The spread of tick-borne borreliosis in West Africa and its relationship to sub-Saharan drought

In West Africa, tick-borne relapsing fever is due to the spirochete Borrelia crocidurae and its geographic distribution is classically limited to the Sahel and Saharan regions where the vector tick Alectorobius sonrai is distributed. We report results of epidemiologic investigations carried out in the Sudan savanna of Senegal where the existence of the disease was unknown. A two-year prospective investigation of a rural community indicated that 10% of the study population developed an infection during the study period. Transmission patterns of B. crocidurae to humans and the small wild mammals who act as reservoirs for infection were similar to those previously described in the Sahel region. Examination of 1,197 burrows and blood samples from 2,531 small mammals indicated a considerable spread of the known area of distribution of A. sonrai and B. crocidurae. The actual spread of the vector and the disease has affected those regions where the average rainfall, before the start of the extended drought in West Africa, reached up to 1,000 mm and corresponds to the movement of the 750-mm isohyet toward the south from 1970 to 1992. Our findings suggest that the persistence of sub-Saharan drought, allowing the vector to colonize new areas in the Sudan savanna of West Africa, is probably responsible for a considerable spread of tick-borne borreliosis in this part of Africa.

Reducing the oral quinine-quinidine-cinchonin (Quinimax) treatment of uncomplicated malaria to three days does not increase the recurrence of attacks among children living in a highly endemic area of Senegal

A 3 d shortened course of the quinine-quinidine-cinchonin association Quinimax was compared to the usual 7 d regimen for routinely treating 462 acute uncomplicated Plasmodium falciparum malaria attacks in 72 children under the age of 10 years in Dielmo, a holoendemic village in Senegal. 25 mg/kg Quinimax salt daily, given in 3 equal doses, improved clinical status in 99.6% of the patients receiving the course and in all of those treated for 7 d. Even if the 3 d course did not systematically eliminate parasitaemia, reducing oral Quinimax treatment of uncomplicated malaria from 7 to 3 d did not increase the recurrence of attacks, even among the youngest children. Both the quinine sensitivity of the Senegalese strains of P. falciparum and the partial acquired immunity of the children were probably responsible for the absence of any difference between the courses. Oral Quinimax for 3 d is a possible alternative regimen to chloroquine and sulfadoxine-pyrimethamine for treating uncomplicated malaria in highly endemic areas of Africa where clinical resistance to these drugs exists.

HLA-A, -B, -C, -DR, -DQ typing in a population group of Senegal: distribution of HLA antigens and HLA-DRB1*13 and DRB1*11 subtyping by PCR using sequence-specific primers (PCR-SSP)

One-hundred-and sixteen Senegalese Serere were typed for HLA antigens and compared with other ethnic groups in Gambia. We did not find significant differences (Fisher's exact test; P < 0.01) in the HLA antigens distribution between the Serere and Mandinka groups in Senegal and the Serere, Mandinka and Wolof in The Gambia. The most common HLA haplotypes found (P < 0.01; Chi square with Yates' correction) were: A1, B8; A2, B51; A32, B44; A33, B58; A2, Cw2; A2, Cw4; A33, Cw3; A2, DR17; A10, DR10; B35, Cw4; B53, Cw6; B57, Cw3; B65, Cw8; B50, DR15; B52, DR4; Cw2, DR17; DR7, DQ2; DR18, DQ4. The HLA-DRB1*13 and DRB1*11 alleles were subtyped by PCR-SSP and the frequencies of these alleles in the studied population given. HLA-DRB1*1304 and DRB1102 were the most common alleles found respectively 15.0 and 18.5%.

Plasmodium falciparum induces apoptosis in human mononuclear cells

The level of spontaneous apoptosis in short-term lymphocyte cultures was evaluated in different human immunodeficiency virus-negative groups of either healthy control individuals or patients with clinical malaria. The mean percentage of spontaneous apoptosis found in patients during a malaria attack was significantly higher than in sex- and age-matched healthy controls. The healthy asymptomatic controls were individuals with different degrees of exposure to Plasmodium falciparum as reflected by their various mean levels of specific anti-P. falciparum (immunoglobulin G and M) antibodies. The percentages of apoptotic nuclei were found to be significantly higher in lymphocytes from subjects living in an area where malaria is holoendemic than in lymphocytes from subjects less exposed. Concentrations of soluble plasma interleukin-2 receptor were also higher in subjects from areas where malaria is endemic than in other groups, revealing different levels of lymphocyte activation. Of particular relevance to the in vivo situation, a P. falciparum schizont-rich extract induced a systematic and significant elevation of apoptotic nuclei at day 6 in 87.5% (35 of 40) of the subjects tested. In additional studies with different concentrations of extract, [3H]thymidine incorporation was concomitant with a low or limited level of apoptosis. Taken together, our results strongly suggest that acute as well as chronic asymptomatic P. falciparum infections were consistently associated with a marked increase in the level of mononuclear cell apoptosis. This process could be implicated in some of the alterations reported for the proliferative T-cell responses in areas where malaria is endemic.

