Naturally acquired antibodies to sporozoites do not prevent malaria: vaccine development implications.

The first human vaccines against the malaria parasite have been designed to elicit antibodies to the circumsporozoite protein of Plasmodium falciparum. However, it is not known whether any level of naturally acquired antibodies to the circumsporozoite protein can predict resistance to Plasmodium falciparum malaria. In this study, 83 adults in a malaria-endemic region of Kenya were tested for circumsporozoite antibodies and then treated for malaria. They were monitored for the development of new malaria infections for 98 days. Antibody levels, as determined by four assays in vitro, were indistinguishable between the 60 individuals who did and the 23 who did not develop parasitemia during follow-up, and there was no apparent relation between day of onset of parasitemia and level of antibodies to circumsporozoite protein. Unless immunization with sporozoite vaccines induces antibodies that are quantitatively or qualitatively superior to the circumsporozoite antibodies in these adults, it is unlikely that such antibodies will prevent infection in areas with as intense malaria transmission as western Kenya.

Isolation of HIV-1 from the blood of seropositive adults: patient stage of illness and sample inoculum size are major determinants of a positive culture. The Walter Reed Retroviral Research Group.

Results of HIV-1 blood cultures from 609 seropositive adults across all stages of illness at the Walter Reed Army Medical Center were reviewed. HIV-1 was isolated by coculturing of patient peripheral blood mononuclear leukocytes (PBMCs) with normal blood donor target PBMCs that had been stimulated with phytohemagglutinin and interleukin-2. The HIV-1 isolation success rate at Walter Reed increased progressively each year from 1986 to 1989. In 1989, HIV-1 was isolated from a single blood specimen from patients in Walter Reed stages 1-2, 3-4, and 5-6 in 75% (49/65), 90% (37/41), and 97% (30/31) of cases, respectively. None of 22 blinded negative control specimens was positive. PBMC cultures from late stage patients regularly became positive within 7 days (92%), compared to only 46% of positive cultures from early stage patients. For most patients, the lowest number of serially diluted PBMCs that resulted in a positive culture was 30,000 patient PBMCs, but the range was 300 to 3 million cells. HIV-1 was isolated less frequently from plasma (5/18, 28%). Plasma viremia was detected only in patients with relatively high titers of infected PBMCs. Forty-six blood specimens from "at-risk" seronegative adults were also cocultured; none was positive.

Malaria in asplenic mice: effects of splenectomy, congenital asplenia, and splenic reconstitution on the course of infection.

In investigations on the role of the spleen in host defense against malaria, we studied the course of murine malaria in three groups of mice with altered splenic function: congenitally asplenic mice, adult-splenectomized mice, and adult-splenectomized mice which were reconstituted with spleen-cell suspensions. Intact mice infected with either Plasmodium yoelii or P. chabaudi adami experienced infections which resolved spontaneously, with low mortality. Congenitally asplenic and splenectomized-reconstituted mice were unable to clear their primary infections, and experienced high mortality; infections in the latter two groups of mice differed little from those in splenectomized, nonreconstituted controls. However, when asplenic and splenectomized mice were treated with cloroquine during their primary infections and then rechallenged with the homologous Plasmodium species, they experienced mild infections similar to those of intact controls. These observations support the concept that the host defense in primary malaria infections requires an architecturally intact spleen, and therefore is not solely dependent upon the presence of a subpopulation of immune spleen cells.

Susceptibility of clinically sensitive and resistant Leishmania to pentavalent antimony in vitro.

Standard courses of pentavalent antimonials frequently fail to cure cutaneous, mucocutaneous, and visceral leishmaniasis, and characteristically fail to cure diffuse cutaneous disease. We have determined the in vitro sensitivity of clinical isolates of Leishmania to pentavalent antimony to determine if inherent drug resistance of the parasite is responsible for treatment failures in human beings. Intracellular amastigotes resulting from promastigote-initiated infection of human macrophages were exposed to pentavalent antimony for 6 days at 34.5-35 degrees C. Amastigotes from clinically sensitive simple cutaneous lesions exhibited a range of in vitro sensitivity. Four strains were greater than or equal to 90% eliminated and two strains were 70-75% eliminated in vitro by concentrations of antimony (15-20 micrograms Sb/ml), comparable to peak achievable serum levels in humans. Amastigotes from initially clinically resistant simple cutaneous lesions showed a wider range of sensitivities. Five strains were greater than or equal to 90% eliminated, but one strain was only 40% eliminated and another strain was completely insensitive in vitro. The clinically resistant diffuse cutaneous strain was 61% eliminated. The techniques described herein permit determination of the in vitro antimicrobial susceptibility of Leishmania from all major human forms of leishmaniasis. The data from this series indicate that in a minority of initially resistant cases parasite resistance to the drug may be contributing to clinical resistance, and use of non-antimonial drugs might be recommended for future therapy.

