detection of both animal and human trypanosomiasis has been reported using primer sets that are designed from 5.8S rRNA-internal transcribed spacer 2 (ITS2) gene for T. b. gambiense, 18S rRNA for both T. congolense and T. cruzi, and the RoTat 1.2 VSG encoding gene for T. evansi (Thekisoe et al. Reference Thekisoe, Kuboki, Nambota, Fujisaki, Sugimoto, Igarashi, Yasuda and Inoue2007).

Genes that allow a differential diagnosis of T. b. rhodesiense and T. b. gambiense will be particularly attractive candidates for the molecular diagnosis of HAT, as they allow specific diagnosis in foci where there is a risk of overlap of both infections, such as in Uganda (Picozzi et al. Reference Picozzi, Fèvre, Odiit, Carrington, Eisler, Maudlin and Welburn2005). Examples of such genes are the tgsgp gene specific for T. b. gambiense marker (Radwanska et al. Reference Radwanska, Claes, Magez, Magnus, Perez-Morga, Pays and Büscher2002 b), and the sra gene, specific for T. b. rhodesianse (Radwanska et al. Reference Radwanska, Chamekh, Vanhamme, Claes, Magez, Magnus, De Baetselier, Büscher and Pays2002 c).

BIOMARKER DISCOVERY AND VALIDATION: THE LONG AND UNCERTAIN PATH

In recent years, many laboratories have embarked on the research path of host-biomarker discovery for trypanosomiasis, using various proteomic approaches. Effective biomarkers that could improve diagnosis, predict disease progression, guide drug treatment and monitor the therapeutic outcomes, would be very welcome. Of particular value would be a serum biomarker for stage determination, replacing the current invasive methods requiring lumbar puncture, and allowing avoidance of the highly toxic stage 2 disease treatments as much as possible.

One target of several biomarker discovery programmes has been the study of cytokine and chemokine concentrations in the serum and CSF of HAT patients. The results show that CSF and serum of second-stage patients infected with T. b. gambiense and T. b. rhodesiense contain high levels of cytokines, interleukins and chemokines (Lejon et al. Reference Lejon, Lardon, Kenis, Pinoges, Legros, Bisser, N'Siesi, Bosmans and Büscher2002 a, Reference Lejon, Roger, Mumba Ngoyi, Menten, Robays, N'siesi, Bisser, Boelaert and Büscher2008; MacLean et al. Reference MacLean, Chisi, Odiit, Gibson, Ferris, Picozzi and Sternberg2004; Courtioux et al. Reference Courtioux, Boda, Vatunga, Pervieux, Josenando, M'Eyi, Bouteille, Jauberteau-Marchan and Bisser2006; Hainard et al. Reference Hainard, Tiberti, Robin, Lejon, Ngoyi, Matovu, Enyaru, Fouda, Ndung'u, Lisacek, Müller, Turck and Sanchez2009). As various cytokines and chemokines are involved in the early recruitment of leukocytes into the CNS, monitoring the levels of MCP-1 (monocyte chemoattractant protein-1/CCL-2), MIP-1α (macrophage inflammatory protein-1 alpha), CCL-3, CXCL-8, IL-8, and IL-1β could be a valuable aid in stage determination. For example, while CXCL-8 was found to be higher in first-stage patients when compared with second-stage, CSF expression of the other selected cytokines was associated with the presence of neurological second-stage HAT signs (Courtioux et al. Reference Courtioux, Boda, Vatunga, Pervieux, Josenando, M'Eyi, Bouteille, Jauberteau-Marchan and Bisser2006). The use of cytokines and chemokines as HAT-staging biomarkers requires considerable further experimental study. One problem is that in all studies to date the control group has been made up by ‘healthy’ volunteers, with individuals diagnosed with other infections commonly encountered in endemic areas excluded from studies. This creates a bias towards a positive outcome. Almost all host immune molecules studied to date as potential HAT biomarkers are also being evaluated as markers for general inflammatory and anti-inflammatory responses, and as such cannot be used for specific HAT diagnosis. For example, levels of IL-8, IP-10, MCP-1 and MIP-1β are all elevated in patients with tuberculosis (Djoba Siawaya et al. Reference Djoba Siawaya, Beyers, van Helden and Walzl2009; Ruhwald et al. Reference Ruhwald, Bjerregaard-Andersen, Rabna, Eugen-Olsen and Ravn2009). CXCL13, IP-10 and MCP-1 levels are also increased in human immunodeficiency virus (HIV)-seropositive patients, and the latter are affected in this setting by HIV/tuberculosis co-infections (Juffermans et al. Reference Juffermans, Verbon, van Deventer, van Deutekom, Belisle, Ellis, Speelman and van der Poll1999). Hence, the search continues for a host immune molecule that is regulated in a unique way during HAT, and is not affected by any other disease condition prevalent in HAT endemic areas. Even if the currently proposed set of cytokine/chemokine biomarkers are only used for stage determination, after detecting the actual presence of tryposomes by more conventional means, it is likely that other common infections and diseases would still have a major impact on screening outcome.

Besides the modulation of cytokines/chemokines, the presence of antibodies in CSF has also been suggested as a biomarker for second-stage HAT complication. Determining the IgM content in the CSF can however be very difficult or even impossible to carry out in endemic areas and under field conditions. In order to address this problem, an experimental adaptation of the CATT assay has been reported, which relies on latex agglutination by host IgM (Lejon et al. Reference Lejon, Legros, Richer, Ruiz, Jamonneau, Truc, Doua, Djé, N'Siesi, Bisser, Magnus, Wouters, Konings, Vervoort, Sultan and Büscher2002 b). While this has been demonstrated to have stage diagnostic value in field studies when used in addition to conventional CSF markers, it remains to be determined if interference by other CNS infections such as neuroborreliosis, neurosyphilis, mumps meningoencephalitis and non-Hodgkin lymphoma involving the central nervous system will underline the effectiveness of this test in large-scale screening studies (Sindic et al. Reference Sindic, Monteyne and Laterre1994; Ellie et al. Reference Ellie, Vital, Steck, Julien, Henry and Vital1995; Schutzer et al. Reference Schutzer, Coyle, Krupp, Deng, Belman, Dattwyler and Luft1997; Rawal et al. Reference Rawal, Gavin and Sturgis2006).