Jpn. J. Infect. Dis., 59, 306-310, 2006

Original Article Intestinal Microsporidiosis in Diarrheal Patients Infected with Human Virus-1 in Addis Ababa, Ethiopia Tekola Endeshaw*, Amha Kebede, Jaco J. Verweij1, Ayele Zewide2, Kebede Tsige3, Yodit Abraham3, Dawit Wolday, Tilahun Woldemichael, Tsehaynesh Messele, Anton M. Polderman1 and Beyene Petros4 Ethiopian Health and Nutrition Research Institute/Ethio-Netherlands AIDS Research Project; 2Police Force General Hospital; 3Armed Force General Hospital; 4Addis Ababa University, Faculty of Science, Department of Biology, Addis Ababa, Ethiopia; and 1Leiden University Medical Centre, Parasitology Department, Leiden, The Netherlands (Received January 12, 2006. Accepted July 31, 2006) SUMMARY: Intestinal microsporidiosis is most commonly associated with persistent in advanced AIDS cases. To determine the prevalence and clinical manifestations of this infection in HIV/AIDS patients, a single fresh stool sample and blood were collected from 243 (214 HIV-positive and 29 HIV-negative) diarrheal patients. The presence of intestinal microsporidiosis in the stool was determined by Uvitex-2B staining and a PCR-based detection method. HIV screening was done by using ELISA, and reactive samples were confirmed by Western blotting. The CD4+ cell count was analyzed using FACScan. Out of 243 diarrheal patients, 39 (16.0%) cases were positive for intestinal microsporidial infection by either of the methods used. Of the 39, only 18 cases positive by microscopy were also positive by PCR. Based on PCR and microscopic analyses the microsporidial parasites were identified as (30), Ecephalitozoon intestinalis (6), and double infections (3). All -positive cases were HIV-positive, and 92.3% had diarrhea for over 4 weeks. The diarrhea was watery in 79.5% of the patients. Weight loss >10% was recorded in 37 (94.9%) cases. The CD4+ cell count was <100 cells/mm3 in 84.4% of subjects, and 59.4% of the patients had a CD4+ cell count of ≤50 cells/mm3, with a mean of 22.8 cells/mm3. This study revealed that intestinal microsporidiosis is a common cause of chronic diarrhea and severe weight loss in advanced AIDS patients in Ethiopia. This condition is attributable mainly to E. bieneusi. Thus, early diagnosis of intestinal microsporidiosis in HIV/AIDS patients would certainly be helpful in the understanding and management of diarrheal illness.

in AIDS patients (1,2,10). Different studies have indicated INTRODUCTION that infection by E. bieneusi and that by Ecephalitozoon Microsporidiosis, an infection caused by obligate intra- intestinalis are the main causes of chronic watery diarrhea cellular -forming parasites, is frequently reported and and wasting syndrome in AIDS patients when the CD4+ cell well documented in HIV/AIDS patients and immuno- count is <100 cells/mm3 (2,16,17). compromised individuals worldwide. It is still a major health Except for a few published cases, there is insufficient problem in developing countries (1-7). However, the wide information on the prevalence and magnitude of the clinical and intensive application of highly active antiretroviral importance of microsporidial infections in HIV/AIDS patients therapy (HAART) with protease inhibitors to treat HIV/AIDS in tropical and developing countries. Some studies in Africa patients in developed countries has substantially decreased the suggested that the incidence of microsporidial infections prevalence and incidence of opportunistic intestinal parasites, might be less as might be expected given the higher preva- including Enterocytozoon bieneusi (6,8). lence of immunosuppression due to HIV/AIDS in the conti- Six genera of microsporidia are known to infect humans: nent (6,14,18,19). However, this could be attributed to a lack Enterocytozoon, Encephalitozoon, Nosema, Pleistophora, of adequate diagnostic methods specific to this parasite in Trachipleistophora, and Vittaforma (1,9). Infection in humans that part of the world. The purpose of the present study was can occur in many tissues, including the small intestine, to determine the incidence and clinical features of intestinal kidney, liver, and cornea (10,11). Before 1985, reports of clini- microsporidiosis in diarrheal