The American Journal of Tropical Medicine and Hygiene - Volume 32, Issue 2, March 1983

It is indeed a great honor that the Society has given to me in inviting me to come here to present the 47th Annual Charles Franklin Craig Lecture. I feel greatly indebted, and want to begin by expressing my best thanks to its officers and to the members of the selection committee. It is a unique privilege that not many foreigners have received and that I highly appreciate. It is natural that on this occasion I should speak about malaria, the disease which has occupied most of my professional life, and that I refer particularly to the work carried out in Venezuela, and to its antecedents and development.

Malaria was the most deadly endemic and epidemic disease prevalent in this country before 1936, the year in which the activities for its control on a national scale were established under my direction. It covered two-thirds of the territory, 600,000 km.

Results are presented from the 1,000 slide-confirmed malaria cases seen during the period August 1979–September 1981 at San Lazaro Hospital, in Manila, Philippines; 56% were caused by Plasmodium falciparum, 38% by P. vivax, 6% were mixed infections, and 0.1% by P. malariae. The overall case fatality rate was 1%, all due to P. falciparum. Cerebral involvement occurred in 7% and the case fatality rate was 20% compared to a case fatality rate of 0.2% among P. falciparum cases without cerebral involvement. In vivo chloroquine-resistant P. falciparum was seen in 4% of the cases, but of those treated in 1981, 9% of the cases showed resistance. The distribution of chloroquine-resistant cases by province in the Philippines is shown, with resistance being reported for the first time from Isabela, Bulacan, Zambales, Rizal and Bataan provinces. Diagnostic, clinical, and epidemiologic aspects of the cases are discussed, as well as the trend in malaria cases over the last 20 years.

The first two cases from East Africa of RII chloroquine- and Fansidar®-resistant falciparum malaria are described. The first case occurred in a non-immune Swedish expatriate 2 weeks after arrival in Tanzania, and the second in a semi-immune Tanzanian soldier. Fansidar has not yet been marketed in Tanzania. The significance and etiology of the occurrence of combined chloroquine/Fansidar resistance in East Africa are discussed.

Evidence from in vivo experiments and clinical reports in humans has indicated that some antibiotics demonstrate antimalarial activity. Twelve antibiotics have been tested against Plasmodium falciparum in vitro, including cycloheximide, streptomycin, erthromycin, tetracycline, chloramphenicol, clindamycin, actinomycin D, rifampin, nalidixic acid, penicillin G, chlorhexidine, and isoniazid. Inhibitory effects obtained at various drug concentrations in vitro were compared to drug levels reported to be effective against bacteria in vivo. Several antibacterial drugs, including erythromycin, chloramphenicol, clindamycin, rifampin and tetracycline, demonstrated significant antiparasitic effects at concentrations within or near those observed during therapy in vivo; the potency of erythromycin and tetracycline was greater at 96 hours of exposure than at 48 hours.

Variants (K-) of three strains of Plasmodium falciparum which do not produce the erythrocyte surface alterations that have been called knobs have been compared with their wildtype knobby (K+) parents. The K- variants achieve higher parasitemias, incorporate radiolabeled isoleucine more rapidly, and produce a higher percentage of multiply-infected cells than do their K+ parents. Nevertheless, immune owl monkey sera cause approximately the same percentage inhibition of growth of both K+ and K- organisms when included in the growth medium at a 1% concentration.

Compounds (265) of widely diverse structures were appraised for radical curative activity in rhesus monkeys infected with sporozoites of the B strain of Plasmodium cynomolgi, using an evaluation system that provided a preliminary assessment with from 0.1–1.0 g of compound and tests against one to five active infections. None of 32 compounds in a miscellaneous structure category, none of seven agents of antibiotic origin, none of 12 1,5-naphthyridines, and none of seven 7-aminoquinolines exhibited curative activity at the largest test doses. There was a suggestion that one of 12 newly synthesized pyrocatechols was curative. Two of 20 6-aminoquinolines effected cure at or near maximum tolerated doses. In contrast, 90 of 174 newly synthesized 8-aminoquinolines effected cure; 18 of the 90 being as active as primaquine, eight twice as active, and six four times as active. There were major disagreements between the above results and those recorded by others in mice inoculated with sporozoites of P. berghei yoelii or P. yoelii nigeriensis. These discrepancies were of serious dimensions in evaluations of the 8-aminoquinolines. This, plus previous near flawless performances of P. cynomolgi in identifying agents that would cure naturally acquired P. vivax infections, led to the suggestion that the abbreviated simian model employed in these studies be used hereafter in primary screening of new agents for radical curative activity.