Rapid turnover of Plasmodium falciparum populations in asymptomatic individuals living in a high transmission area

A polymerase chain reaction (PCR) typing technique, based on the amplification of polymorphic regions from the merozoite surface protein 1 (MSP-1) and MSP-2 Plasmodium falciparum genes, was used to characterize parasites collected in a longitudinal study of asymptomatic carriers of malaria parasites living in two distinct epidemiologic situations. Blood samples were collected from children and adults living in the village of Dielmo, Senegal, when malaria transmission was 3-6 infective bites/week/individual. For each individual, every sample collected at two-week intervals over a period of three months showed a specific PCR pattern. Changes involved both appearance and disappearance of specific alleles. Analysis of blood samples collected at a few-days interval showed that modifications of the PCR patterns occurred rapidly. Most alleles were detected over a period of 2-3 weeks, but some alleles could be detected only for a few days. The frequent modifications of the PCR patterns indicate significant changes in allelic balance over time, and importantly, this was observed both in children and adults. These results strongly contrast with the stability of the parasite types harbored by asymptomatic individuals living in Pikine, Senegal during a period in which malaria transmission was interrupted, and therefore suggest that the rapid turnover observed in Dielmo may reflect the introduction of new parasite populations by mosquitoes.

Malaria: even more chronic in nature than previously thought; evidence for subpatent parasitaemia detectable by the polymerase chain reaction

In high endemicity areas, malaria is a chronic disease: examination of blood films reveals that up to half of the population, particularly children, harbour parasites at any one given time. The parasitological status of the remainder was addressed using the polymerase chain reaction, a technique 100 to 1000 times more sensitive than microscopy, on a series of samples from Dielmo, a holoendemic area of Senegal. Two-thirds of the microscopically negative individuals were found to harbour subpatent levels of Plasmodium falciparum, suggesting that more than 90% of the exposed population at any one time, i.e. in a cross-sectional survey, are chronically infected. This also means that the range of parasite loads harboured by humans with various degrees of exposure is remarkably large, probably reflecting a large range of effectiveness of the defence mechanisms against malaria parasites, none of which is fully efficient.

Age-dependent carriage of multiple Plasmodium falciparum merozoite surface antigen-2 alleles in asymptomatic malaria infections

Genetic diversity of the merozoite surface antigen-2 gene of the human malaria parasite Plasmodium falciparum has been analyzed in a Senegalese village where malaria is holoendemic. A cross-sectional survey of 65 residents was performed in 1992 during the high transmission season. Plasmodium falciparum was detected both by microscopy (77% positive samples) and DNA amplification using a single (29% or 38% positive samples, depending on the primers used) or nested polymerase chain reaction (PCR) (78% positive samples). The overlap between the positive nested PCR and microscopic examination was not complete. The PCR fragments were analyzed for size polymorphism on agarose gels, and were subsequently assigned to the major allelic families 3D7 or FC27 by hybridization with family-specific probes. Both allelic families were found, with a slightly higher prevalence for FC27. Chimeric alleles that failed to hybridize under stringent conditions to the reference probes were also observed. Some were typed using a novel PCR approach, using hybrid pairs of primers, consisting of a family-specific sense oligonucleotide combined with an antisense oligonucleotide specific for the other family. Combining typing techniques, 82% of the positive PCR results yielded more than one band. Both the overall number of fragments and the number of allelic types per carrier were markedly reduced around the age of 15 years. The number of DNA fragments decreased abruptly from an average of four per carrier before the age of 15 years to an average of two in individuals more than 15 years of age.(ABSTRACT TRUNCATED AT 250 WORDS)

High prevalence of natural antibodies against Plasmodium falciparum 83-kilodalton apical membrane antigen (PF83/AMA-1) as detected by capture-enzyme-linked immunosorbent assay using full-length baculovirus recombinant PF83/AMA-1