patients with HIV/AIDS in cal disease related to intestinal microsporidial infections in Ethiopia. humans were rare. Since then, E. bieneusi has been identified as an etiologic agent of diarrhea (12), and major clinical symp- MATERIALS AND METHODS toms associated with intestinal microsporidiosis in HIV/AIDS patients have been frequently reported (1,2,13-15). Chronic Study subjects: The study was conducted from March diarrhea, malabsorption, and severe weight loss are the most 2002 to October 2004 in the army hospital, the police hospi- common clinical syndromes associated with these infections tal, and St. Paul hospital, Addis Ababa, Ethiopia. A total of 243 individuals were selected based on the following cri- *Corresponding author: Mailing address: Ethiopian Health and teria: age greater than 14 years; either sex; and either chronic Nutrition Research Institute, P.O. Box 1242, Addis Ababa, or acute diarrhea (chronic diarrhea was defined as two or more Ethiopia. Tel: +251-1-0911172856/251-1-0112751522, E-mail: watery or loose stools per day for more than 28 days, while [email protected] acute diarrhea was defined as two or more loose or watery

306 stools per day for less than 28 days). Informed consent was (MSP-1, TGA ATG KGT CCC TGT; MSP-2A, TCA CTC obtained for each subject. Pre- and post-test counselling was GCC GCT ACT; MSP-2B, GTT CAT TCG CAC TAC ) were given to the volunteer patients for HIV serological tests by used. The second (nested) PCR was run by taking 2 μl of the trained counsellors. Clinical and basic demographic informa- first PCR product to the mix containing three inner primers tion was recorded by the attending physicians of the respec- (MSP-3, GGA ATT CAC ACC GCC CGT CRY TAT; MSP- tive hospitals. 4A, AA ARG GGT; MSP-4B, CAA AGC TTA TGC TTA Ethical clearance for the project was obtained from the AGT CCA GGG AG). PCR was amplified in a volume of 40 Ethiopian Health and Nutrition Research Institute (EHNRI) μl reaction mixture with 10 × PCR buffer 100 M Tris- and the National Ethical Committee of the Ethiopian Science HCl, pH 9.0, 15 nm MgCl2 , 500 nm KCl, 1% Triton X-100, and Technology Commission (NEC/ESTC). 0.1% (w/v) gelatin (HT Biotechnology, Cambridge, UK), Stool sample collection and processing: A single fresh 200 μl of each of A, T, G, and C nucleotides; 25 pmol of stool sample was obtained in a labeled cup from each con- each primer, 1 U of Taq polymerase (Super Taq HC; HT senting patient selected for the study. A portion of each fresh Biotehnology) and 2 μl of the DNA samples and run using stool sample was processed for Cryptosporidium parvum and the GeneAmp PCR system 9600® (Perkin-Elmer Cerus, Isospora belli oocysts by the modified Ziehl Neelsen method. Norwalk, Conn., USA). The PCR cycling parameters were The remaining portion was processed by a water-ether sedi- 25 cycles following initial denaturation of DNA at 94°C mentation method and stained by Uvitex-2B method as for 4 min, followed by denaturation at 94°C for 30 s, primer described previously (20). Briefly, 1 g of fresh stool was mixed annealing at 55°C for 45 s, elongation at 72°C for 45 s, and thoroughly with 8 ml of distilled water in a 15-ml conical test 72°C for 7 min for final extension. tube. After sieving with cotton gauze, 3 ml of ether was added. The nested PCR was performed in the same condition with The mixture was shaken for a minute and centrifuged at 2,000 the addition of primers MSP3, MSP4a, and MSP4b on 2 μl g for 2 min. From the sediment, a thin smear was prepared on of the first PCR product. The final PCR product was resolved a microscope slide, allowed to air dry, and fixed in methanol by 2% agarose gel electrophoresis with TBE buffer (1 M Tris for 5 min. base, 0.8 M boric acid, 0.01 M ethylenediamine tetra acetic The slides were again allowed to air dry and stained with acid, pH 8.0) at 100 v. The gel was read under UV illumina- Uvitex-2B (Ciba-Giegy, Basel, Switzerland) for 10 min. Then tion using an instant camera (Polaroid, Waltham, Mass., USA). the slides were washed in PBS and counterstained with Evans blue (Sigma Chemical, St. Louis, Mo., USA) for 30 s, RESULTS further washed in PBS for 5 s, and gently rinsed in tap water. Finally, the