Thirty-four rabbits were experimentally infected with trypomastigotes of either the Ernestina or the Albuquerque strains of Trypanosoma cruzi. These animals showed patent parasitemias, as demonstrated by xenodiagnosis, in the acute phases of the infections. Typical chagoma signs developed in two rabbits 1 week after parasite inoculation in the skin, although the acute phase of Chagas' disease in the rabbit model was usually asymptomatic. In the 6th month of infection the parasitemias became negative and the infections remained subpatent, as indicated by the persistence of positive serologic tests and of delayed-type skin reactions elicited in Chagas' rabbits against a microsomal T. cruzi antigen. This latent infection continued asymptomatically, in the absence of electrocardiographic (ECG) alterations. However, ECG changes consistent with enlargement and overload of cardiac chambers, alterations of ventricular repolarization, S-T changes and bundle-branch blocks were frequently recorded later in the chronic phase of Chagas' disease. The pathological manifestations of these ECG alterations were confirmed at the autopsy of each experimental rabbit. Congestive heart failure and pulmonary thromboemboli related to chronic myocarditis of Chagas' disease were frequent causes of death. Megacolon was seen in two rabbits inoculated with the Ernestina strain of the parasite. The relatively limited duration of detectable parasitemia even when xenodiagnosis is used, the lack of correlation between parasitemia and severity of pathological manifestations, and the fact that all infected animals showed histopathological evidence of myocarditis, destructive inflammatory lesions characterized by mononuclear infiltrates in skeletal muscles, as well as cutaneous delayed hypersensitivity to T. cruzi antigens, are notable observations in this animal model of the human disease.

The protective effects of immunization with an antigen derived from epimastigote forms of Trypanosoma cruzi against challenges with vector metacyclic or bloodstream forms of the parasite were investigated. A marked degree of protection was observed in the immunized mice after challenged with bloodstream or insect-transmissible metacyclic forms. High rates of survival in the immunized groups were accompanied by relatively low, short-lasting parasitemias in some of the animals and significant percentages of immunized mice never developed a measurable parasitemia during the course of the experiments. In contrast, all of the non-immunized animals showed high levels of parasitemias. Similar results were obtained whether the metacyclic challenge was by the intraperitoneal or by the ocular route for which conditions mimicking a natural infection were selected. These results emphasize the protection that immunization confers against challenge with insect-transmissible forms of T. cruzi, and the feasibility of protecting a highly susceptible host against an otherwise lethal acute infection similar to one occurring naturally.

BALB/c, C57Bl/6 and (BALB/c × C57Bl/6)F1 mice all proved susceptible to infection by a strain of Leishmania isolated from a Central Brazilian with espundia. The course of disease differed markedly between BALB/c and C57Bl/6 mice. BALB/c mice suffered from a rapidly progressive and widely metastatic, but non-ulcerative, disease resembling diffuse cutaneous leishmaniasis. In contrast, C57Bl/6 mice initially contained parasite multiplication effectively and appeared clinically cured. However, the parasite could persistently be cultured up to about 1 year post-infection. At that time, the parasite load in the infected footpad increased and a patent disease developed characterized by distinctive ulcerative metastases with destruction of soft-tissue in the nasal region similar to the one observed in espundia. Development of disease in both strains of mice was associated with depression of cell-mediated immunity as monitored by delayed-type hypersensitivity in vivo and lymphocyte transformation in vitro. Thus, our study suggests that diffuse cutaneous leishmaniasis and espundia can be caused by the same strain of parasite, and that the particular clinical expression in the individual mouse is determined by the host response.

A regimen of combined immunostimulation and chemotherapy for the elimination of Leishmania donovani amastigotes was evaluated. An in vitro experimental model utilized cultured peritoneal macrophages from C57Bl/6 mice infected with L. donovani tissue forms. Partial or complete activation of macrophages as judged by killing of tumor cells significantly enhanced the efficacy of sodium antimony gluconate (Pentostam®). The quantity of drug required for elimination of parasites from immunostimulated cells was considerably lower than that required to achieve comparable amastigote killing in thioglycolate-elicited macrophages. In contrast, amphotericin B cleared infected cells of amastigotes at comparable drug levels when tested with immunostimulated and unstimulated macrophages. Several drugs tested inhibited the conversion of amastigotes to promastigotes in vitro but were ineffective in killing of intracellular tissue forms. Allopurinol and difluoromethylornithine (DMFO) blocked amastigote conversion significantly. These drugs at high concentrations, however, exerted only minimal toxicity for amastigotes residing within macrophages. Efficacy of combined therapy was also demonstrated in vivo. Immunoenhancement of L. donovani-infected mice with Corynebacterium parvum vaccine combined with a regimen of sodium antimony gluconate was significantly more effective than was immunotherapy or drug therapy alone.