The 83-kilodalton (kD) apical membrane antigen of Plasmodium falciparum (PF83/AMA-1) is a potential asexual blood stage vaccine component. This antigen has been expressed as a full-length, nonfusion, recombinant baculovirus protein (PF83-7G8-1) using the authentic predicted signal peptide for appropriate postsynthetic routing. When purified by a novel high-performance, ion exchange chromatography (HPIEC) method, PF83-7G8-1 induced polyclonal antibodies in rats that immunoprecipitated both 83- and 66-kD forms of PF83/AMA-1 from 35S-methionine metabolically labeled parasite extracts. Using HPIEC-purified PF83-7G8-1 in combination with a rat monoclonal antibody against the highly conserved carboxy-terminal (CT) region of PF83/AMA-1, we developed a CT-capture-enzyme-linked immunosorbent assay to measure naturally acquired responses against the entire PF83/AMA-1 molecule. Analysis of populations from villages in Guinea-Bissau and in an area of high malarial transmission in Senegal demonstrated a very high prevalence (94-100%) of naturally acquired serum IgG responses to PF83/AMA-1. Analysis of these natural responses showed that PF83/AMA-1 may be a well-recognized asexual parasite antigen. A statistically significant age-related change in antibody levels to PF83/AMA-1 was observed in Guinea-Bissau. No such correlation was observed in the Senegalese population, although an age-related antibody response was seen for total parasite antigen. No significant correlation was observed between PF83/AMA-1 responses and the parameters of parasite load and malaria-related fever.

[Biology of the vectors and transmission of Plasmodium falciparum, P. malariae and P. ovale in a village in the savanna of west Africa (Dielmo, Senegal)]

From April 1990 to March 1992 a longitudinal entomological study was carried out in Dielmo village, Senegal, an area of Sudan-type savanna. Mosquitoes were sampled by night-bite collections and pyretnrum spray collections. Seven anopheles species were identified: An. gambiae s.s. An. arabiensis, An. funestus, An. pharoensis, An. rufipes, An. squamosus and An. ziemanni. Present throughout the year, An. gambiae s.l. and An. funestus represented more than 98% of anopheles captured on man. A yearly wave of An. gambiae s.l. was observed in the rainy season and An. funestus was generally more abundant in the dry season. The sporozoite rate was 1.5% and 1.3%, respectively, for these two species. Sporozoite typing by monoclonal antibodies indicated that the proportion of infected salivary glands was 92.7% P. falciparum, 18.2% P malariae and 8.2% P. ovale. The inoculation rate was calculated to be respectively 111, 21 and 8 infective bites per human for P. falciparum, P. malariae and P. ovale during the first year. Transmission was highest in the second year, with respectively 272, 54 and 25 infective bites per human.

The Dielmo project: a longitudinal study of natural malaria infection and the mechanisms of protective immunity in a community living in a holoendemic area of Senegal

The Dielmo project, initiated in 1990, consisted of long-term investigations on host-parasite relationships and the mechanisms of protective immunity in the 247 residents of a Senegalese village in which malaria is holoendemic. Anopheles gambiae s.l. and An. funestus constituted more than 98% of 11,685 anophelines collected and were present all year round. Inoculation rates of Plasmodium falciparum, P. malariae, and P. ovale averaged respectively 0.51, 0.10, and 0.04 infective bites per person per night. During a four-month period of intensive parasitologic and clinical monitoring, Plasmodium falciparum, P. malariae, and P. ovale were observed in 72.0%, 21.1% and 6.0%, respectively, of the 8,539 thick smears examined. Individual longitudinal data revealed that 98.6% of the villagers harbored trophozoites of P. falciparum at least once during the period of the study. Infections by P. malariae and P. ovale were both observed in individuals of all age groups and their cumulative prevalences reached 50.5% and 40.3%, respectively. Malaria was responsible for 162 (60.9%) of 266 febrile episodes; 159 of these attacks were due to P. falciparum, three to P. ovale, and none to P. malariae. The incidence of malaria attacks was 40 times higher in children 0-4 years of age than in adults more than 40 years old. Our findings suggest that sterile immunity and clinical protection are never fully achieved in humans continuously exposed since birth to intense transmission.