slides were examined under a fluorescent Out of 243 diarrheal stool samples, 39 (16.0%) cases were microscope with a 50-W mercury high-pressure lamp, an positive for intestinal microsporidial infections by PCR. Of excitation filter transmitting 355 - 425 nm, and a suppression these, only 18 (7.6%) cases detected by microscopy were filter of 460 nm (Leitz Ploemopak Filter Block D; Ernst Leitz also positive by PCR (Table 1). All 39 (18.2%) of the cases GmbH, Germany) at the magnification of ×1,000. Each slide were from among the 214 HIV/AIDS diarrheal patients. was observed for about 10 min to determine whether it was Based on PCR and microscopic analyses, the microsporidial positive or negative. parasites were identified as E. bieneusi (30), E. intestinalis HIV serology: About 7 -10 ml of whole blood was col- (6), and double infections (3) (Figure 1). The of E. lected in a labeled vacutainer tube containing EDTA. HIV-1 intestinalis were larger than those of the E. bieneusi and serology was done using ELISA (Vironostika®, HIV-Uni- varied considerably in both shape and size, either broad and Form II Plus O; Biomerieux, Boxtel, the Netherlands) and rod-like or kidney-shape (Figures 2 and 3). reactive samples were confirmed by Western blot assay (HIV Of the 39 patients with microsporidial infections, 36 Blot 2.2; Genelabs Diagnostics, The Cavendish Singapore (92.3%) had diarrhea for more than 4 weeks. The majority Science Park, Singapore). were male and all were adults, with an age range of 24-48 CD4+ cell count: The CD4+ cell count was determined years (mean 36 years). Severe (>10%) weight loss was by a FACScan flow cytometer (Becton Dickinson, San recorded in 37 (94.4%) of the cases. Other clinical indications Jose, Calif., USA) and analyzed with Simulset software were found, including: anorexia in 29 (74.4%), abdominal (Becton Dickinson) included with the FACScan, following pain with cramping in 25 (64.1%), and vomiting in 18 (46.2%), the manufacturer’s instruction. among others (Table 2). Viral load determination: The HIV-1 viral load was Concurrent infection with other opportunistic intestinal determined by measuring the HIV-1 RNA in plasma samples parasites was also observed. These included C. parvum in using Nucleic Acid Sequence-Based Amplification, Nuclisens 4/39 and I. belli in 4/39 cases. NASBA (Organon Teknika BV, Boxtel, the Netherlands). The CD4+ cell count was below 100 cells/mm3 in 27/32 DNA extraction and polymerase chain reaction (PCR) (84.4%) of the cases, and 59.4% of the patients had cell counts for detection of intestinal microsporidia from stool samples: of less than or equal to 50 cells/mm3, with a mean of 22.8 DNA was isolated from fresh stool suspension prepared in cells/mm3. A few individuals had counts greater than 100 200 μl PBS containing 2% polyvinylpolypyrolidone (PVPP) (Sigma) and heated for10 min in a heat block at 100°C. Table 1. Results of microscopy and PCR for intestinal Sodium dodoecylsulphate-proteinase K was added and microsporidiosis on 243 stool samples allowed to incubate in the heat block at 55°C for 2 h. After PCR Microscopy incubation, DNA was extracted by using the QIAamp Tissue Positive Negative Total kit spin columns in an Eppendorf microcentrifuge (Qiagen, Hilden, Germany) following the manufacturer’s instructions. Positive 18 0 18 Negative 21 204 225 Amplification: DNA was amplified by nested PCR as de- scribed previously (21). In the first PCR, three outer primers Total 39 204 243

307 Fig. 1. An agarose gel showing PCR products amplified from human fecal samples by nested microsporidium PCR. Lane 1, 100 bp marker; lane 17, negative control; lane 16, Enterocytozoon bieneusi positive control; lane 15, Encephalitozoon intestinalis positive control; lane 4, E. bieneusi positive product; lane 9-10, product of double infection. Fig. 2. Encephalitozoon intestinalis spores from stool sample of HIV positive diarrhea patient stained with Uvitex-2B under fluorescent Table 2. Demographic and clinical features of diarrheal HIV/ microscope (magnification ×1,000). AIDS patients with intestinal microsporidiosis Features No. (%) (n = 39) Outpatient 10 (25.6) Inpatient 29 (79.4) Age 24-48 (mean 36 years) Sex