Indian Pediatrics 2003; 40:1029-1034 

Use of Vaccines for the Prevention of Typhoid Fever

The period 1990 to the present has been a hallmark era in the history of typhoid fever because of the emergence and dissemination in Asia and parts of Africa of S. typhi strains carrying resistance to multiple clinically relevant antibiotics(1,2). Even as antibiotic-resistant typhoid fever spread in the Indian sub-continent, new knowledge was being generated on the pathogenesis of typhoid fever and the human immune response to the pathogen and new typhoid vaccines were being developed. This editorial will review this period.

Typhoid fever, the generalized infection of the reticuloendothelial system (spleen, liver, bone marrow), gut-associated lymphoid tissue and gall bladder caused by Salmonella enterica serovar Typhi, is the quintessential infectious disease associated with inadequate sanitation, and the pediatric population disproportionately bears the burden of clinical disease. In most endemic areas, chronic gall bladder carriers (usually adult females) who excrete large numbers of typhoid bacilli constitute the main reservoir of infection. Where sanitation and treatment of water supplies is inadequate, fecal contamination from carriers and patients can make water sources important vehicles of transmission, particularly under crowed conditions. Consumption of contaminated water and food vehicles by susceptibles results in clinical or sub-clinical infection (depending on dose ingested, precise vehicle and host factors). Depending on age, the presence of pre-existent gall bladder pathology and the antibiotic used to treat the acute typhoid illness, up to a few percent of infected persons can become chronic carriers, thereby maintaining the reservoir of infection. S. typhi is highly restricted to human hosts and a fairly long incubation period (8-14 days) passes before the onset of clinical disease.

Typhoid fever first emerged as a recognized clinico-epidemiological entity in the early 19th century in Europe and North America in association with the industrial revolution when the population of cities exploded and the combination of extreme crowding, unsanitary conditions and lack of potable water amplified the transmission of typhoid. By the mid-19th century, an encounter with S. typhi had become virtually an urban rite of passage. In the late 19th and early 20th century, the treatment of municipal water supplies in Europe and North America interrupted the amplified transmission of S. typhi and led to a plummeting of incidence rates and typhoid mortality(3). Conse-quently, by the mid-point of the 20th century typhoid had become a disease whose endemicity was increasingly restricted to developing countries. It is often forgotten that in the pre-antibiotic era typhoid fever was commonly fatal. Indeed, until the late 1940s there was approximately a 15% case fatality among cases of typhoid fever worldwide.

In 1948 a breakthrough occurred in the treatment of typhoid, as Woodward et al.(4) working in Southeast Asia found that chloramphenicol could successfully treat typhoid fever, dropping the case fatality to <1%. Over the next quarter century the control of typhoid fever mortality was accomplished using oral chloramphenicol as a simple, practical and effective treatment for typhoid fever. Curiously, during the 1950s and 1960s, when Shigella and non-typhoidal Salmonella strains rapidly acquired R factor plasmids encoding resistance to multiple antibiotics, S. typhi remained sensitive. The reasons for this were not apparent but it was a welcome fact. Thus, the appearance of epidemics of chloramphenicol-resistant S. typhi strains in Mexico in 1971-72(5) and in Vietnam in 1973-74(6), constituted a rude awakening. This led to a quest for alternative effective oral antibiotics (trimethoprim/sulfamethoxa-zole and amoxicillin were found to serve this purpose)(5) and to a renewed interest in typhoid vaccines.

Enigmatically, in Mexico circa 1974 and in Peru in 1981 chloramphenicol-sensitive strains reappeared to replace the resistant strains that had caused epidemic disease during the previous two years in each country. In the late 1980s, oral fluoroquinolones, particularly ciprofloxacin, came on the scene as virtual "miracle drugs" for the treatment of typhoid. However, concerns about their safety (effect on growth cartilage) in young children initially limited their use in pediatrics. In contrast with earlier oral antibiotics, full-dosage ciprofloxacin therapy was rarely followed by relapse or chronic carrier state. It is against this background that the emergence circa 1989 in Asia and Northeast Africa of strains of S. typhi carrying resistance to multiple antibiotics posed a public health crisis. For the first time, S. typhi was stably carrying a plasmid encoding resistance to almost all the clinically relevant antibiotics including chloramphenicol, tri-methoprim/sulfamethoxazole, ampicillin and amoxicillin. In the Indian sub-continent and in Southeast Asia cases of severe typhoid were being seen that were difficult to treat, manifested complications and exhibited increased case fatality(1,7).

