Safety and Efficacy of a Typhoid Conjugate Vaccine in Malawian Children

Background Typhoid fever caused by multidrug-resistant H58 Salmonella Typhi is an increasing public-health threat in sub-Saharan Africa. We present phase 3 efficacy data from an African trial of a Vi-polysaccharide typhoid conjugate vaccine (Vi-TCV). Methods Children aged 9 months to 12 years in Blantyre, Malawi were randomized (1:1) in a double-blind trial to receive Vi-TCV (single dose) or group-A meningococcal control vaccine (MenA).The primary outcome was blood culture-confirmed typhoid fever. We present the primary vaccine efficacy (VE) and safety outcomes after 18–24 months of follow-up. Results This intention-to-treat (ITT) analysis included 28,130 children, comprising 14,069 children who received Vi-TCV and 14,061 children who received MenA. Blood culture-confirmed typhoid fever occurred in 12 children in the Vi-TCV group (46.9 per 100,000 person-years) and 62 children in the MenA group (243 per 100,000 person-years). Overall VE was 80.7% (95% confidence interval (CI): 64.2% to 89.6%) in an ITT analysis, and 83.7% (95% CI: 68.1%−91.6%) in a per-protocol analysis. In total, 130 serious adverse events occurred in the first 6 months after vaccination (52 in Vi-TCV group and 78 in MenA group), including 6 deaths (all in MenA group). No serious adverse event was considered by the investigator as related to study vaccination. Conclusions Vi-TCV reduced blood culture-confirmed typhoid fever among Malawian children aged 9 months to 12 years. (Funded by the Bill & Melinda Gates Foundation; ClinicalTrials.gov number NCT03299426.)

No serious adverse event was considered by the investigator as related to study vaccination. Typhoid fever, a systemic febrile illness caused by Salmonella enterica serovar Typhi (S. typhi), is responsible for more than 9 million infections and over 110,000 deaths globally each year, with the highest disease burden among school-age and pre-school children. 1,2 An estimated 1.2 million typhoid cases and 18,703 deaths occur annually in sub-Saharan Africa, with 383-843 cases per 100,000 person-years reported in some urban settings. [3][4][5] The increased public health importance of typhoid fever across sub-Saharan Africa over the past decade is due in part to emergence and spread of several multidrug resistant (MDR; resistant to first-line agents chloramphenicol, ampicillin, and co-trimoxazole) S. typhi lineages, particularly H58 (clade 4.3.1) and H56 (clade 3.1.1). 6,7,8 In Malawi and other countries in East and Southern Africa, MDR H58 S. typhi emerged in 2010 following its introduction from Asia, 9,7 becoming the predominant blood-stream infection among adults and children, with 2% case fatality and 5% rate of small bowel perforation. 10,11,12 Emerging antimicrobial resistance to fluoroquinolones has been documented in East Africa, 7,13 Nigeria, 6 and Democratic Republic of Congo. 14 Extensively drug-resistant (XDR) typhoid, resistant to fluoroquinolones and 3 rd generation cephalosporins, is established in Pakistan. 12 The dual threat in Africa of local emergence or introduction of untreatable XDR typhoid from Asia underscores the need for typhoid fever prevention. 15 In 2018, the World Health Organization (WHO) recommended typhoid conjugate vaccine (TCV) for children 6 months through 15 years of age in countries with high incidence of disease or antimicrobial resistance. 16 Typbar TCV® (Bharat Biotech International) is a WHO-prequalified typhoid conjugate vaccine. The Typhoid Vaccine Acceleration Consortium (TyVAC) was launched in 2017 with the aim to accelerate Vi-TCV introduction in low-income settings. TyVAC is conducting large, randomized controlled efficacy trials of a Vi-TCV in diverse epidemiological settings in Malawi, Nepal, and Bangladesh. [17][18][19] Here we present vaccine efficacy and safety results for Vi-TCV from the African continent, 20,21 through 18-24 months of follow-up, from a clinical trial of single dose Vi-TCV in Blantyre, Malawi. 22

Study design and participants
This is a single center, phase 3, double-blind, individually randomized active-controlled trial in two urban townships in Blantyre, Malawi. Detailed methods have been published. 22,23 Briefly, a target of approximately 28,000 healthy children were enrolled, aged 9 months through 12 years residing within the urban townships of Ndirande and Zingwangwa, whose parents/guardians provided written consent, with no previous history of typhoid vaccination, and no acute illness or history of allergy or hypersensitivity. Written assent was required for children aged ≥8 years.
HIV status was solicited verbally and confirmed, if positive, using the participant's healthpassport, where possible. Participants were recruited through government health centers and primary schools. Safety data (Adverse Events (AEs) and Serious Adverse Events (SAEs)) were prospectively recorded.