Male 32 (82.1) Female 7 (17.9) Marital Single 9 (23.1) Married 27 (69.2) Widowed 3 ( 7.7) Diarrhea duration 1-4 weeks 3 ( 7.7) Fig. 3. Enterocytozoon bieneusi spores from stool sample of HIV posi- >4 weeks 36 (92.3) tive diarrhea patient stained with Uvitex-2B under fluorescent microscope (magnification 1,000). Bowl motion per day × 2-6 BM/D 35 (89.8) NASBA was 80 HIV-1 RNA copies/ml. >6 BM/D 4 (10.2) Appearance of diarrhea Watery 31 (79.5) DISCUSSION Loose 7 (17.9) Mucoid 1 ( 2.6) Diarrhea is a common clinical feature in HIV-1-infected Bloody 0 ( 0.0) individuals. In tropical and subtropical countries, HIV-infected CD4+ cell count (cells/mm3) individuals had chronic diarrhea associated mostly with ≤50 19/32 (59.4) significant weight loss (14,16,17,22). The aetiology of diar- 51-100 8/32 (25.0) rhea in HIV/AIDS patients is quite various: bacterial, viral, 101-200 5/32 (15.6) parasitic, mainly opportunistic intestinal parasites or HIV- Viral load (copies/ml) induced entropathy (23). Similarly, in this study most of the <10,000 3/28 (10.7) HIV/AIDS patients with microsporidial infection had severe ≥10,000 25/28 (89.3) weight loss (>10% of body weight) and chronic diarrhea Stage I 0 lasting more than 4 weeks. In the present study, the majority II 3 ( 7.6) of the parasites identified were E. bieneusi (77%) followed III 18 (46.2) by E. intestinalis (15.4%), excluding the 3 cases (7.7%) of IV 18 (46.2) double infection. This finding partially deviates in magni- Clinical indications tude from previous findings, where 90% of chronic diarrhea Weight loss >10% 37 (94.4) in HIV/AIDS patients caused by intestinal microsporidiosis Abdominal pain 25 (64.1) was attributed to E. bieneusi (14,18,19). Based on this obser- Vomiting 18 (46.2) vation, in Ethiopia the cause of chronic diarrhea in HIV/AIDS Nausea 22 (56.4) patients might not be predominantly one species but could Anorexia 29 (74.4) be both species, with a high preponderance of E. bieneusi Fever (low grade) 19 (47.7) over E. intestinalis. Other studies elsewhere have reported Flatulence 12 (30.8) that the incidence of E. intestinalis in the diarrhea HIVAIDS patients is rare (20,24). The severity of diarrheal illness is or less than 200 cells/mm3. The majority (25/28; 89.3%) of aggravated by co-infection with coccidian intestinal parasites diarrhea patients with intestinal microsporidiosis had HIV-1 (18,25). In this study, 4 (10.2%) of C. parvum and I. belli each, plasma viral load counts of at least 10,000 copies/ml, with a were found to be co-infected with intestinal microsporidiosis mean HIV-1 RNA viral load of 22,217 copies/ml, a reflec- which might worsen the condition of diarrheal illness in some tion of low CD4+ levels. The lower detection limit of Nuclisen HIV/AIDS patients.

308 Like most intestinal parasites, the route of transmission is with protease inhibitors, which may help to reduce the risk of mainly fecal-oral contamination of food or drinking water. developing diarrhea. This makes it more difficult to reduce the risk of trans- mission of intestinal microsporidial infection in developing ACKNOWLEDGMENTS countries (1,4,6). Where antiretroviral therapy is not acces- sible, it is essential for HIV-infected individuals to maintain This study was conducted in the Ethiopian-Netherlands good personal hygiene and to handle drinking water safely AIDS Research Project (ENARP), which was a collaborative (mostly by boiling it before drinking). effort of the Ethiopian Health and Nutrition Research Insti- It has been reported that the AIDS patients who are at high- tute (EHNRI), Addis Ababa, the Municipal Health Service, est risk for intestinal microsporidiosis are those with CD4+ Amsterdam, Department of Parasitology, Leiden University cell counts below 100 cells/mm3 (16). In the present study, Medical Centre, Amsterdam, and Addis Ababa University, more than half of HIV/AIDS diarrheal patients with intesti- Ethiopia. nal microsporidiosis had CD4+ cell counts below 50 cells/ We are grateful to the technical staff members of Parasit- mm3, and the majority were below 100 cells/mm3, indicating ology Laboratory of EHNRI, ENARP, army hospital, police that infection with intestinal microsporidiosis occurred hospital, and St. Paul hospital at Addis Ababa. under a high degree of immunosuppression during the late stage of HIV infection. The majority of the diarrhea patients REFERENCES were categorized as having an HIV-1 viral load ≥10,000 copies/ml, which is directly associated with low CD4+ cell 1. Weber, R., Bryan, R.T., Shruratz, A.D. and Owen, R.L. counts. This indicates that intestinal microsporiodiosis in HIV/ (1994): Human microsporidial infection. Clin. Microbiol. AIDS patients is more apparent when patients are severely Rev., 7, 426-461. immunocompromised. 