Concurrent infection with Leishmania donovani and Leishmania major in a Kenyan patient: clinical description and parasite characterization.

Leishmania isolates aspirated a few months apart from the spleen of an indigenous adult male kala-azar patient from Baringo District, Kenya, were biochemically characterized and compared. The patient lived within a dual focus of L. donovani kalazar and L. major cutaneous leishmaniasis. A primary Leishmania isolate from splenic aspirates was cryopreserved (NLB-294). The patient was treated with sodium stibogluconate for kala-azar and discharged. Three months later, he had clinical relapse and returned for retreatment. During his second visit, the patient participated in a diagnostic study in which urine and nasopharyngeal samples were cultured for leishmaniasis. Urine, nasopharyngeal, and splenic samples were positive for Leishmania. Secondary isolates from splenic (NLB-294-I) and urine (NLB-318) cultures were cryopreserved and characterized by cellulose acetate electrophoresis (CAE) using 20 enzymes. Whereas the urine isolate was typed as L. donovani, the splenic aspirate culture revealed a mixed infection with L. donovani and L. major. The primary isolate (NLB-294) was then characterized and also showed a mixed infection. To exclude the possibility of protein post-translational modifications in electrophoretic assays, the primary and secondary isolates were grown and processed under identical cultural and lysis conditions, and compared using CAE. The results were identical to the first electrophoretic assays showing mixed promastigote banding patterns. Stationary-phase promastigotes of the secondary splenic isolate (NLB-294-I) inoculated subcutaneously, intraperitoneally, and intracardially into Syrian hamsters and BALB/c mice produced both kala-azar and cutaneous leishmaniasis within 6.5 months.(ABSTRACT TRUNCATED AT 250 WORDS)

Predicting outcome in malaria: correlation between rate of exposure to infected mosquitoes and level of Plasmodium falciparum parasitemia.

The level of Plasmodium falciparum parasitemia at clinical presentation has repeatedly been shown to correlate with severity of disease. Using data collected in western Kenya over 21 months, we examined associations between exposure variables, especially exposure to infective mosquitoes, and prevalence and density of P. falciparum parasitemia among 1,007 children six months to six years of age. The prevalence of P falciparum infection was similar at all exposure levels, but there was a correlation between exposure to sporozoite-infected mosquitoes over the previous 28-day period, and geometric mean parasite density of each cohort (Spearman rank coefficient = 0.724, P = 0.002). The relative odds of having a parasite density > or = 5,000/microliters was increased almost two-fold among individuals exposed to more than 10 infective bites during the prior 28-day period. Children enrolled during the highest incidence period were 80% more likely to have a density > or = 5,000/microliters relative to individuals enrolled during periods of lower incidence. The data suggest that measures, such as malaria vaccines, that reduce parasite densities by limiting numbers of sporozoites reaching the liver, or merozoites released from the liver, will reduce malaria-associated morbidity and mortality, even when they do not prevent all infections.

Evaluation of promastigote and amastigote antigens in the indirect fluorescent antibody test for American cutaneous leishmaniasis.

Leishmania braziliensis panamensis promastigotes, temperature-induced in vitro-cultivated amastigotes, Vero cell-derived amastigotes, and rodent lesion-derived amastigotes were evaluated as antigens in the indirect immunofluorescent antibody (IFA) test for American cutaneous leishmaniasis. Test sensitivity was determined using sera from 34 U.S. soldiers with leishmaniasis diagnosed by demonstrating parasites in their skin lesions. Sera were collected from 3 to 24 months after exposure to Leishmania. Positive IFA reactions among patient sera were 82% with promastigotes or lesion amastigotes, 79% with in vitro amastigotes, and 76% with Vero cell amastigotes (P = N.S.). Positive titers ranged from 1:8 to 1:128 using all antigens. Test specificity was determined with 30 sera from healthy individuals. False positive reactions ranged from 0-5% depending on the antigen and all titers were less than or equal to 1:8. Test cross-reactivity was assessed with 47 sera from patients with other diseases. Depending on the antigen, cross-reactions occurred with sera from patients with Chagas' disease, lupus erythematosus, malaria, toxoplasmosis and amebiasis. None of the antigens cross-reacted with sera from patients with viral hepatitis, coccidioidomycosis, syphilis, schistosomiasis, and trichinosis. In replicate experiments, 99-100% of the sera varied no more than +/- 1 titer dilution. As sensitivity, specificity, cross-reactivity, and reproducibility of the four antigens were statistically similar, promastigotes, which can be easily and economically cultured in large numbers in vitro are recommended for use in the IFA test for American cutaneous leishmaniasis.