A comparative study of three methods of detection of Borrelia crocidurae in wild rodents in Senegal

In a rural area in Sénégal with a high incidence of tick-borne relapsing fever in humans, Borrelia crocidurae was studied in the blood and brain of wild rodents (Mastomys erythroleucus, Arvicanthis niloticus and Rattus rattus) using 3 methods: (i) direct examination of thick blood films; (ii) intraperitoneal inoculation of blood into white mice; (iii) intraperitoneal inoculation of homogenized cerebral tissue into white mice. Of the 82 rodents examined, the proportion of infected animals was respectively 2.4%, 7.3% and 14.6% for each method, and 18.3% for all 3 methods combined. Of the 12 animals with infected cerebral tissue, only 3 were found to have infected blood. These results suggest that isolated infections of the brain occur frequently in Senegalese wild rodents. Measurement of the real prevalence of B. crocidurae should therefore take into account these infections in addition to blood infections.

Aspects of immunity for the AMA-1 family of molecules in humans and non-human primates malarias

The apical membrane antigen (AMA-1) family of malaria merozoite proteins is characterised by a high degree of inter-species conservation. Evidence that the protein (PK66/AMA-1) from the simian parasite Plasmodium knowlesi was protective in rhesus monkeys suggested that the 83kDa P. falciparum equivalent (PF83/AMA-1) should be investigated for protective effects in humans. Here we briefly review pertinent comparative data, and describe the use of an eukaryotic full length recombinant PF83/AMA-1 molecule to develop a sensitive ELISA for the determination of serological responses in endemic populations. The assay has revealed surprisingly high levels of humoral response to this quantitatively minor antigen. We also show that PK66/AMA-1 inhibitory mAb's are active against merozoites subsequent to release from schizont-infected red cells, further implicating AMA-1 molecules in red cell invasion.

Lymphocyte response in vitro to Plasmodium falciparum merozoite antigens in donors from a holoendemic area

Crude merozoite antigens from P. falciparum were used to analyse the proliferative response of peripheral blood mononuclear cells from 114 inhabitants of the village of Dielmo (Senegal, West Africa), who are exposed continuously to malaria transmission. The high or low responses to merozoite antigens obtained in lymphocyte stimulation assays were correlated to the presence or absence of parasites, to the IFN-gamma production and to the HLA-phenotype. High responders produced high levels of IFN-gamma while low responders did not secrete IFN-gamma (23/27). The two HLA phenotypes HLA-B51 and HLA-DR1 were significantly associated with high response (p < 0.05).

Immune response in mouse and malaria-exposed humans to peptides derived from Pf11-1, a highly repetitive megadalton protein of Plasmodium falciparum

We have investigated the immune response against the Plasmodium falciparum gametocyte-specific antigen Pf11-1. This megadalton parasite molecule has been implicated in the process of erythrocyte rupture during gametogenesis. The molecule is composed in great part of degenerated nonapeptide motifs which are tandemly repeated several hundred times. A computer algorithm searching for T sites predicted that the entire repeat region of the Pf11-1 represents potential T cell antigenic major histocompatibility complex class II-binding sites. To test this hypothesis, synthetic peptides corresponding to two nonamer subtype repeats, differing only at two amino acid positions, were used to immunize congenic mouse strains. Both peptides were shown to contain both B and T cell epitopes. The immune response is restricted to the H-2d and H-2k haplotypes. The T cell response against the peptides appeared to be highly specific, clearly discriminating between the two similar nonamer repeat sequences, whereas the humoral response produced cross-reacting antibodies. We also investigated the humoral and T cell reactivities of P. falciparum-primed individuals in West Africa against the synthetic Pf11-1 peptides. Among 51 individuals 35 had antibodies to at least one of the two peptides and a majority of them (28) had antibodies reacting with both peptides. The cellular response was analyzed by [3H]thymidine incorporation or interferon-gamma release. There was considerable variation in the response to the two peptides. Among the human samples 36% responded to one repeat subtype, while only 13% responded to the second subtype. Interestingly, in individual donors the T cell response to both peptides are associated, suggesting that, as shown for mice, the response is restricted by a genetic element. The data obtained on the two subtypes of the nonamer repeat region suggest that the entire Pf11-1 molecule might induce an unusually heterogenous B and T cell response during natural infection in man.

