Ascertaining the burden of invasive Salmonella disease in hospitalised febrile children aged under four years in Blantyre, Malawi

Abstract

Typhoid fever is endemic across sub-Saharan Africa. However, estimates of the burden of typhoid are undermined by insufficient blood volumes and lack of sensitivity of blood culture. Here, we aimed to address this limitation by exploiting pre-enrichment culture followed by PCR, alongside routine blood culture to improve typhoid case detection. We carried out a prospective diagnostic cohort study and enrolled children (aged 0-4 years) with non-specific febrile disease admitted to a tertiary hospital in Blantyre, Malawi from August 2014 to July 2016. Blood was collected for culture (BC) and real-time PCR after a pre-enrichment culture in tryptone soy broth and ox-bile. DNA was subjected to PCR for invA (Pan-Salmonella), staG (S. Typhi), and fliC (S. Typhimurium) genes. A positive PCR was defined as invA plus either staG or fliC (CT<29). IgM and IgG ELISA against four S. Typhi antigens was also performed. In total, 643 children (median age 1.3 years) with nonspecific febrile disease were enrolled; 31 (4.8%) were BC positive for Salmonella (n = 13 S. Typhi, n = 16 S. Typhimurium, and n = 2 S. Enteritidis). Pre-enrichment culture of blood followed by PCR identified a further 8 S. Typhi and 15 S. Typhimurium positive children. IgM and IgG titres to the S. Typhi antigen STY1498 (haemolysin) were significantly higher in children that were PCR positive but blood culture negative compared to febrile children with all other non-typhoid illnesses. The addition of pre-enrichment culture and PCR increased the case ascertainment of invasive Salmonella disease in children by 62-94%. These data support recent burden estimates that highlight the insensitivity of blood cultures and support the targeting of pre-school children for typhoid vaccine prevention in Africa. Blood culture with real-time PCR following pre-enrichment should be used to further refine estimates of vaccine effectiveness in typhoid vaccine trials.

Fig 1. Flowchart of participant recruitment.

Numbers of children screened, recruited and processed on PCR, blood culture and serology are provided.

Fig 2. Bimodal distribution of CT-values for primers invA, fliC and staG.

Graph shows the distribution of CT-values for the salmonella pan primer and S.Typhi and S.Typhiurium specific primers with a clear distinction between PCR positives and negatives.

Fig 3

CT cut-off set to demarcate non-specific amplification signals for primers invA (a), fliC (b) and staG (c). Graph shows CT-values from serial dilutions of contaminants; Micrococci (mc), Bacillus (bac), and pathogens; S. Typhi (sty), S. Typhimurium (stm), S. aureus (sta), Klebsiella (klb), and E. coli (eco).

Fig 4

Distribution of IgG antibody titres during acute infection (A, C, E, G) and in convalescence (B, D, F, H). IgG antibody titres were measured against four antigens (STY1498 in A & B, STY1479 in C & D, STY1886 in E & F, & Vi in G & H) across distinct groups of children. The groups are bc+PCR+ (cases that had S. Typhi infection confirmed by both blood culture and PCR (or blood culture alone), in acute infection [age 14 to 45, median 37 months], and in convalescence [age 14 to 45, median 33 months]), PCR+ (S. Typhi infection confirmed by PCR only, in cute infection [age 8 to 46, median 22 months] and in convalescence [age 22 to 46, median 40 months]), Negative (febrile but negative for typhoid on blood culture and PCR, in acute infection [age 1 to 48, median 13.5 months] and in convalescence [age 1 to 45, median 16 months]), and Control (afebrile healthy controls [age 0 to 52.5, median 10.2 months]).

Fig 5

Distribution of IgM antibody titres during acute infection (A, C, E, G) and in convalescence (B, D, F, H). IgM antibodies were measured against four antigens (STY1498 in A & B, STY1479 in C & D, STY1886 in E & F & Vi in G &H) across distinct groups of children. The groups are bc+PCR+ (cases that had S. Typhi infection confirmed by both blood culture and PCR (or blood culture alone), in acute infection [age 14 to 45, median 37 months], and in convalescence [age 14 to 45, median 33 months]), PCR+ (S. Typhi infection confirmed by PCR only, in cute infection [age 8 to 46, median 22 months] and in convalescence [age 22 to 46, median 40 months]), Negative (febrile but negative for typhoid on blood culture and PCR, in acute infection [age 1 to 48, median 13.5 months] and in convalescence [age 1 to 45, median 16 months]), and Control (afebrile healthy controls [age 0 to 52.5, median 10.2 months]).