The American Journal of Tropical Medicine and Hygiene - Volume 53, Issue 4, October 1995

Between 1991 and 1994, an intervention program with permethrin- and lambdacyhalothrin-impregnated bed nets was carried out over a period of nine months in each of five endemic, malarious areas of Ecuador, Peru, and Colombia. This program was evaluated through household surveys, blood sampling, in-depth longitudinal studies, and entomologic analysis. Eighty-four communities (including approximately 35,000 individuals) were paired according to malaria incidence, size, and coverage with bed nets and then randomly allocated to intervention and control groups. The results showed that peoples' acceptance of the measure was related to their perception of an immediate protective effect against insects. The effectiveness of the bed nets, measured as a reduction of malaria incidence in intervention communities as against control communities, showed large variations between and within the study areas. The protective efficacy varied between 0% and 70% when looking only at the postintervention differences between intervention and control groups. The average protection was 40.8% when considering a four-month incidence of clinical malaria attacks and 28.3% when considering a two-week malaria incidence. Important factors for the success of the bed net program were insect susceptibility to pyrethroids, high coverage with impregnated bed nets, high malaria incidence, good community participation, high mosquito densities when people go to bed, and a high proportion of Plasmodium falciparum. In one area, where DDT spraying in the control communities was executed, the effectiveness of bed net impregnation was slightly better than that of spraying.

An evaluation of the Lao Aedes aegypti control program and of the predatory abilities of copepods from Vientiane, Lao People's Democratic Republic was undertaken before a field release of copepods in Thongkankam village, Vientiane. Copepods were transported to Australia for evaluation of predatory abilities and their survival under various nutrient and pH conditions. Mesocyclops guangxiensis was chosen for release over M. aspericornis due to its higher reproduction rate and its ability to survive in lower nutrient environments. Mesocyclops guangxiensis was released into 142 containers and 20 wells in a village in Vientiane. Copepods were present in 7% of the containers after one month and were absent six months postinoculation. In comparison, 100% of wells were still positive after six months, with average numbers of Ae. aegypti in the wells decreasing from 59.5 ± 18.5 (± SEM) to 0 after six months. Numbers of Culex quinquefasciatus and Anopheles maculatus also decreased to 0 after six months. This study indicates that predacious copepods will be accepted by the community and could be integrated as a low-cost, persistent control agent into new strategies for improving surveillance and control of dengue vectors.

In this paper, we consider the movement of Rift Valley fever (RVF) virus from infected mosquito midgut epithelial cells into the hemocoel as an important factor in the ultimate ability of the insect to transmit the virus. Our results are therefore significant in the context of vector competence. The mosquito Culex pipiens was identified as the primary vector of RVF in an epidemic that occurred in Egypt in the 1970s. On this basis, we have carried out several studies of RVF virus in this mosquito species. In the research reported here, we used immunocytochemical and transmission electron microscopic techniques to study the occurrence of RVF virus in the mosquito cardia and aspects of the histology and ultrastructure of this organ. The cardia is a complex organ consisting of both foregut and midgut tissue and is the location of the foregut-midgut junction. The cardia is of interest because it appears to provide routes of RVF virus egress from the midgut lumen and it is consistently infected in mosquitoes with disseminated infections, making it a potentially important site of viral amplification and an ideal site for studying RVF viral morphogenesis. In orally infected mosquitoes, large numbers of RVF virions were observed budding into the basal labyrinth associated with the outer cardial epithelial cells and into the noncellular matrix associated with the inner cardial epithelial cells and the cells of the intussuscepted foregut. In mosquitoes infected by injection of virus into the hemocoel and then held for different incubation periods, viral antigen was first detected in the cells of the intussuscepted foregut in the cardia and later in the cardial epithelial cells. In both orally and hemocoelically infected mosquitoes, large numbers of virions were observed in the matrix. It is therefore clear that RVF virus can infect the cardial epithelial cells both from the lumen of the midgut and from the hemocoel, suggesting that the cardial tissues, including the intussuscepted foregut, provide a ready conduit for virus.