Malaria morbidity among children exposed to low seasonal transmission in Dakar, Senegal and its implications for malaria control in tropical Africa

To measure morbidity due to malaria and to study its relationship with transmission and parasitemia in children living in an area of low malaria endemicity, a cohort study of 343 schoolchildren was undertaken during a one-year period in Dakar, Senegal. From parallel investigations on transmission and the frequency of malaria as a cause for outpatient visits, three different seasons were chosen for close monitoring of different clinical parasitologic, and sero-immunologic parameters. The daily incidence rates of malaria parasitemia and primary attacks were at a maximum level during the high transmission season (0.00198 and 0.00185 new cases/person/day, respectively) and decreased considerably during the season of low transmission. For each given period, the values of these two rates were close to each other, suggesting that each new infection was followed by a clinical attack. During the period of maximum transmission, clinical malaria prevalence was 1.36% and malaria was responsible for 36% of school absences due to medical reasons. At the end of the period of minimum transmission, clinical malaria prevalence was 0.15% and malaria was responsible for 3% of school absences due to medical reasons. In contrast, parasite prevalence hardly varied with the season (minimum 3.6%, maximum 7.5%). In a one-year period, the total number of new malarial infections was estimated between 173 and 230. Because of the existence of a vector density gradient in the area concerned, the annual malaria incidence varied considerably according to the children's place of residence.(ABSTRACT TRUNCATED AT 250 WORDS)

Malaria attacks in children exposed to high transmission: who is protected?

In persons naturally exposed to high transmission, how significant, in terms of immune protection, is the occurrence or non-occurrence of a malaria attack during a given observation period? This question was studied in a West African village where Plasmodium falciparum malaria is holoendemic with intense and perennial transmission. A cohort of 94 children aged 4 months-14 years from Dielmo village, Sénégal, was studied over 4 months, June-September 1990. 41 children had no malaria attack and 53 children suffered between one and 6 attacks. The average number of attacks was 1.80, 2.25, 1.87, 0.29 and 0.07, respectively, in children aged 4-11 months, 1-2, 3-6, 7-10 and 11-14 years. The transmission level was 75 infective bites per person. Analysis of the distribution of the number of attacks in individual children suggested that all children within the same age group had either the same, or a very similar, level of protection. This suggests that the acquisition of clinical protection in areas where malaria is highly endemic involves a progressive and homogeneous decrease of the probability of having a malaria attack (attacks occur less frequently as age increases, in all children), rather than the acquisition of complete protection by an increasing number of children. The differences between the number of clinical attacks observed in young children and their presumed exposure suggest that protective mechanisms become effective from the first reinfection onwards, and are independent of the cumulative exposure to a great variety of antigens.

Lymphocyte response in vitro to Plasmodium falciparum merozoite antigens in donors from a holoendemic area

Crude merozoite antigens from Plasmodium falciparum were used to evaluate the proliferative response of peripheral mononuclear cells (PBMCs) from 114 inhabitants of the village of Dielmo (Senegal, West Africa) exposed continuously to malaria transmission. The high or low responses to merozoite antigens obtained in lymphocyte stimulation assays were correlated with the presence or absence of parasites, IFN-gamma production and HLA phenotype. The high responders produced high levels of IFN-gamma, in contrast to the low responders, most of whom did not secrete IFN-gamma (23/27). Among others, the two HLA phenotypes HLA-B51 and HLA-DR1 were significantly associated with a high response (P < 0.05).

Antibodies and reactive T cells against the malaria heat-shock protein Pf72/Hsp70-1 and derived peptides in individuals continuously exposed to Plasmodium falciparum

Pf72/Hsp70-1, a heat-shock protein of m.w. 72 kDa from Plasmodium falciparum is one of the Ag of interest to be included in a polyvalent vaccine against malaria. It is one of the major immunogens present in a fraction of purified blood stage parasites that elicited protection against experimental infection of Saimiri monkeys with blood stages of P. falciparum. It is present at all blood stages and one of its B cell epitopes is also detected on the surface of the infected hepatocyte. Moreover, Pf72 appears to be well conserved among different isolates of P. falciparum. We have examined the immune response against Pf72/Hsp70-1 in individuals from different age groups living in a holoendemic area (West Africa). The immune response against the native Ag (purified from schizonts and called Pf/Hsp70) was analyzed both at the humoral level by ELISA and at the cellular level by assessing in vitro proliferation and IFN-gamma production of PBMC. Of the individuals studied 52% had a statistically significant level of anti-Pf/Hsp70 antibodies as compared with unexposed individuals. These positive individuals showed a heterogeneous distribution because significant levels of antibodies were found in 70% of the adults but in only 26% of the children. The presence of Pf/Hsp70-specific reactive T cells in the blood was detected in 32% of the individuals. The total anti-Pf/Hsp70 antibody level (IgG+IgM) appeared strongly age related and correlated positively with parasite exposure, whereas the T cell response failed to correlate either with the antibody level or with age. Moreover, PBMC of donors responded to the Pf/Hsp70 in a dissociated way, namely, by either T cell proliferation or IFN-gamma production. Ten synthetic peptides based on sequences found in the C-terminal part of Pf72/Hsp70-1 were further tested as potential T cell epitopes. The proliferative response of PBMC from individuals continuously exposed to the parasite showed that three peptides more frequently trigger significant T cell proliferation (in 21% to 27% of the individuals) and three others less frequently (10%). None of these peptides allowed detection of reactive T cells in PBMC of Europeans with no previous exposure to malaria. Some of the stimulating peptides are highly similar to human heat-shock Hsc and Hsp70 with large stretches of identical amino acids.(ABSTRACT TRUNCATED AT 400 WORDS)

