LETTERS nationally agreed upon. This case 2. John TJ, Jesudason MV, Lalitha MK, Study design and methods have illustrates another potential pitfall in Ganesh A, Mohandas V, Cherian T, et al. been described elsewhere (5,6). Melioidosis in India: the tip of the iceberg? melioidosis serodiagnosis. Most iso- Indian J Med Res. 1996;103:62–5. Twenty-four children that were peri- lates express a conserved LPS antigen, 3. John TJ. Melioidosis, the mimicker of mal- natally infected with HIV were which allows use of a single reference adies. Indian J Med Res. 2004;119:vi–viii. included in this substudy. During strain for determination of anti-LPS 4. Currie BJ, Jacups SP. Intensity of rainfall monthly visits during the child’s first and severity of melioidosis, Australia. antibodies (8). However, because Emerg Infect Dis. 2003;9:1538–42. 2 years of life, malaria and HIV inci- some strains express different LPS 5. Jesudason MV, Anbarasu A, John TJ. dence were recorded (5,6). Both chil- antigens, serologic tests must be per- Septicaemic meliodosis in a tertiary care dren with malaria-positive blood formed with the patient’s own strain. hospital in south India. Indian J Med Res. smears and those with fever but no 2003;117:119–21. This case illustrates potential pit- 6. Dance DAB. Melioidosis: the tip of the ice- smear result available were treated falls in diagnosing melioidosis, which berg? Clin Microbiol Rev. 1991;4:52–60. with sulfadoxine-pyrimethamine requires clinical and laboratory aware- 7. Novak RT, Glass MB, Gee JE, Gal D, Mayo according to national guidelines. At ness and knowledge of its geographic MJ, Currie BJ, et al. Development and eval- the time of this substudy, none of the uation of a real-time PCR assay targeting distribution. LPS-based serologic the type III secretion system of study participants were taking anti- assays should use a range of isolates Burkholderia pseudomallei. J Clin retroviral drugs. representative of known LPS types. Microbiol. 2006;44:85–90. HIV and malaria diagnoses were 8. Pitt TL, Aucken H, Dance DA. determined by using standard meth- Homogeneity of lipopolysaccharide anti- Acknowledgment gens in Pseudomonas pseudomallei. J ods (5–7). To reduce the chance of We are grateful to the patient for per- Infect. 1992;25:139–46. including infants infected through mission to publish this case report. breast-feeding, perinatal infection was Address for correspondence: Andrew J. Brent, defined as >2 consecutive HIV-posi- A. J. Brent and R. Handy had clini- Nuffield Department of Infectious Diseases & tive tests, with the first positive PCR cal responsibility for the patient. R. Microbiology, John Radcliffe Hospital, result by 4 months of age (7). The so- Handy and P.C. Matthews made the initial Headington, Oxford, United Kingdom OX3 called baseline viral load was the pre- microbiologic diagnosis of melioidosis, 9DU; email: dr.a.brent@ gmail.com malaria value measured 1 month and T.L. Pitt confirmed the isolate as before the first observation in the Burkholderia pseudomallei. T.L. Pitt per- analysis. To be included in the analy- formed the serology and SDS-PAGE sis, follow-up visits had to have data analysis of lipopolysaccharide antigens. available on the current and previous All authors contributed to preparation of months’ viral load and malaria status the manuscript. A.J. Brent is guarantor for and occur at roughly monthly inter- the article, had full access to all the clini- Subclinical vals at >4 months of age. cal and microbiologic data, and had final Plasmodium Malaria parasites were found at 53 responsibility for the decision to submit falciparum of 146 visits in the month before viral for publication. Infection and HIV-1 load measurement, although at 89% of visits in which children were Andrew J. Brent,* Philippa C. Viral Load malaria-positive, the children’s sam- Matthews,* David A. Dance,† To the Editor: Studies indicate ples had <1,000 parasites/µL, and in Tyrone L. Pitt,‡ and Rupert Handy§ that Plasmodium falciparum infection only 13% of visits in which children *John Radcliffe Hospital, Headley Way, increases HIV replication in adults had parasitemia did they also have Headington, Oxford, United Kingdom; (1,2). Although malaria-related illness fever (8). Median number of observa- †Health Protection Agency (South West), Plymouth, Devon, United Kingdom; and death are more common in chil- tions per child was 7 (range 2–18). No ‡Laboratory of HealthCare Associated dren, and HIV-1 generally progresses significant demographic or clinical Infection, London, United Kingdom; and faster in children than in adults (3,4), differences were found between HIV- §Heatherwood and Wexham Park to our knowledge the effect of inter- positive children in this substudy and Hospitals, Wexham, Slough, United mittent malaria on HIV-1 viral load those enrolled in the full cohort (data Kingdom has not been directly explored in chil- not shown). dren. To investigate this issue, we Clinical and demographic vari- References monitored HIV-positive infants from ables were evaluated in univariate 1. Dance DA, Smith MD, Aucken HM, Pitt a 1996–2001 birth cohort study in repeated measures analysis to deter- TL. Imported melioidosis in England and Kisumu, Kenya, a P. falciparum– mine associations with log-trans- Wales. Lancet. 1999;353:208. holoendemic area. formed HIV-1 viral load. Age and