Ciprofloxacin and other fluoroquinolones were initially highly efficacious but the inappropriate use of these antibiotics (inade-quate dosage, ultra-abbreviated courses and sub-potent drug) eventually led to the selection of S. typhi strains with chromosomal muta-tions that conferred partial resistance to fluoroquinolones(8). Infection with these partially resistant strains resulted in a slow clinical response even when a full course of potent fluoroquinolone was given at proper dosage.

In some regions of Asia, as in Jakarta during the 1970s and 1980s, a severe form of clinical typhoid accompanied by extreme obtundation was not uncommonly seen in tertiary care hospitals(9). Despite appropriate antibiotic therapy, the case fatality of Jakarta severe typhoid was ~50% unless steroids (e.g., dexamethasone) were concomitantly adminis-tered in high dosage; the adjunct steroid therapy dropped the case fatality to ~10%(9). It is not known if this severe typhoid syndrome was due to enhanced virulence of the Javanese S. typhi strains, host factors, or ingestion of huge inocula in vehicles that protected the bacilli during gastric transit. It is also possible that such cases were in fact rare events if calculated as a population-based incidence but appeared to be common numerator events in tertiary care hospitals because they derived from an enormous overall typhoid disease burden in Jakarta. S. typhi affects many organ systems and during the last decade when multiply resistant typhoid fever was prevalent in Asia, complications were seen in South and Southeast with a much greater frequency than when strains were antibiotic-susceptible.

The entire genome of S. typhi has been sequenced(10) generating hope that this information can be exploited to achieve improved diagnostic tests and to identify targets for new antimicrobials. In the meantime, the genomic data have provided insights on pathogenesis, molecular epidemiology and the relationship of S. typhi to other Salmonella.

Even as multiply-resistant strains of S. typhi were spreading in Asia in the 1990s, two new typhoid vaccines were becoming widely used to protect travelers from industrialized countries(11). Attenuated strain Ty21a, a live oral vaccine, and purified Vi polysaccharide, a parenteral vaccine, replaced the highly reactogenic heat-inactivated phenol-preserved whole cell parenteral vaccine that had been around for a century(11). The two new vaccines are very well tolerated and provide as good (or better) protection than the killed whole cell parenteral vaccine. Ty21a exists in two formulations, enteric-coated capsules and a so-called "liquid formulation" (lyophilized vaccine that is reconstituted with water and buffer to make a vaccine cocktail)(12,13). Three doses of Ty21a are administered on an every other day schedule. Vi polysaccharide vaccine is given as just a single parenteral dose because serologic responses are not boostable by administering additional doses(14). A large experience attests to the safety, practicality and effective-ness of these vaccines as public health tools in other parts of the world. Notably, Ty21a confers rather long-term protection(15). In Chile in the 1980s, three doses of enteric-coated Ty21a conferred upon school children 62% protection over seven years and three doses of the liquid formulation provided 78% protection over five years of follow-up(15). There is evidence that large-scale immuniza-tion with Ty21a led to a herd immunity effect(16). Since good results have been achieved with these two vaccines when administered in school-based mass immuniza-tion programs in other venues(17,18) one cannot help but wonder what might have been seen in some sites in the Indian Sub-continent if public health authorities had undertaken demonstration projects with these vaccines to assess their utility as control measures during the height of the epidemic of resistant typhoid in the 1990s.