Leishmania donovani parasites in the nasal secretions, tonsillopharyngeal mucosa, and urine centrifugates of visceral leishmaniasis patients in Kenya.

In the early 1930s, investigators of visceral leishmaniasis stated that Leishman-Donovan bodies are found in body fluids of kala-azar patients, for example, in urine, feces, semen, and nasal and pharyngeal secretions. Based on this finding, we investigated the diagnostic potential of nasal secretions, tonsillopharyngeal mucosal swabs, and urine centrifugates inoculated into Schneider's Drosophila Medium (containing antibiotics and antifungal agents) as well as with Giemsa-stained smears. Consequently, 64 randomly selected patients with visceral leishmaniasis from Kenya (59 who were splenic culture or Giemsa stain positive and five who were culture negative but Giemsa stain positive) were tested by three noninvasive methods. These tests were all performed before the patients were treated with Pentostam. Cultures of nasal and tonsillopharyngeal swabs and urine centrifugates produced 28 positive samples representing 24 patients (37.5%). Moreover, a set of 25 Giemsa-stained slide smears made from the nasal and tonsillopharyngeal mucosa of 25 patients with visceral leishmaniasis who had not tested positive in cultures produced nine positives. Therefore, the overall total of patients who tested positive by all of the above methods was 33 or 51.6%. The cryopreserved Leishmania isolates were characterized by cellulose acetate electrophoresis using 20 enzyme systems. The isoenzyme profiles produced by the parasites were represented in five different L. donovani s.l. zymodemes. Representatives of these isolates were also characterized by DNA Southern blotting analysis, which corroborated the isoenzyme results.

Pre-exposure rabies immunization with human diploid cell vaccine: decreased antibody responses in persons immunized in developing countries.

In November 1982, a U.S. Peace Corps volunteer in Kenya completed pre-exposure rabies prophylaxis with a standard 3 dose intradermal (ID) series of human diploid cell rabies vaccine (HDCV). In May 1983, she was bitten by a dog and died of rabies 3 months later. An initial investigation revealed that the patient, as well as 9 of 11 others immunized at the same time, had no rabies antibody titers (less than 1:5). We therefore instituted investigations into the immunogenicity of pre-exposure HDCV both in the United States and in developing countries. A serosurvey revealed unexpectedly low rabies titers in both Peace Corps volunteers and others immunized in developing countries. Antibody titers measured 2-3 weeks after ID immunization were compared in 9 groups totaling 271 persons in the United States and Kenya. There was no statistically significant difference in antibody titers in the 6 U.S. groups immunized from 1980-1984 (P greater than 0.15); however, groups immunized in the United States had significantly higher titers than a group of Kenyan nationals (P less than or equal to 0.0001), and the Kenyans had significantly higher titers than 2 Peace Corps groups immunized in Kenya (P less than or equal to 0.0001). No single hypothesis proposed (laboratory error, vaccine potency, vaccination technique, or specific immune suppression) accounted for the observed differences. Although we cannot fully explain the poor response to HDCV, it is probably due to multiple factors. We conclude that persons immunized with ID pre-exposure HDCV in developing countries should have rabies antibody titers determined to ensure their seroconversion; for persons immunized in the United States, such titers need not be routinely determined.

Plasmodium falciparum incidence relative to entomologic inoculation rates at a site proposed for testing malaria vaccines in western Kenya.