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 2, February 2007 351 LETTERS baseline viral load were strong predic- different HIV/malaria interactions through conducting monthly surveil- tors of current load (Table). Although than seen in adults. lance and encouraging mothers to not statistically significant, clearing Viral load increased in infants that bring children in during febrile the previous month’s malaria infec- were incorrectly treated presumptive- episodes, ability to assess the effect of tion was associated with a drop in ly (due to fever) for malaria (Table). high-density malaria was diminished viral load (Table, p = 0.09). It was not Most of these children were found to because parasitemia levels never possible to distinguish between the have other infections. Fever in malar- reached clinically significant levels. effects of treatment versus malaria ia-endemic areas is often assumed to Finally, because malaria was diag- clearance because 87% of malaria be malaria-related, but delay in treat- nosed by microscopy, rather than infections were treated with anti- ment of nonmalarial infections may PCR, some subclinical malaria infec- malarial drugs. However, viral load be harmful in HIV-infected children tions may have been misclassified as increased in those incorrectly treated Our assessment was limited in size malaria-negative. for malaria presumptively (Table). and duration. Furthermore, in attempt- Although we found no evidence of After adjusting for age and base- ing to provide optimal patient care an association between subclinical, line viral load, we assessed log10 HIV viral load in relation to malaria clear- ance, persistence, absence, or new infection using a repeated measures model with autoregressive covariance structure. No differences were statisti- cally significant, although clearing an infection versus no malaria had a 0.22 log viral load decrease (Table, p = 0.10). When 15 malaria episodes with malaria-free visits 1 month before and after the episode were compared, mean difference (signed-rank test) in viral load “before” and “after” malar- ia was not significant. Our findings suggest that low-den- sity malaria infection may not dramat- ically affect plasma HIV-1 levels in infants. This finding is similar to results of studies of perinatally HIV- infected children in which, although viral loads were unavailable, number of malaria episodes did not signifi- cantly affect development of AIDS- related symptoms (9,10). While clini- cal malaria leads to at least short-term HIV viral load increases in adults (1,2), the effect of subclinical malaria is unclear, and even less is known about coinfection in children. Children usually have higher baseline viral loads than adults; thus, the rela- tive effect of malaria on viral load may not be as great. To reduce the impact of passively transferred mater- nal antibodies, analyses were done on visits after the child was 4 months old. However, lack of fully acquired anti- malarial immunity may have led to

352 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 2, February 2007 LETTERS low-density malaria and infant HIV-1 RNA blood plasma concentration. AIDS. African Tickbite viral load, the consequences of high- 1999;13:487–94. 2. Kublin JG, Patnaik P, Jere CS, Miller WC, Fever in Travelers, density or clinical malaria need to be Hoffman IF, Chimbiya N, et al. Effect of explored. If clinical malaria in infants Plasmodium falciparum malaria on concen- Swaziland increases HIV-1 viral load as it does tration of HIV-1-RNA in the blood of adults in adults (1,2), our study underscores in rural Malawi: a prospective cohort study. To the Editor: African tickbite Lancet. 