Serologic markers of hepatitis B virus (HBV), including HBV DNA, and of hepatitis delta virus were measured in three villages in Gabon. Of 303 subjects studied, 19% were carriers of hepatitis B surface antigen (HBsAg); 8.5% of these had anti-delta antibodies. No difference among the three villages was observed. All HBV DNA carriers were children less than 11 years of age. In the 2–9-year-old group, 71% of the HBsAg-positive children tested were HBV DNA carriers. These results indicate that HBV prevalence is high in Gabon and emphasize the importance of horizontal transmission of HBV in Africa. Antibodies to hepatitis C virus, assessed in one of the three villages, were found with a prevalence of 35% with a second generation enzyme-linked immunosorbent assay (ELISA) and 24% with a third generation ELISA. None of 35 subjects tested for antibodies to hepatitis E virus was positive.

Twenty-two Costa Rican dairy herds were actively monitored for vesicular diseases between November 1991 and March 1992, in anticipation of the annual dry season occurrence of vesicular stomatitis (VS). Thirty-nine confirmed cases of VS were concentrated on nine farms, resulting in an average herd incidence rate of 11.1% of mature cattle for the affected farms, or 2.6% for all farms. Affected cattle were generally older (average age = 5.4 years), with seven-year-old cows having the highest age-specific incidence rate (6.5%). The New Jersey serotype of VS was diagnosed for all but four cases, primarily manifesting as teat vesicles. Cases had high acute and convalescent antibody titers for both major VS serotypes, New Jersey and Indiana, with significant titer increases after clinical disease. Screening and multiple variable regression analyses of 140 cow, farm, and ecologic variables revealed that clinical disease was significantly associated with cows being in lactation, and with higher acute antibody titers to serotype Indiana, but not with any predisposing diseases. Significant farm factors were the presence of poultry and a longer calving interval on the farm, but not higher rates of other diseases. According to a prior epidemiologic hypothesis, two ecologic factors were forced to be included into the farm model: the reported presence of sand flies, and a higher proportion of the farm being in forest land, with subsequent stepwise regression. The resulting model containing only the two forced variables was significant (P < 0.003) and correctly classified 78% of farms into the correct group, as compared with 79% for the final stepwise model. Study results suggested that the occurrence of clinical VS on a farm was more likely related to ecologically dependent vector transmission to older lactating animals, rather than to predisposing diseases that may immunocompromise affected animals. This was the first prospective case-control study aimed at finding host and ecologic risk factors associated with an outbreak of clinical VS.

The prevalence of Vibrio cholerae O1 and non-O1 has been investigated in numerous Somali regions of the Horn of Africa from 1983 to 1990. From January 1983 to January 1985 and between December 1986 and December 1990, no strains of V. cholerae O1 and 226 strains (5.3%) of V. cholerae non-O1 were isolated from 4,295 diarrhea cases. During a cholera epidemic in 1985 and 1986, the overall case-fatality rate was 13% and the attack rate was 3–3.5 per 1,000 population. Matched case-control studies identified a waterborne route of transmission. A drug-susceptible Ogawa strain from Ethiopia caused the introduction of the disease into northern Somalia. There were two major resistant derivatives of the original strain, and the one resistant to ampicillin, kanamycin, streptomycin, sulfonamide, and tetracycline (TC) predominated in the spreading disease. In 1986, susceptible Ogawa strains quickly displaced this resistant strain. The two incompatibility group C plasmids responsible for the resistance patterns had complex and scattered differences in their structures. Physical analysis of the plasmid DNA region coding for TC resistance demonstrated its genetic amplification in highly resistant variants of Ogawa strains.

A five-year follow-up of the natural history of 924 children exposed to or infected with Leishmania chagasi was conducted in a rural area of northeast Brazil. Seventy-eight percent of the children sought were located. There was no evidence of smoldering disease or long subclinical latency in this population. The overall prevalence of clinical visceral leishmaniasis in this population was 6.1%, with a mortality rate of 10%.