Relationships between Plasmodium falciparum incidence and entomologic inoculation rates (EIRs) were determined for a 21-month period in Saradidi, western Kenya, in preparation for malaria vaccine field trials. Children, ranging in age from six months to six years and treated to clear malaria parasites, were monitored daily for up to 12 weeks to detect new malaria infections. Overall, new P. falciparum infections were detected in 77% of 809 children. The percentage of children that developed infections per two-week period averaged 34.7%, ranging from 7.3% to 90.9%. Transmission by vector populations was detected in 86.4% (38 of 44) of the two-week periods, with daily EIRs averaging 0.75 infective bites per person. Periods of intense transmission during April to August, and from November to January, coincided with seasonal rains. Relationships between daily malaria attack rates and EIRs indicated that an average of only 7.5% (1 in 13) of the sporozoite inoculations produced new infections in children. Regression analysis demonstrated that EIRs accounted for 74% of the variation in attack rates. One of the components of the EIR, the human-biting rate, alone accounted for 68% of the variation in attack rates. Thus, measurements of either the EIR or the human-biting rate can be used to predict corresponding attack rates in children. These baseline epidemiologic studies indicate that the intense transmission patterns of P. falciparum in Saradidi will provide excellent conditions for evaluating malaria vaccine efficacy.

Plasmodium falciparum gametocytemia in Kenyan children: associations among age, intensity of exposure to transmission, and prevalence and density of subsequent gametocytemia.

Recently, an association was described between the density of Plasmodium falciparum asexual parasitemia in Kenyan children and the entomologic inoculation rate (EIR) measured prior to measurement of asexual parasitemia. This study examined whether transmission pressure, as represented by the EIR, was associated with the prevalence or density of gametocytemia in Kenyan children. Each month for 19 months, a cohort of approximately 50 children was given a radical cure and enrolled in the study. Blood films were taken on days 0, 7, and 14. The EIR was calculated for the 28-day period ending 14 days prior to enrollment: the relationship between blood film data from day 7 and exposure variables was explored. We found that younger children were more likely to be gametocytemic than older children and, if gametocytemic, were more likely to have a dense gametocytemia. There was an inverse relationship between the number of infective bites per night received and prevalence but not density of gametocytemia, even after age adjustment. Concordance of gametocytemia prevalence on days 0 (64%), 7 (66%), and 14 (52%) was poor; 84% of the children were positive on at least one day. This indicates that in many subjects the detectable gametocytemia varied over the 14 days. Under these holoendemic transmission conditions, the EIR is inversely correlated with prevalence of gametocytemia, and point measurements of gametocytemia by conventional microscopy underestimate the number of infective donor hosts.

American cutaneous leishmaniasis: a comparison of three sodium stibogluconate treatment schedules.

Thirty-six patients with American cutaneous leishmaniasis were randomized to receive intravenous sodium stibogluconate for 10 days at a dose of 600 mg antimony (Sb) per day by one of three schedules: once daily by rapid infusion (A), by continuous 24 hr infusion (B), or in divided doses every eight hours by rapid infusion (C). Patients not cured after initial treatment were rerandomized to one of the other treatment schedules. An additional 19 patients who were not part of the randomized study received standard (STD) sodium stibogluconate treatment (600 mg Sb once daily by rapid infusion for 10 days, identical with schedule A). In the randomized study, the overall cure rate after the first course of treatment was 64%, but was higher for schedule A (100%) than for B (50%) or C (42%) (P less than 0.01). Considering all courses of treatment, schedule A was more effective (94%) than B (53%) or C (43%) (P less than 0.01). Paradoxically, patients in group A had a higher cure rate than patients in group STD (42% after the first course of treatment and 51% when all courses of treatment were considered). Side effects were mild and well tolerated. Total side effects were more frequent in groups B + C (52%) than A + STD (23%) due to an increased incidence of subjective complaints (26% vs. 10%, P less than 0.05) in patients receiving other than once daily rapid infusion. We conclude that giving the same total amount of sodium stibogluconate in three divided doses or by continuous infusion offers no advantage over standard, once daily treatment of American cutaneous leishmaniasis.

Iontophoresis of vinblastine into normal skin and for treatment of Kaposi's sarcoma in human immunodeficiency virus-positive patients. The Military Medical Consortium for Applied Retroviral Research.