2005;365:233–40. dual benefits of malaria treatment in fever (ATBF), which is caused by 3. Rogers MF, Thomas PA, Starcher ET, Noa Rickettsia africae, is well documented the context of HIV: 1) keeping malar- MC, Bush TJ, Jaffe HW. Acquired immun- in travelers to southern Africa (1–3) ia in check, and 2) preventing an odeficiency syndrome in children: report of increase in HIV viral load. Ethical the Centers for Disease Control National and transmitted by ungulate ticks of Surveillance, 1982 to 1985. Pediatrics. issues prevent prospective studies to the genus Amblyomma. Positive sero- 1987;79:1008–14. logic results were reported in 9% of assess the impact of coinfection early 4. Pizzo PA, Wilfert CM. Markers and deter- patients (1) and 11% of travelers (4) in life, but alternatives include using minants of disease progression in children animal models or stored specimens. with HIV infection. The Pediatric AIDS from southern Africa. We report an Siena Workshop II. J Acquir Immune Defic outbreak of ATBF with an attack rate Syndr Hum Retrovirol. 1995;8:30–44. of 100% among 12 Dutch travelers to Acknowledgments 5. Ayisi JG, van Eijk AM, ter Kuile FO, Swaziland. We are grateful to the Kenyan chil- Kolczak MS, Otieno JA, Misore AO, et al. The 12 travelers (9 male and 3 dren who participated in this study and Risk factors for HIV infection among asymptomatic pregnant women attending female) visited Mkhaya Game their families. We also thank the field and an antenatal clinic in western Kenya. Int J Reserve in Swaziland in May 2003 for laboratory staff of the Centers for Disease STD AIDS. 2000;11:393–401. several days. Upon retuning to the Control and Prevention (CDC)/Kenya 6. Ayisi JG, van Eijk AM, Newman RD, ter Netherlands, they consulted our clinic Medical Research Institute (KEMRI). We Kuile FO, Shi YP, Yang C, et al. Maternal malaria infection and perinatal HIV trans- for assessment for fever, malaise, and appreciate Davy Koech, director of mission, western Kenya. Emerg Infect Dis. skin eruptions. Epidemiologic and KEMRI, for his support and approval with 2004;10:643–52. clinical data were obtained after the regard to publication of this article, and 7. Yang C, Li M, Newman RD, Shi YP, Ayisi patients provided informed consent. Feiko ter Kuile and Laurence Slutsker for J, van Eijk AM, et al. Genetic diversity of HIV-1 in western Kenya: subtype-specific All symptomatic patients were treated their helpful comments. differences in mother-to-child transmission. before serum samples were collected. AIDS. 2003;17:1667–74. This study was supported by 8. Bloland PB, Boriga DA, Ruebush TK, Acute-phase and convalescence- AOT0483-PH1-2171 from the United McCormick JB, Roberts JM, Oloo AJ, et al. phase serum samples were obtained States Agency for International Develop- Longitudinal cohort study of the epidemiol- from 8 patients at 3 and 9 weeks, ment. The Institutional Review Boards of ogy of malaria infections in an area of respectively, after symptoms were intense malaria transmission II. Descriptive CDC and KEMRI approved the methods. epidemiology of malaria infection and dis- reported. Only convalescent-phase ease among children. Am J Trop Med Hyg. serum samples were obtained from Kimberly C. Brouwer,* 1999;60:641–8. the other 4 patients. Serologic assays 9. Greenberg AE, Nsa W, Ryder RW, Medi M, were conducted for screening and Lisa B. Mirel,* Chunfu Yang,* Nzeza M, Kitadi N, et al. Plasmodium fal- Renu B. Lal,* ciparum malaria and perinatally acquired confirmation in Rotterdam, the Margarette S. Kolczak,* human immunodeficiency virus type 1 Netherlands (Department of Virology, Anne M. Van Eijk,† John Ayisi,† infection in Kinshasa, Zaire. A prospective, Erasmus University Hospital) and longitudinal cohort study of 587 children. N Marseille, France (Unité des Juliana A. Otieno,‡ Engl J Med. 1991;325:105–9. Bernard L. Nahlen,*§ 10. Kalyesubula I, Musoke-Mudido P, Marum Rickettsies, Faculté de Médecine, Richard Steketee,* Ya Ping Shi,*† L, Bagenda D, Aceng E, Ndugwa C, et al. Université de la Mediterranée), and Altaf A. Lal* Effects of malaria infection in human respectively. immunodeficiency virus type 1-infected In Rotterdam, immunofluores- *Centers for Disease Control and Ugandan children. Pediatr Infect Dis J. Prevention, Atlanta, Georgia, USA; 1997;16:876–81. cence assays for immunoglobulin G †Kenya Medical Research Institute, (IgG) and IgM against R. conorii, R. Kisumu, Kenya; ‡New Nyanza Provincial Address for correspondence: Kimberly C. typhi, and R. rickettsii were per- General Hospital, Kisumu, Kenya; and formed with multiwell slides on §World Health Organization, Geneva, Brouwer, Division of International Health and Switzerland Cross-Cultural Medicine, Department of which antigens were fixed (Panbio Family and Preventive Medicine, School of Inc., Columbia, MD, USA). Serum References Medicine, University of California, San Diego, samples with fluorescent rickettsiae 1. Hoffman IF, Jere CS, Taylor TE, Munthali 9500 Gilman Dr, MC 0622, La Jolla, CA at dilutions >1:32 were considered P, Dyer JR, Wirima JJ, et al. The effect of 93093-0622, USA; email: [email protected] positive. Plasmodium falciparum malaria on HIV-1

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 13, No. 2, February 2007 353