Randomization and masking
Participants were randomized at a 1:1 ratio to receive a single dose of Vi-TCV or control meningococcal group A conjugate vaccine (MenA), using block randomization with block sizes from 6-12. The random allocation sequence was generated by the blockrand package (version 1.3) in R (version 3.4.1) and concealed before randomization, which occurred in real time immediately before vaccination. Parents, guardians, participants, and study staff involved in screening, eligibility assessment, and follow-up were fully blinded to vaccine-group assignment.
Unblinded nurses prepared and administered vaccine in a private area and had no further role in the study after vaccination.

Procedures and vaccines
Typbar TCV® consists of Vi polysaccharide conjugated to a tetanus toxoid protein carrier (25 µg Enhanced fever and safety surveillance All participants were monitored for 30 minutes after vaccination for immediate AEs. Enhanced passive surveillance for fever and SAEs was conducted at four primary health centers (Ndirande, Zingwangwa, Gateway, Nayo) in addition to Queen Elizabeth Central Hospital, a government referral hospital, where parents/guardians were instructed to bring unwell children at any time.
Usual health service provision was also enhanced by telephone and community messaging to participants. Children presenting with febrile illness (subjective fever for ≥72 hours; axillary temperature ≥38°C; or hospitalization with history of fever of any duration), had blood-culture collected (5 mL <5-year old; 10 mL >=5 year old), and malaria Rapid Diagnostic Test (RDT).
Antimicrobial resistance of S. typhi isolates was tested by disc-diffusion. 24 Isolates showing pefloxacin-resistance had confirmatory ciprofloxacin e-test (BioMerieux), minimum inhibitory concentration >0.06mg/L indicating resistance. Hospital admission and antimicrobial treatment were at the facility clinician's discretion. Participants with blood-culture -confirmed S. typhi were followed up biweekly, until asymptomatic, to monitor treatment response and outcome.

Outcomes
The primary outcome was blood-culture -confirmed typhoid fever occurring at any time after vaccination. Vaccine efficacy (VE) was calculated as (1-IRR) × 100%, where IRR is the incidence rate ratio (ratio of incidence in the Vi-TCV group compared with the MenA group). Secondary

Statistical analysis
Details of sample size and power calculations have been reported. 22 Briefly, assuming 75% VE, the minimum number of cases needed to test the null hypothesis that the vaccine has no protective efficacy (i.e., VE≤0), with 90% power, was 30. The primary analysis to test VE was based on the intention-to-treat (ITT) principle, which included all randomized children who were vaccinated and all first episodes of blood-culture -confirmed typhoid fever occurring after vaccination. In the ITT analysis, the vaccine group was defined by the vaccine randomly assigned, not by the vaccine received. Per-protocol VE analysis included children who completed the study, without any protocol deviations, received the vaccine to which they were assigned, and accrued first episodes of blood-culture -confirmed typhoid fever at least 14 days after vaccination. The incidence rate was calculated as the number of first episodes of blood-culture -confirmed typhoid fever divided by the total follow-up time. Individual follow-up time was the smallest of the following: Time to first episode of typhoid fever; time to withdrawal, loss to follow-up, death, or relocation out of study area; or time to the end of the analysis period. The incidence rate ratio (IRR) was calculated as the ratio of the incidence rate in the Vi-TCV to the MenA group, and VE was calculated as (1-IRR) X 100%. Subgroup analysis was conducted to evaluate VE for sex, study site, and <5 years vs. ≥5 years of age at vaccination; Poisson regression with the interaction term between each pre-planned subgroup of interest and the vaccine group was used to compare VE across subgroups.
Absolute risk reduction was calculated as the risk of blood-culture -confirmed typhoid fever in the MenA group minus that in the Vi-TCV group. The number needed to vaccinate was calculated as 1/absolute risk reduction, representing the number of children who need to be vaccinated to prevent one additional blood-culture -confirmed case of typhoid fever. The cumulative incidence of typhoid fever for each vaccine group was presented using the Kaplan-Meier method; and VE was estimated at 12, 18, and 24 months after vaccination using the life table method. All analyses were performed according to the pre-specified statistical analysis plan, using Stata/SE (version 16). For full details of study design and conduct see the protocol at nejm.org.  (Figure 1). Median age was 6 years (range: 0.8-12), and baseline characteristics were similar for the two study groups (Table 1).