Antibodies and reactive T cells against the malaria heat-shock protein Pf72/Hsp70-1 and derived peptides in individuals continuously exposed to Plasmodium falciparum

Pf72/Hsp70-1, a heat-shock protein of m.w. 72 kDa from Plasmodium falciparum is one of the Ag of interest to be included in a polyvalent vaccine against malaria. It is one of the major immunogens present in a fraction of purified blood stage parasites that elicited protection against experimental infection of Saimiri monkeys with blood stages of P. falciparum. It is present at all blood stages and one of its B cell epitopes is also detected on the surface of the infected hepatocyte. Moreover, Pf72 appears to be well conserved among different isolates of P. falciparum. We have examined the immune response against Pf72/Hsp70-1 in individuals from different age groups living in a holoendemic area (West Africa). The immune response against the native Ag (purified from schizonts and called Pf/Hsp70) was analyzed both at the humoral level by ELISA and at the cellular level by assessing in vitro proliferation and IFN-gamma production of PBMC. Of the individuals studied 52% had a statistically significant level of anti-Pf/Hsp70 antibodies as compared with unexposed individuals. These positive individuals showed a heterogeneous distribution because significant levels of antibodies were found in 70% of the adults but in only 26% of the children. The presence of Pf/Hsp70-specific reactive T cells in the blood was detected in 32% of the individuals. The total anti-Pf/Hsp70 antibody level (IgG+IgM) appeared strongly age related and correlated positively with parasite exposure, whereas the T cell response failed to correlate either with the antibody level or with age. Moreover, PBMC of donors responded to the Pf/Hsp70 in a dissociated way, namely, by either T cell proliferation or IFN-gamma production. Ten synthetic peptides based on sequences found in the C-terminal part of Pf72/Hsp70-1 were further tested as potential T cell epitopes. The proliferative response of PBMC from individuals continuously exposed to the parasite showed that three peptides more frequently trigger significant T cell proliferation (in 21% to 27% of the individuals) and three others less frequently (10%). None of these peptides allowed detection of reactive T cells in PBMC of Europeans with no previous exposure to malaria. Some of the stimulating peptides are highly similar to human heat-shock Hsc and Hsp70 with large stretches of identical amino acids.(ABSTRACT TRUNCATED AT 400 WORDS)

[Platelet antibody activity in malaria thrombocytopenia]

To assess immune responses to malaria-induced thrombocytopenia, an haematologic and immunologic study was performed on 25 patients with imported malaria upon admission and 8 days after treatment. Thrombocytopenia (150 x 10(9)/litre) was detected in 19 cases (P. falciparum: 11 cases, P. ovale: 6 cases, P. vivax: 2 cases). No laboratory evidence of disseminated intravascular coagulation impairment was found in any of the patients. Bone marrow examination performed in 9 cases showed no abnormality in the megakaryocyte series. Platelet count was independent of circulating parasite levels (r = 0.27) and inversely related to the number of antibody binding sites (ABS) on platelets (r = -0.6, p. less than 0.01). The indirect Coombs test (r = -0.54; p less than 0.01) and IgG and IgE levels (p less than 0.02) gave similar findings. A statistical correlation was observed between the level of circulating immune complex and the number of ABS (r = 0.525, p less than 0.01). Thus malaria-induced thrombocytopenia seems to mainly involve IgG type antiplatelet antibody activity. Although they may be implicated in the binding of antibodies to platelets, circulating immune complexes do not appear to mediate thrombocytopenia.