An electrophoretic survey of 42 populations of Anopheles pseudopunctipennis collected throughout its known geographic distribution was performed to clarify the taxonomic status of this important malaria vector species. The results indicated strong differences in the allele frequencies of three enzyme loci (glycerol dehydrogenase, 6-phosphogluconate dehydrogenase, and phosphoglucomutase) of the 33 loci analyzed. No fixed electromorphic differences separate the populations of An. pseudopunctipennis. The populations of An. pseudopunctipennis showed little genetic divergence, with Nei distances ranging from 0 to 0.079. A comparison of An. pseudopunctipennis data with either one of three other Anopheles species showed a high genetic distance of 0.335 with a closely related species, An. franciscanus; 0.997 with An. crucians, and 2.355 with An. (Nyssorhynchus) albimanus. Geographic populations of An. pseudopunctipennis were classified into three clusters; one cluster included populations collected in North America (United States and Mexico) and Guatemala, one cluster included populations from Belize and South America (Colombia, Ecuador, Peru, Chile, and Argentina); and one cluster was represented by populations from the Island of Grenada (type-locality of An. pseudopunctipennis). Based on our isozyme analyses, we defined these clusters as three geographic populations of An. pseudopunctipennis. Of the two mainland populations, one extends from the southern United States south through Mexico and Guatemala, and the other extends north from southern South America through Central America to Belize. These two geographic populations converge in southern Mexico and northern Central America. One part of the convergence zone was identified in the area of eastern Guatemala and southern Belize.

This study was designed to determine the peak period of infective mosquito biting and thus the greatest potential for filarial transmission to occur. It was found that biting density, natural infection, and infectivity rates of Culex quinquefasciatus were significantly higher in the third quadrant of the night (from midnight to 3:00 am) than at other times. This was true in both an urban and a rural environment. Avoidance of mosquito bite by any means during this time period might reduce and limit filarial transmission to a great extent.

Enlarged regional lymph nodes have been reported to accompany the cutaneous lesions of Leishmania (Viannia) braziliensis (= L. braziliensis). A survey in Ceara State, Brazil indicated that 77% of persons (456 of 595) with parasitologically confirmed cutaneous leishmaniasis reported lymphadenopathy in addition to skin lesions. A group of 169 persons with recently diagnosed leishmaniasis and lymph nodes measuring ≥ 2 cm in diameter (mean = 3.6 cm, maximum = 10.5 cm) underwent detailed clinical examination. Lymphadenopathy preceded the skin lesions in more than two-thirds of these, on the average by two weeks. Cultures of lymph node aspirates yielded Leishmania more frequently (86%) than cultures of aspirates of skin (53%) or biopsies of skin (74%). Parasites were isolated from the peripheral blood of one patient. Persons with lymphadenopathy gave a history of fever and had enlarged livers or spleens more often than a comparison group of 50 persons with cutaneous lesions but no lymphadenopathy. Persons with lymphadenopathy had more intense leishmanin skin reactions and lymphocyte proliferation following stimulation with specific antigens, whereas persons without lymphadenopathy had a higher frequency of previous infection. Isolates of parasites from both groups were identified as L. braziliensis. These data demonstrate the early spread of L. braziliensis beyond the skin and suggest differences in host immunity between persons with and without lymphadenopathy. Leishmaniasis braziliensis should be considered in cases of unexplained lymphadenopathy in endemic areas.

Four hundred patients with suspected tetanus were studied to determine the value of the spatula test to diagnose tetanus. A positive test result (reflex spasm of the masseters on touching the posterior pharyngeal wall) was seen in 359 (94%) of 380 patients with tetanus and in no patient without tetanus. Thirty-three of 400 patients (13 with tetanus and 20 with other diagnoses) had a negative test result (a gag reflex with attempted expulsion of the spatula). Thus, the test performed on presentation had a high specificity (100%) and sensitivity (94%) for diagnosing tetanus.