2. Bryan, T.B. (1995): Microsporiosis as an AIDS related The major clinical symptoms observed were chronic diar- opportunistic infections. Clin. Infect. Dis., 21 (Suppl. 1), rhea associated with severe weight loss. Another common 562-565. clinical symptom was anorexia. Similar clinical observations 3. Lock Wood, J.N.D. and Weber, J.N. (1987): Parasite were documented by others elsewhere (2,4). infection in AIDS. Parasitol. Today, 5, 310-315. Based on the World Health Organization clinical staging 4. Chukwuma, C. (1996): Microsporidia in AIDS patients: criteria (26), most of the intestinal microsporidia-infected A prospective. East Afr. Med. J., 73, 72-75. patients were categorized as stage III or IV. The associated 5. Goodgame, R., Stager, C., Marcantel, B., Alcoler, E. and clinical features commonly noted in the late stage of HIV Segura, A.M. (1999): Intensity of infection in AIDS infection with intestinal microsporidiosis were chronic diar- related intestinal microsporidiosis. J. Infect. Dis., 180, rhea for more than 1 month, severe weight loss, and oral 929-932. , and one individual had pulmonary tuberculosis 6. Drobnieruski, F., Kelly, P., Carew, A., Ngwenya, B., Luo, within the previous year. N., Pankhurst, C. and Farthing, M. (1995): Human In Africa, information regarding intestinal microsporidiosis microsporidiosis in African AIDS patients with chronic in HIV/AIDS patients is very limited. This is due to a lack of diarrhoea. J. Infect. Dis., 171, 515-516. appropriate diagnostic methods and to a lack of experienced 7. Sianongo, S., MacDonald, V. and Kelly, P. (2001): A and skilled heal care workers. In addition, the spores are very method for diagnosis of microsporidiosis adapted for use small; those of E. bienuesi averaged 1.5×1.0 μm and those in developing countries. Trans. Roy. Soc. Trop. Med. Hyg., of E. intestinalis were 2.2 × 1.2 μm (7); these values are 95, 605-607. consistent with our observation. In the present study, PCR 8. Conteas N.C., Berlin W.G.O., Speck, E.E., Pundhumas, detected more positive microsporidia cases than did light S.S., Lariviere, MJ. and Fu, C. (1998): Modification of microscopy; this was also similar with a previous study the clinical course of intestinal in AIDS patients by conducted elsewhere (21). Moreover, when stool contained immune status and anti-HIV therapy. Am. J. Trop. Med. undetectable levels of spores, the light microscope would not Hyg., 58, 555-558. have been useful. Similar observation was recorded by Gumbo 9. DeGirolami, G.P., Ezratty, R.C. and Desal, G. (1995): et al. (19) in Zimbabwe, where microscopy detected intesti- Diagnosis of intestinal microsporidiosis by the exami- nal microsporidia in 18% of diarrhea patients, whereas PCR nation of stool and duodenal aspirates with Weber’s detected them in 51%. In the same country, Van Gool et al. modified trichrome and Uvtex-2B stains. J. Clin. (18) reported 11% of intestinal microsporidia using the Microbiol., 33, 800-810. Uvitex-2B staining method in AIDS patients. Both studies 10. Kotler, D.P. and Orenstein, J.M. (1998): Clinical syn- detected E. bieneusi. The other importance of PCR is that dromes associated with microsporidiosis. Adv. Parasitol., the use of appropriate PCR primers enables species identifi- 40, 322-343. cation (27), so that it could be important for treating E. 11.Garcia, L.S. (2002): Laboratory identification of the intestinalis with albendazole (25,28). So far, there is no widely microsporidia. J. Clin. Microbiol., 40, 1892-1901. recognized and applied treatment for E. bieneusi. 