BACKGROUND: Patients who test positive for human immunodeficiency virus type 1 (HIV-1) and who have disfiguring and/or painful cutaneous lesions of Kaposi's sarcoma (KS) may not be candidates for systemic chemotherapy and/or immunotherapy. Intralesional vinblastine sulfate, as a single-agent chemotherapeutic drug, has been used with some success to treat KS in patients who are HIV-1 positive. However, some patients may not tolerate the pain associated with injection of vinblastine. Transcutaneous iontophoresis of vinblastine was evaluated for therapy of KS in HIV-1-infected patients. Prior to therapy of patients, we iontophoresed vinblastine into the normal skin of volunteers who were not infected with HIV-1 to document the clinical and histologic features that occurred. OBSERVATIONS: Iontophoresis produced a localized erythematous papular eruption in non-HIV-infected volunteers but not in HIV-1-infected patients. Histologic changes in the biopsy specimens taken from non-HIV-infected volunteers consisted primarily of scattered necrotic keratinocytes and a mild to moderate superficial lymphohistiocytic infiltrate. Thirty-one lesions of KS were treated with partial to complete clearing and symptomatic improvement. CONCLUSION: Clinical and histologic features of iontophoresed normal skin suggest an immunologic mechanism of action. Iontophoresis of vinblastine for KS is well tolerated and results in symptomatic improvement as well as varying degrees of clearing of the lesions.

Mouse foot-pad inoculation as an aid to the isolation of Leishmania spp. from patients.

Portions of splenic or subcutaneous saline aspirates from suspected visceral or cutaneous leishmaniasis patients were inoculated into NNN media with an overlay of Schneider's medium or Schneider's medium alone for routine parasitological diagnosis. The remaining portions of the aspirates were used for preparing Giemsa-stained smears and for subcutaneous inoculation into hind foot-pads of Balb/c mice. Saline aspirates obtained from the foot-pads 2-14 d after inoculation were inoculated into Schneider's medium and examined for promastigotes. Parasite isolation was achieved from 90% of confirmed leishmaniasis patients by either culture method alone. Mouse foot-pad aspiration demonstrated parasites in 95% of all patients, and in over 80% of the confirmed cases of leishmaniasis. Combined culturing and aspirate smear examination was more efficient than foot-pad inoculation alone for the demonstration of leishmanial infection. Foot-pad aspiration does not entail killing animals and was sensitive for parasite isolation; it may be a useful short-term adjunct to existing parasite isolation methods, especially under field conditions where the risks of culture contamination may be high.

Isolation of Leishmania donovani from Phlebotomus martini in Baringo district, Kenya.

An 18-month sandfly survey was conducted at 4 locations in Baringo District, Rift Valley Province, Kenya. 3 collection techniques were used: aspiration, sticky paper trap, and light trap in sites selected because of their proximity to homes of visceral leishmaniasis patients diagnosed and treated within 6 months before the survey. Over 2000 female Phlebotomus martini were collected of which 6 females were found to have flagellate protozoan infections. 3 of these infections were cultured successfully and cryopreserved. 2 isolates were identified as Leishmania donovani by cellulose acetate electrophoresis. The zymogram of the third isolate was different from all Old World Leishmania reference strains examined, and it is still unidentified. The finding of 2 P. martini naturally infected with L. donovani strongly supports the hypothesis that this species is a vector of visceral leishmaniasis in this area.

Cutaneous leishmaniasis caused by Leishmania tropica in Kenya.

9 leishmanial strains, isolated from cutaneous papulonodular lesions on 3 patients, were characterized by cellulose acetate electrophoresis using 7 enzymes. The patterns obtained were indistinguishable from those of a Leishmania tropica reference strain and these 9 strains were similar to L. tropica in failing to infect mice. Although these 3 patients were Americans, their only potential exposure to sandflies was in Kenya, and thus they are believed to be the first cases of cutaneous leishmaniasis due to L. tropica in Kenya.

Visceral leishmaniasis unresponsive to antimonial drugs. III. Successful treatment using a combination of sodium stibogluconate plus allopurinol.

Five patients with long-standing visceral leishmaniasis who were unresponsive to sodium stibogluconate, 20 mg antimony/kg body-weight once or twice daily, were treated for 14 to 54 days with a combination of sodium stibogluconate at the same dose plus allopurinol at a dose of 20 mg/kg body-weight per day in three divided doses. This combination was safe and effective. Negative splenic aspirate smears were obtained from all patients within 19 days, and none has relapsed in at least 12 months of follow-up.

Safety and immunogenicity of a Plasmodium falciparum sporozoite vaccine: boosting of antibody response in a population with prior natural exposure to malaria.