Three culture media were compared with Giemsa-stained smears for the detection of Leishmania in splenic aspirates from Kenyan patients with visceral leishmaniasis. Ninety-nine splenic aspirates obtained from 26 patients at various times before, during, and after treatment were cultured in Schneider's Drosophila medium and RPMI medium 1640 (each supplemented with 20% fetal bovine serum) and McConnell's modification of Senekje's medium overlayed with 0.9% saline. From 13 splenic aspirates obtained before treatment, amastigotes were identified microscopically in all and promastigotes were cultured in 12. During and after treatment, Schneider's medium was the most sensitive method for detecting parasites, followed by microscopic examination of stained smears which was more sensitive than either of the other two media tested. Results indicate that, for initial diagnosis, both culture and direct microscopy of aspirates should be employed.

Infection of mouse peritoneal macrophages in vitro was used to examine the effect of elevated temperature on the intracellular replication of various strains of Leishmania. Of eight cutaneous strains examined, all grew optimally at 35°C. At 37°C the reduction in growth was most pronounced for the New World cutaneous strains, and at 39°C three of four New World cutaneous strains were completely destroyed whereas all of the L. tropica strains survived and exhibited at least 100% growth after 3 days. The results of these in vitro studies correlate closely with the outcome of heat therapy on two patients with cutaneous disease, suggesting that, in general, cutaneous lesions due to L. tropica strains might be less responsive to heat therapy than lesions due to L. mexicana and related strains.

St. Louis encephalitis (SLE) virus has continued to be active in the Colorado Desert region of Southern California (Coachella and Imperial Valleys) since 1973, while it has virtually disappeared from the Sacramento and San Joaquin Valleys of Central California. Consequently, comparative vector competence studies were undertaken from 1978 to 1981 to evaluate the susceptibility of populations of Culex quinquefasciatus and Culex tarsalis from the San Joaquin and Coachella/Imperial Valleys to oral infection with representative SLE viral strains from each geographical area. Culex quinquefasciatus females from both areas were equally susceptible to infection with both viral strains by the pledget and viremic chick feeding techniques. Although susceptibility profiles were similar by both feeding techniques, infection thresholds (i.e., ID50s) obtained with Cx. quinquefasciatus females were at least 10,000-fold lower when fed on viremic chicks than on virus soaked pledgets. Culex tarslis females from the two geographical areas were equally susceptible to infection with the indigenous viral strain by feeding on pledgets, and were uniformly more susceptible than sympatric populations of Cx. quinquefasciatus. Thus, differences observed in SLE viral activity in different geographical areas of California since 1973 cannot be explained by differences in the susceptibility of Cx. quinquefasciatus or Cx. tarsalis to oral infection with SLE virus.

Two hundred and fifty-seven Egyptian patients were classified into three groups: patients with schistosomal colonic polyposis, those with simple schistosomiasis without polyposis, and a non-schistosomal group. A diagnosis of schistosomiasis was made by clinical history and examination plus three fresh stool examinations or a rectal biopsy. The presence of schistosomal colonic polyps was established by sigmoidoscopy and biopsy of polyps. Stool examinations were made on all individuals, using the merthiolate-iodine-formaldehyde technique to detect Entamoeba histolytica. We found the prevalence of amebiasis in the group with schistosomal colonic polyposis (37%) to be significantly higher than that in the non-schistosomal group (11%) and in the schistosomal group without polyposis (15%). The difference in prevalence of amebiasis between the simple schistosomal and non-schistosomal groups was not significant.

Liver biopsies of four patients with hepatosplenic schistosomiasis, two patients with schistosomiasis and chronic active hepatitis, two patients with chronic active hepatitis and four control patients with no clinical evidence of either disease, were examined by standard light microscopic techniques, electron microscopy and immunocytochemical staining for collagen type I, III and B. Pure schistosomiasis showed the classical “clay-pipe stem fibrosis” and granulomata composed of eosinophils, macrophages and lymphocytes. In that group, the hepatocellular damage was less conspicuous than in the groups with chronic hepatitis and was usually confined to the granulomatous or fibrotic areas. Destruction of the normal architecture and infiltration by macrophages and lymphocytes with severe damage of hepatocytes was found only in the cases of chronic active hepatitis, with or without associated schistosomiasis. Increased collagen deposits were demonstrated in all three groups. Types I, III and B were found in the enlarged portal triads and fibrotic septa. The intranodular or intralobular collagen stained negatively for type I and strongly positive for types III and B.

The proliferative responses of peripheral blood mononuclear cells from patients with Schistosoma mansoni infections were studied in response to heterogeneous schistosomal antigenic preparations derived from whole adult worms, cercariae or schistosomula, and soluble tegumental preparations from adult worms or schistosomula. The use of these preparations was standardized with a Brazilian patient population, and comparisons were made between previously used and newly, variously prepared antigenic preparations. Responses of cells from these intestinal/ambulatory patients were generally strong to whole adult worm antigens and low to moderate to whole cercarial or schistosomular materials. Although most patients responded well to whole worm extracts, they were not stimulated by the soluble adult tegumental preparation. In contrast, the responses to the soluble schistosomular tegumental material were vigorous. These responses were usually much higher than to whole schistosomula-derived materials. Thus it appears, using these particular preparations, that human schistosomal patient lymphocytes recognize and are stimulated by schistosomular tegumental antigens but they do not demonstrate good reactivity when exposed to an adult worm tegumental preparation.