12. Desportes, I., Charpentier, Y. and Galian, A. (1985): Occur- This study demonstrates that intestinal microsporidiosis (E. rence of a new microsporidia: Enterocytozoon bieneusi bieneusi and E. intestinalis) is an important cause of chronic n.g., n.sp. in the entrocytes of human patients with AIDS. diarrhea in AIDS patients in Ethiopia. Furthermore, cases J. Protozoal., 32, 250-254. of intestinal microsporidiosis appear to be associated with 13. Lucas, B.S., Papadaki, L. and Conlon, C. (1989): Diag- severe immunosuppression, i.e., with CD4+ cell counts below nosis of intestinal microsporidiosis with AIDS. J. Clin. 50 cells/mm3, where E. bieneusi is the dominant species. Thus, Pathol., 42, 885-890. it is imperative to treat this kind of parasite along with 14. Dallabetta, A.G. and Miotti, G.P.(1992): Chronic diar- other opportunistic intestinal parasites by initiating HAART rhoea in AIDS patients in tropics: a review. Trop. Doc.,

309 22, 3-9. species determination of microsporidia DNA from stool 15. Dascomb, K., Clark, R., Aberg, J., Pulvienti, J., Hewitt, specimens. Trop. Med. Intern. Health, 1, 373-378. R.G., Kissinger, P. and Diders, E.S. (1999): Natural 22. Rebeneck, L., Gyorkey, F., Genta, R.M., Gyorkes, P., history of intestinal microsporidiosis among patients Foote, L.W. and Risser, M.H. (1993): The role of infected with HIV. J. Clin. Microbiol., 37, 3421-3422. microsporidia in the pathogenesis of HIV related chronic 16. Asmuth, M.D., DeGirolami, C.P., Federman, M., Ezratty, diarrhoea. Ann. Intern. Med., 119, 895-899. C.R., Plesknow, D.K., Dasai, G. and Wanke, C.A. (1994): 23. Germani, Y., Minssart, P., Vohito, M., Yassibanda, S., Clinical features of microsporidiosis in patients with Glazion, P., Hocqvet, D., Berthelemy, P. and Morvan, J. AIDS. Clin. Infect. Dis., 18, 819-825. (1998): Aetiologies of acute, persistent and dysenteric 17. Navin, R.T., Weber, R., Vugia, J.D., Rimland, D., diarrhoea in adults in Bangui, Central African Republic, Roberts, J.M., Addis, G.D., Visvesvara, G.S., Wahlquist, in relation to HIV-serostatus. Am. J. Trop. Med. Hyg., S.P., Hongan, S.E., Gallagher, L.E., Juranck, D.D., 59, 1008-1014. Schwartz, D.A., Wilcox, C.M., Stewart, J.M., Thompson 24. Molina, J.M., Oksenhender, E., Beavuais, B., Sarfati C., III, S.E. and Bryan, R.T. (1999): Declining CD4+ T- Jaccord, A., Derouin, N. and Moday, J. (1995): Dissemi- lymphocyte counts are associated with increased risk of nated microsporidiosis due to Septata intestinalis in enteric parasitosis and chronic diarrhoea: results of a year patients with AIDS: clinical features and response to longitudinal study. J. Acquired Immune Defic. Syndr. Albendazole therapy. J. Infect. Dis., 171, 245-249. Hum. Retrovirol., 20, 154-159. 25. Tarimo, S.D., Killeno, J.Z.J., Mirij, A.S. and Masmanga, 18. Van Gool, T., Lundehoff, E., Nathuoo, K.J., Kiire, C.F., I.G. (1996): Prevalence of intestinal parasites in adults Dankert, J. and Manson, P.R. (1995): High prevalence patients with entropathic AIDS in North-Eastern Tanza- of Enterocytozoon bieneusi infections among HIV- nia. East Afr. Med. J., 7, 397-399. positive individuals with persistent diarrhoea in Harara, 26. The World Health Organization International Collabo- Zimbabwe. Trans. Roy. Soc. Trop. Med. Hyg., 89, 478- rating Group for the Study of the WHO Staging System 480. (1993): Proposed World Health Organization staging 19. Gumbo, T., Sarbah, S. and Gangaidoz, T.T. (1999): system for HIV infection and disease: preliminary testing Intestinal parasites in patients with diarrhoea and HIV by an international collaborative cross-sectional study. infected in Zimbabwe. AIDS, 13, 819-822. AIDS, 7, 711-718. 20. Van Gool, T., Cunning, E.U. and Denkert, J. (1994): 27. Fedorko, D.P. and Hiyazi, Y.M. (1996): Application of An improved practical and sensitive technique for the molecular techniques to the diagnosis of microsporidial detection of microsporidia spores in stool samples. Trans. infection. Emerg. Infect. Dis., 2, 183-191. Roy. Soc. Trop. Med. Hyg., 88, 189-190. 28. Conteas, N.C., Berlin, W.G.O., Ash, L.R. and Peruth, 21. Katzwnikel-Wladarsch, S., Lieb, M., Heise, W., Loscher, J.S. (2000): Therapy for human gastrointestinal micro- T. and Rinder, H. (1996): Direct amplification and sporidiosis. Am. J. Trop. Med. Hyg., 63, 121-127.

310