Recombinant sporozoite vaccine or placebo were administered once to 25 volunteers from an area endemic for malaria. Antibody to R32tet32 rose in 9 of 15 receiving vaccine and remained elevated in 6 for 6 months. Mean absorbance increase was 0.43 +/- 0.40 with vaccine, 0.01 +/- 0.23 with placebo, and 0.72 +/- 0.19 in responders. Six non-responders had significantly lower pre-immunization levels (0.07 +/- 0.05) than responders (0.39 +/- 0.25). There was an association between an increase in immunofluorescence (n = 4) and an increase in absorbance (n = 9) among vaccine recipients (n = 15). Vaccine-induced increase in antibody to natural circumsporozoite antigen was indicated by increases in immunofluorescence and by increases in circumsporozoite precipitation score in 2 of the 5 responders with highest antibody increase measured by enzyme-linked immunosorbent assay. Response to subunit sporozoite vaccine paralleled response to prior natural sporozoite exposure and was significant and prolonged in a population with prior natural exposure to malaria.

Effect of human circumsporozoite antibodies in Plasmodium-infected Anopheles (Diptera: Culicidae).

Human circumsporozoite (CS) antibodies to Plasmodium falciparum were detected in blood meals from 45.0% of 1,547 field-collected Anopheles gambiae Giles sensu lato and Anopheles funestus Giles from western Kenya. Possible effects on malaria infections within the Anopheles host were investigated. Circumsporozoite antibodies were detected in blood meals up to 36 h after feeding. Antibodies crossing the midgut were detected experimentally in hemolymph from 4 to 36 h after feeding; human IgG also was present in hemolymph from fully gravid field-collected Anopheles. Ingestion of high-titer human CS antibodies or 2A10 monoclonal antibody to P. falciparum sporozoites by P. falciparum-infected An. gambiae, 10 d after feeding on an infected human, had no effect on oöcyst maturation, sporozoite rates, or sporozoite loads. Contact between CS antibodies and sporozoites in the hemocoel did not block sporozoite invasion of salivary glands. Human IgG antibodies were detected by an indirect fluorescent antibody technique on salivary gland sporozoites from 83.3% of 114 field-collected Anopheles. In 65.4% of 26 infections, antibodies persisted on sporozoites for at least three days. Thus, a high proportion of naturally infected An. gambiae s.l. and An. funestus in western Kenya transmit sporozoites that are bound with human IgG acquired during previous blood meals. The infectivity of such sporozoites needs to be determined in relation to natural transmission and to the potential use of malaria sporozoite vaccines.

Characterization of malaria transmission by Anopheles (Diptera: Culicidae) in western Kenya in preparation for malaria vaccine trials.

Malaria transmission was studied for 33 mo in the villages of Kisian and Saradidi in western Kenya in preparation for field trials of malaria vaccines. Abundance estimates of Anopheles gambiae Giles sensu lato and Anopheles funestus Giles, which constituted over 99% of 26,645 anophelines collected, were compared for all-night biting collections inside houses, outdoors, and in tents. The overall numbers of Anopheles per man-night were 2.3 times greater in Kisian than in Saradidi. For the three types of collections, mean sporozoite rates by dissection ranged from 2.2 to 5.4% for 13,072 Anopheles in Kisian and from 9.9 to 13.6% for 7,058 Anopheles in Saradidi; greater than 90% of the infections were Plasmodium falciparum, either alone or mixed with P. malariae or P. ovale. Heaviest transmission from April to July coincided with the end of the long rainy season. Entomological inoculation rates (EIR) averaged 0.82 infective bites per man per night inside houses in Kisian and 0.65 in Saradidi. Outdoors, EIRs averaged 0.09 in Kisian and 0.52 in Saradidi. In tents, which were evaluated to identify methods for exposing nonindigenous volunteers during vaccine efficacy trials, EIRs were 3.3 and 2.5 times less than inside houses for Kisian (EIR = 0.25) and Saradidi (EIR = 0.26), respectively. Exposure in tents averaged one infective bite every 4.0 d in Kisian and every 3.8 d in Saradidi. The use of tents in vaccine efficacy trials should provide adequate exposure for nonindigenous volunteers. Malaria vaccine trials could be conducted efficiently in western Kenya, with timing dependent upon the intensity of transmission required by vaccine trial objectives.