A monoclonal antibody has been used to identify and characterize an antigenic tegumental surface membrane glycoprotein of Schistosoma mansoni. Direct binding of 125I-labeled monoclonal antibody showed that this glycoprotein was present in eggs, cercariae, and worms of both sexes. The glycoprotein had an apparent molecular weight of 180,000. Indirect and direct immunofluorescent microscopy showed that this antigen was located on the interlinked tegumental folds of both larval and adult parasites. These findings are discussed in relation to parasite development and the mechanism by which schistosomes evade the host's immune defenses.

Introduction of synthetic antigens into the surface of schistosomula of Schistosoma mansoni was achieved by brief incubation of the worms with liposomes carrying the lipid-bound antigens in their bilayers. Three-hour-old schistosomula were surface-labeled with lipid-conjugated dinitrophenyl (DNP) groups by using liposomes made of egg lecithin-N-dinitrophenil-ε-aminocaproyl-phosphatidylethanolamine (5:1). The DNP groups incorporated in this way could be detected for more than 21 hours in vitro by using rabbit anti-DNP antibodies stained with fluorescein isothiocyanate-conjugated goat anti-rabbit IgG. Immunofluorescence microscopy showed the lipid antigen to be uniformly distributed over the entire surface of the worms. Electron microscope studies, performed with purified rabbit anti-DNP antibodies followed by ferritin-conjugated goat anti-rabbit IgG, showed that the DNP groups were evenly and densely distributed over the entire outer membrane of the schistosomula, including spines. The distance between the ferritin molecules and the parasite's surface was 24 ± 5 nm, indicating that the lipid antigen had been incorporated into the outer membrane of the schistosomula.

The presence of cross-reacting antigens between Paragonimus westermani, Schistosoma mansoni, and Fasciola hepatica adult worms was demonstrated by Ouchterlony immunodiffusion and enzyme-linked immunosorbent assay (ELISA). A serum bank was developed against the three trematode genera to serve as probes to determine the presence of cross-reacting antibodies to P. westermani worm extracts. In this manner, it was possible to demonstrate that antigens common to F. hepatica and S. mansoni tegument were also present in P. westermani worm extracts. Likewise, it was possible to demonstrate that the F. hepatica antigens which bind to Concanavalin A, as well as the subfraction which in isoelectric focusing has a pI of 4.2, were also found in P. westermani worms. Also, a monospecific polyclonal serum to a Fasciola/Schistosoma cross-reacting antigen and the anti-P. westermani serum both reacted in Ouchterlony immunodiffusion with the P. westermani antigenic extract, each producing a line which linked with each other indicating common antigenic determinants and suggesting a common antigen among the digenetic trematodes. Finally, the P. westermani antigenic extracts induced in mice the production of antibodies which reacted with S. mansoni adult worm antigens by ELISA. As all of the Fasciola and Schistosoma sera were prepared against antigenic preparations which induced in mice protection to challenge infection with S. mansoni, this suggested that the P. westermani worms also contain protective antigens against S. mansoni. Immunity to Schistosoma mansoni infection was induced in mice by vaccination with Paragonimus westermani whole worm extracts (PwWWE). Immunized mice showed as high as a 67% worm burden reduction over controls. High doses of PwWWE did not confer protection to S. mansoni infection. Thus, in this study, immunity in heterologous systems was demonstrated and the existence of a common protective antigen shared by the digenetic trematodes was suggested.

Significant chemotactic activity for eosinophils was detected in soluble egg antigen (SEA) preparations of both Schistosoma japonicum and S. mansoni in dose-dependent fashion. The activity of S. japonicum SEA was higher than that of S. mansoni SEA. Gel filtration on Sephadex G-150 showed that S. japonicum SEA was composed of two groups of eosinophil chemotactic factors (ECFs), one of high molecular weight (JEE-H) and the other of low molecular weight (JEE-L). S. mansoni SEA showed ECF composition similar to that of S. japonicum SEA. JEE-H was stable on heating (100°C, 60 min) and resistant to pronase digestion, but was sensitive to periodate oxidation. JEE-L was also stable on heating and resistant to pronase and carboxypeptidase A digestions. These properties of the ECFs were also held in common with those of S. mansoni SEA. JEE-L was extractable by toluene, indicating a hydrophobic nature. These results suggest that schistosome eggs themselves contain ECFs, and that the composition of S. mansoni and S. japonicum SEA-derived ECFs is essentially the same. However, they differ from the other ECFs which have already been described in schistosome infections.