The in vitro activity of atovaquone (566C80) was evaluated and compared with that of chloroquine, quinine, mefloquine, halofantrine, artemether, pyrimethamine, and cycloguanil against African isolates and clones of Plasmodium falciparum using an isotopic, semimicro, drug susceptibility test. Atovaquone was highly active against the chloroquine-susceptible L-3 (geometric mean 50% inhibitory concentration [IC50] = 0.978 nM) and L-16 clones (mean IC50 = 0.680 nM) and against the multidrug-resistant FCM 29 clone (mean IC50 = 1.76 nM). Similar low IC50 values for atovaquone were observed against the chloroquine-susceptible isolates (n = 35; geometric mean IC50 = 0.889 nM) and the chloroquine-resistant parasites (n = 26; geometric mean IC50 = 0.906 nM). The in vitro responses between atovaquone and the other antimalarial drugs were not correlated, indicating the absence of in vitro cross-resistance. The high in vitro activity of atovaquone without any in vitro evidence for cross-resistance with other antimalarial drugs against the naturally occurring malaria parasites is a factor that favors further development of the drug for clinical use.

Two strains (RH and GC, the latter of which is a Taiwan isolate of porcine origin) of Toxoplasma gondii were kept at -20°C, -60°C, and in liquid nitrogen (-196°C) to follow the time course change in viability and virulence of the parasites by direct count and animal inoculation methods. Changes in antibody titers in some of the mice inoculated with the thawed organisms were assayed by the indirect immunofluorescent antibody test. Viability and virulence of T. gondii were best preserved by storage in liquid nitrogen. Tachyzoites kept in liquid nitrogen for eight years still can lead to the death of the injected mice in 2–3 weeks. Virulence of the tachyzoites could be maintained for eight weeks at most at -20°C and -60°C. Dimethylsulfoxide (DMSO) seemed to be a better cryoprotectant for T. gondii than glycerol, but the DMSO-preserved organisms resulted in fewer tachyzoite-containing peritoneal exudates in inoculated mice than the glycerol-preserved organisms. The local isolate (GC strain) tachyzoites tolerated cryopreservation less well than the RH strain parasites. Low antibody titers (at most 1:64) were produced in mice that survived more than 16 days after inoculation with thawed tachyzoites.

We have studied the growth of Borrelia burgdorferi in nymphal ticks (Ixodes scapularis) feeding on mice using confocal fluorescence microscopy to follow the distribution of spirochetes. In starved nymphs, the bacteria were only detected in the midgut and each nymph had a mean of 496 spirochetes. Upon attachment of nymphs to the host, the bacteria grew with a doubling time close to 4 hr and reached a mean of 7,848 spirochetes per nymph 15 hr after attachment. During this initial period (36 hr) of rapid growth, the bacteria appeared to be restricted to the gut, but after 48 hr, the spirochetes had disseminated to the salivary glands in the majority of nymphs examined. Thus, a critical event that allows the spirochetes to disseminate and infect the salivary glands takes place 36–48 hr after attachment. A maximum number of 166,575 spirochetes per nymph was noted 72 hr after attachment. Soon after completion of feeding and detachment from the host (96 hr), the mean number of spirochetes decreased to 95,410 per nymph and the spirochetes appeared to be cleared from organs other than the midgut. Thus, dissemination of spirochetes within the vector appears to be a transient phenomenon. These results provide strong evidence in favor of a salivary route of disease transmission while also demonstrating the utility of confocal microscopy to study vectorpathogen interactions in general.