Several new typhoid vaccines that offer improvements over Ty21a and Vi have been progressing through clinical trials. These include: (1) several engineered strains of S. typhi(19-21) that appear promising as candi-date single-dose live oral vaccines because of their increased immunogenicity over Ty21a, and; (2) a parenteral Vi-conjugate vaccine that stimulates higher titers of Vi antibodies than unconjugated Vi polysaccharide and that elicits immunologic memory(22). In addition to their administration to school age children, these vaccines appear amenable to use in infants in the Expanded Program on Immunization (EPI).

Events of the last decade have taught that antibiotics can no longer be counted on as the main public health measure to control typhoid fever. We are rapidly approaching the point where the only reliable therapy may derive from the use of expensive parenteral antibiotics like ceftriaxone, thereby making treatment costly and cumbersome in developing countries. If all urban and sub-urban populations in India can be provided with treated, bacteriologically monitored water supplies and sanitation to dispose of human waste this will constitute the best solution to eliminate endemic typhoid fever. While that may be a laudable and achievable long-term goal, in the short and medium term it is unlikely that such services will become universally available to solve the typhoid ‘problem. This leaves vaccination as a potential option in the short and medium term. If they are to be used, vaccines will have to be a reasonably cost effective investment as a measure to control typhoid fever. Recognizing this, careful consideration will have to be given to identify the precise target popula-tions, the types of vaccine to be used, and the infrastructure to deliver vaccine. Other relevant factors in the equation will be the expected duration of protection, the economic cost of disease and the cost of vaccine.

In most endemic areas, the syndrome recognized by clinicians as typhoid fever exhibits its peak incidence in school age children 5-19 years of age(23-25). However, a series of systematic studies has shown that bacteremic S. typhi infection also occurs in children <5 years, including in infants and toddlers <2 years of age(23-25). These epidemiologic data were generated by systematically collecting blood cultures from children with fever seen at health centers or detected during household surveillance visits. Circa 2-4% of febrile children sampled in this way yield S. typhi or S. paratyphi in their blood cultures(23-25). In pre-school children 2-4 years of age, the clinical syndrome often resembles typhoid fever as seen in older children. However, in children <24 months of age, although occasional cases may be clinically severe, most of the cases of bacteremic S. typhi infection detected by such active surveillance methods exhibit a non-specific mild febrile illness. This mild illness is indistinguishable from the other 96-98% of febrile children, a few percent of whom have "occult" pneumococcal bacteremia and most of whom have self-limiting viral infections. This information is important to decision makers. Although the total numbers of toddlers with mild S. typhi febrile illness and bacteremia may be large when detected by household surveillance, the S. typhi disease burden that results in health care visits, hospitalizations, disease complications and deaths mostly falls within older pre-school and school age children. Consequently, older pre-school and school age children are the most important target groups for disease prevention.

While exciting new vaccine candidates are in late development and may be options available in the future, oral Ty21a and parenteral Vi constitute the well tolerated vaccines available today. With respect to delivering typhoid vaccines to target populations, the EPI offers an established infrastructure that is already delivering multiple vaccines to infants at scheduled ages. However, EPI is not an option for Vi vaccine since as a polysaccharide it is poorly immunogenic in infants. Moreover, even if future typhoid vaccines such as Vi conjugate or new live oral vaccines are to be considered for delivery through the EPI, it would have to be clearly demonstrated that they confer upon infants long-term immunity that endures through the critical pre-school and school age risk period. These considerations argue that if demonstration projects are to be carried out with Ty21a and Vi they should target school children and be delivered through urban school-based immunization. Other forms of mass immunization campaigns would be required to reach 2-4 year old pre-school children (polio vaccine campaigns could serve as a model).

Attempts to control typhoid fever using vaccine will require a significant financial investment. If immunization is limited to high incidence populations in regions where antibiotic-resistant strains are prevalent, and if the vaccine is highly effective and elicits long-term protection, it is likely that vaccination will prove to be a sound public health investment.

Myron M. Levine,
Center for Vaccine Development,
University of Maryland School of Medicine,
685 W Baltimore St.,
Baltimore, MD 21201, USA.
E-mail: [email protected]

 References