Genomic epidemiology of Klebsiella pneumoniae neonatal and infant sepsis in Kenyan hospitals

Background Klebsiella pneumoniae (Kpn) is a leading cause of sepsis among hospitalised neonates and commonly causes outbreaks. Lack of surveillance data from low- and middle-income countries hampers development of vaccines and new measures to control infections. We describe the genomic epidemiology of Kpn neonatal infections admitted to Kenyan hospitals. Methods We analysed clinical and genome sequence data from neonates ([≤]28 days old) from a study of bacteraemia at admission at three Kenyan hospitals from October 2020 to April 2023 (NeoBAC), and, at one site, Kilifi, both neonates and infants (29 to 182 days old) from bacteraemia surveillance at admission and during admission from January 2001 to April 2023. Results We identified 75 neonatal cases of Kpn bacteraemia in NeoBAC, and 267 neonatal cases and 36 infant cases in the Kilifi surveillance. In NeoBAC, most neonatal infections were early onset and in-born whilst in the Kilifi surveillance most were late onset and out-born. Mortality was 32% (24/75) in NeoBAC, 41% (110/267) in Kilifi neonates and 56% (20/36) among Kilifi infants. Of 13 different STs identified in NeoBAC, ST6775 (29/75, 39%) and ST14 (26/75, 37%) were the most common. In Kilifi neonates 112 STs were identified, with ST17 (47/270, 17%) and ST14 (24/270, 9%) as the most common. Kilifi neonates had 61 KL- and 10 O-loci with predominant K-loci changing over the study period. Outbreak transmission clusters comprised 61/75 (81%) cases in NeoBAC and 140/270 (48%) cases in Kilifi neonates. Cumulative distribution suggested >30 K types would be needed in a vaccine to cover >85% of isolates. Conclusions Kpn bacteraemia occurred mainly in outbreaks with high mortality. The temporal dynamics of the target surface antigens may challenge future vaccine efficiency.