The 74HB59 strain of Rift Valley fever (RVF) virus, isolated from a human case in the Central African Republic, was shown to be composed of a heterogeneous population of viruses when plaque-purified clones were analyzed for their reactivity with monoclonal antibodies (MAbs) directed against the nucleocapsid (N) protein or the nonstructural (NSs) protein. One of these clones, C13, was of particular interest in that it proved to be avirulent in mice and hamsters, and highly immunogenic. Although C13 showed normal reactivity with a large panel of MAbs directed at the glycoproteins, it failed to react with specific MAbs or polyclonal antibodies directed at the NSs protein and with a specific MAb recognizing the N protein of the Egyptian strains. Consequently, the small RNA segment, which encodes the N and NSs proteins in an ambisense strategy, was sequenced and compared with the existing sequence of the attenuated MP-12 RVF virus strain. We found that the NSs gene contained, in addition to two conservative coding changes, a large internal deletion of 549 nucleotides that removes 69% of the open reading frame but conserves in-frame the N and C termini of the predicted translation product. In addition, the sequence revealed that the N protein of C13 contained a single amino acid change. Clone C13 replicated normally in certain cell types in vitro and in Culex pipiens mosquitoes after intrathoracic inoculation, but established abortive infections in MRC-5 human fibroblasts.

We used in situ hybridization combined with immunocytochemistry, cell sorting, and the polymerase chain reaction (PCR) to investigate clinical events in three asymptomatic carriers of human T lymphotrophic virus type-1 (HTLV-1) and ten patients with HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP). The objective was to determine which T cell subset of peripheral blood mononuclear cells (PBMC), CD4 or CD8, were infected by HTLV-1 and the manner in which HTLV-1 proviral DNA was expressed at the level of the single cell. Both CD4-positive and CD8-positive cells of the PBMC from five patients with HAM/TSP were infected with HTLV-1. The proportion of HTLV-1-infected cells was 2.5–40% in the CD4-positive subset and 1.0–65% in the CD8-positive subset, when quantified by PCR using HTLV-1-infected MT2 cells as a positive standard. Proviral DNA of HTLV-1 was expressed in both CD4-positive cells and CD8-positive cells of the PBMC from six patients with HAM/TSP and three asymptomatic HTLV-1 carriers. In patients with HAM/TSP, the proportion of the cells expressing HTLV-1 proviral DNA was 0.02–0.1% in both subsets. In asymptomatic carriers, the expression of HTLV-1 proviral DNA was 0.01–0.02% in the CD4-positive subset and 0.01% in the CD8-positive subset. Therefore, HTLV-1 possessed similar in vivo cellular tropism for both CD4-positive cells and CD8-positive cells and HTLV-1 proviral DNA was expressed in vivo in both circulating T cell subsets. These results suggest that the immunologic states of patients with HAM/TSP are modulated by the viral gene expression in both T cell subsets.

Healthy Egyptian neonates born to hepatitis B surface antigen (HBsAg)-seronegative mothers were randomly enrolled in one of three vaccination schedules. A dose of 2.5 µg of recombinant HB vaccine was given at birth, two, and six months of age (group A) or two, four, and nine months of age (group B). These two groups and a third control group (group C) also were given the other routine childhood vaccines (BCG, DPT, polio, and measles). Blood samples were taken one month after the third vaccine dose in groups A (seven months of age) and B (10 months of age), and a second follow-up blood sample was taken at the age of 18 months for all three groups. Sera were tested for HBsAg and antibody to hepatitis B core antigen, and quantitatively for antibody to hepatitis B surface antigen (anti-HBs) using commercial enzyme immunoassay kits. The vaccine was well tolerated and side effects were limited to local soreness, redness, or temporary swelling. Among 590 infants who were followed-up, good (51–300 mIU anti-HBs/ml) or excellent (> 300 mIU/ml) immune responses occurred in 85% of the infants in group A and in 96% in group B. Geometric mean titers of anti-HBs at the first and second follow-up were 306 and 55 mIU/ml in group A, and 1,492 and 147 mIU/ml in group B. The recombinant HB vaccine is safe and immunogenic when given in three doses of 2.5 µg in either regimen, but delay of the booster dose of the vaccine until nine months after birth produced a higher immune response.