Vaccine efficacy
Between 21 February 2018 and 3 April 2020, 7,776 children presented to a passive surveillance center and met the primary case definition. Blood cultures were collected from 7,314 (94%).
Seventy-five were positive for S. typhi. These included a 9-year-old with two typhoid fever episodes at 24 weeks and 49 weeks after vaccination; the second episode was therefore excluded from VE analyses. All 75 isolates (100%) were MDR, and 4/75 (5.3%) were ciprofloxacin-resistant.
For the ITT population, incidence of blood-culture -confirmed typhoid fever in the MenA group was similar across age groups (  Figure 1).

Safety
Three male participants in the MenA group had directly-observed AEs, all graded mild, within 30 minutes of vaccination. Two (skin rash and syncope) were deemed related to vaccination, and one (diarrhea) was deemed unrelated.

Discussion
In this field trial of Vi-TCV in Africa, a single dose Vi-TCV was efficacious in preventing typhoid fever among children 9 months to 12 years of age. The endemic typhoid incidence observed across both school-age and pre-school children in the control group (243/100,000 person-years) was reduced by 80.7% in the Vi-TCV group and Vi-TCV efficacy remained consistent across children <5 years or ≥5 years at vaccination, and throughout the observation period.
Encouragingly, our VE (80.7% ITT; 83.7% per-protocol) in Malawi after 18-24 months is consistent with previously reported Vi-TCV VE of 81.6% after 12 months of follow-up among Nepalese children aged 9 months through 16 years of age. 17 The safety profile was reassuring, with no excess SAEs in the Vi-TCV group and no AE or SAE considered related to Vi-TCV. The 6 deaths within the first 6 months occurred in the MenA group and the one death from typhoid, 7 months after vaccination, also occurred in the MenA group.
Typhoid vaccines have previously been trialed in Africa. Among 23,075 South African children aged 5 to 16 years, a single dose of unconjugated Vi capsular polysaccharide vaccine was 55-60% effective over 3 years in a randomized controlled trial. 25,26 A systemic review and metaanalysis of randomized controlled trials, including trials and/or populations in Africa, showed cumulative efficacy at 3 years for the Ty21a oral and the polysaccharide Vi vaccine were similar at 51% (95% CI 36%, 62%) and 55% (95% CI 30%, 70%), respectively. 27 Despite a 2008 WHO recommendation for programmatic use of existing typhoid vaccines in endemic countries, 28 no African country integrated these vaccines into routine schedules, largely due to the unsuitability in the youngest children or the need for repeated doses. Typhoid burden in Malawi, and elsewhere in Africa, is high among both school and pre-school age children, down to one year of age. 29 While these earlier typhoid vaccines were shown to be efficacious in school-age children, our study, importantly, demonstrates single dose Vi-TCV efficacy among African children <5 years of age, whose typhoid incidence was comparably high to that of school-age children. Vi-TCV efficacy was consistent throughout the study and ongoing typhoid surveillance of this cohort (until 36-42 months) will further assess the durability of protection and enable further age-stratified analyses in younger children. Analysis of a sub-study cohort is underway and will provide data on age-stratified immunogenicity.
Routine introduction of Vi-TCV among infants, coupled with catch-up campaigns to the age of 15 years, offers a strategy for typhoid control. 30 32 It is nonetheless reassuring that among 196 HIV-infected children identified and randomized in this study, no excess of SAEs was observed in the Vi-TCV group, and no SAEs were considered related to vaccination. In sub-Saharan Africa, HIV infection is, counterintuitively, epidemiologically associated with a 24-fold (95% CI: 9-100) reduction in diagnosis of blood culture-confirmed typhoid. 33,34 The continued surveillance in this trial, along with an ongoing substudy assessing the immunogenicity of a one or two dose schedule of Vi-TCV among HIV-exposed children at 9-and 15-month immunization visits, will provide additional information in this vulnerable population. MDR S. typhi remains prevalent in sub-Saharan Africa, 8