This study was part of a larger program to develop a vaccine effective against Plasmodium falciparum infection caused by sporozoites and clinical malaria caused by asexual blood stages. In a phase 1 study of safety and immunogenicity, two recombinant proteins (Ro 46–2717, a circumsporozoite [CS] protein) construct with a molecular mass of 35 kD, and Ro 46-2924, a merozoite surface antigen [MSA-2] construct with a molecular mass of 25 kD) adsorbed onto alum were injected in two low (20 µg) or two high (100 µg) doses in the right and left deltoid muscles of 33 healthy Swiss volunteers; six other volunteers received a placebo (alum alone). Twenty-six participants reported 51 immunization-related adverse events, mainly pain at the injection site. Mean antibody titers to CS protein and MSA-2 in an indirect immunofluorescence assay peaked four weeks after the second immunization without evidence of boosting (i.e., sharp increase in titer). By that time, 56% and 31% of the vaccinees seroconverted to CS protein and MSA-2, respectively, with the increase in MSA-2 titer being weaker than that for the CS protein. After a third immunization, five vaccinees volunteered to be challenged by three or four infective bites of Anopheles stephensi. Prepatent and incubation periods in all five were comparable with unvaccinated historic controls challenged under similar conditions, and all had symptoms of clinical falciparum malaria. We conclude that the vaccine components were safe and immunogenic but there was no evidence that this immunization regimen with the CS protein plus MSA-2 component was able to prevent infection. Because the protocol required treatment in ≤ 24 hr after detection of parasitemia, the contribution of MSA-2 as an anti-disease vaccine candidate will be further evaluated in a field trial.

Murine immunoglobulin G (IgG) subclass responses to immunization are restricted to certain subclasses depending on the nature of the immunogen. Immunization with live viruses generally leads to a predominant IgG2a response, which may be the most effective at resisting future challenge due to the unique effector functions of IgG2a. Knowledge of subclass responses following immunization with dengue vaccine candidates may be helpful in determining which candidates are most efficacious. We measured the dengue-specific IgG subclass responses of BALB/c mice following immunization with live dengue-2 virus or with a partially purified recombinant dengue-2 envelope (E) protein. Subclass responses following immunization with live virus were IgG2a > IgG1 > IgG2b > IgG3, as opposed to IgG1 > IgG2a > IgG2b > IgG3 after immunization with recombinant protein. Responses of all subclasses except IgG1 were greater following immunization with live dengue than with the recombinant E protein. Neutralizing antibody titers were also higher after immunization with live virus than with E protein and were positively correlated with dengue-specific IgG2a responses in mice immunized with recombinant E protein. Following separation of the four IgG subclasses by chromatography, the IgG2a fraction exhibited the greatest neutralizing activity. The results seen after immunization with live dengue virus or recombinant E protein in this study are in concordance with studies involving other viruses and viral proteins and may have implications for the development of an effective vaccine for dengue.

The Journal is now accepting advertisements, including those for positions available, seminars and other educational opportunities, and products. For further information or to place an ad, please contact the Editorial office of the Journal.

Visiting Scientist Harvard School of Public Health. The Department of Tropical Public Health invites applications for a Visiting Scientist to participate in research studies conducted in Boston and Brazil on schistosomiasis, leishmaniasis, or malaria funded through a National Institutes of Health/International Centers for Infectious Disease Research Award. Successful applicant(s) will spend from two months to one year at one of the following study sites: Salvador, BAHIA, Fortaleza, CEARA or Rio de Janeiro, RIO DE JANEIRO. Applicants must have a Ph.D. or M.D. degree and interest/relevant experience in epidemiology, immunology and pharmacology. Please send research proposal, curriculum vitae and three letters of reference to: Dr. Donald Harn, Professor, Harvard School of Public Health, Department of Tropical Public Health, 665 Huntington Avenue, Boston, Massachusetts 02115 or to FAX# (617) 432-1257.