Case-control investigation of invasive Salmonella disease in Africa - comparison of human, animal and household environmental isolates find no evidence of environmental or animal reservoirs of invasive clades/strains

Background Invasive Salmonella infections are a major cause of morbidity and mortality in Sub-Saharan Africa (SSA), but the sources and transmission routes are uncertain. We investigated potential sources for cases of invasive disease by sampling healthy people, animals, and the environment in index-case and geographically-matched control households. Methods Sixty index cases of human invasive Salmonella infection were recruited (28 invasive Non-Typhoidal Salmonella (iNTS) disease, and 32 typhoid). Each index-case household was geographically matched to a control household. Extensive microbiological sampling included stool sampling from healthy household members, stool or rectal swabs from household-associated animals and boot-sock sampling of the household environment. Results 1203 samples were taken from 120 households, yielding 43 Salmonella isolates from 25 households (overall sample positivity 3.6%). Isolates from households were all NTS and spanned 15 STs. iNTS disease was caused by 3 STs of Salmonella Typhimurium, mainly ST313. Two S. Typhimurium isolates from index cases closely matched isolates from their respective asymptomatic household members (2 and 3 SNPs different respectively). There was no overlap of STs causing iNTS disease with environmental or animal isolates, despite recovery of diverse NTS. Conclusions The finding of NTS strains from index cases matching household members, coupled with lack of overlap with either animal or environmental isolates, supports a hypothesis that healthy humans are the source of iNTS infections in the household. The breadth of NTS strains found in the household environment across all sites demonstrated the robustness of sampling and methodology to detect NTS, and suggests a diverse ecology of Salmonella in this setting. The lack of S. Typhi isolated from the household environment may suggest a need for further methodological development to culture sources of typhoid.

sampling and methodology to detect NTS, and suggests a diverse ecology of Salmonella in 47 this setting. The lack of S. Typhi isolated from the household environment may suggest a 48 need for further methodological development to culture sources of typhoid. 49 50 51 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. ; https://doi.org/10.1101/2022.01.31.22270114 doi: medRxiv preprint Introduction 71 Salmonella infections are a substantial cause of morbidity and mortality in many low-and 72 phenomenon associated with narrower host range (i.e. becoming human restricted) and 121 increased invasiveness in a range of bacteria, including S. Typhi and S. Paratyphi A [36,37]. 122 Transcriptomic and phenotypic evidence also points towards specific niche adaptations to 123 human hosts among African isolates of NTS causing invasive disease [38,39]. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. October 2016 were approached as index cases. We sought written informed consent from 148 adults, and guardian consent for minors. Patients living outside the Blantyre district and those 149 with recurring iNTS disease were excluded. Index case households were sampled within a 150 maximum of 14 days following initial presentation to QECH. Following recruitment, the field 151 team visited the index cases in their households, where GPS co-ordinates were taken and a 152 household socio-demographic questionnaire was completed. Control households were then 153 selected by random bottle-spin and pacing 100 m from the index case household, recruited with 154 informed consent, and GPS and questionnaire data were also collected. Exclusion criteria for 155 control households were current treatment for invasive Salmonella disease for any family 156 member, or declining consent. In this event, the next-nearest house in the same direction was 157 selected. 158 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Environmental: The investigator's shoe was first covered with a waterproof sterile bag to 174 prevent cross-contamination, then a sterile bootsock was put over this, moistened with sterile 175 water, and single bootsocks were used to collect samples from each of the follow sites: pit-176 latrine, outdoor rubbish area, cooking and bedroom areas, and from the outside perimeter of 177 the household, using a standardised protocol for location and number of paces (10 paces), and 178 a step and twist motion. Bootsocks were then removed and placed into a sterile plastic bag for 179 transport. 180 All samples were transferred from the field to the laboratory in coolboxes, and processed on 181 the same day. 182 183 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint leaving 32 typhoid index case households, with 32 controls, and 28 iNTS index case 231 households, with 28 controls recruited for the study, giving a total of 120 households. All 232 households were in high-density low-income housing locations within Blantyre. 233 234 Isolation of Salmonella by culture: Table 1 shows sampling and positivity rates for Salmonella 235 by sample-type and by household. We collected a total of 1203 samples from the 120 index-236 case and control households (mean 10 samples per household), yielding 43 isolates (all NTS) 237 which were confirmed as Salmonella by WGS, giving an overall sample positivity of 3.6% 238 (Table 1). Overall, sample positivity was very similar across human, animal and bootsock 239 environment samples (human stool 16/491, 3.3%; animals 4/110, 3.6%; bootsocks 23/620, 240 3.8%), and across different household categories (Table 1). Animal sampling was from 110 241 animals in 71/120 (59%) households, and included both domesticated and non-domesticated 242 species (cows, chickens, pigs, dogs, cats, rats, doves, guinea pigs and geckos). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  Relatedness of Sequence Types from different sample categories: Figure 1 demonstrates the 254 wide diversity and relationships of Sequence Types (STs) isolated from different sample 255 categories (invasive disease, healthy human stool, animals, and the household environment). 256 Typhimurium ST313 and ST3257, that were linked to isolates from 2 healthy household 269 members in their respective households, who also had also ST313 and ST3257 isolated from 270 their stools ( Figure 1). Interestingly, although stool isolates from household members were 271 found equally among adults (n=7) and children (n=8), in both cases the household member 272 carrying a matching isolate was an adult. The genetic relatedness between index and human 273 household strains was explored at higher resolution by WGS phylogenetic analysis ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. ; https://doi.org/10.1101/2022.01.31.22270114 doi: medRxiv preprint all 3 sample-types. There was, however, non-co-localised overlap in some STs between 279 healthy humans and household-associated animals. A monophasic ST3262 was found in a 280 cat, a gecko, and a human, but these were all in different households. Similarly, ST3261 was 281 found in 2 healthy humans and a chicken, but these 3 events were also not from the same 282 households (Table 1). Significantly, ST3261 nor ST3262 are not known to have caused 283 human invasive disease in Blantyre. ST325 was found only in an animal (gecko). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. from a healthy human sample of the same ST from the same household. These isolates are indicated by black and red outlines. 296 298

Sequence types and WGS-inferred serovars from different sample categories: 299
The relationship between STs and WGS-inferred serovars for all isolates in the study is 300 documented in  Figure 1). 304 Serovars that were associated with healthy humans were Havana (7), Typhimurium (2), 305 Gaminara (2) is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. case-control studies have sampled adults and children in the same household, and household-371 associated animals, but this is the first study to also sample the household environment. This 372 yielded a similar culture positivity-rate to other sample-types, but also yielded a much wider 373 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. ; https://doi.org/10.1101/2022.01.31.22270114 doi: medRxiv preprint range of STs/serovars, indicating that Salmonella has a more diverse ecology in the 374 household environment than has previously been appreciated. The previously unappreciated 375 broad diversity of Salmonella in the household environment in this setting is in contrast to the 376 narrow range of STs causing human invasive disease, and the lack of overlap between these 377 categories is very striking. 378 379 Ours is also the most comprehensive study of its type, finding the greatest number of non-380 invasive disease isolates of any publication addressing this topic to date. We isolated a total 381 of 27 animal and environmental Salmonella, almost double the number isolated from a 382 comparable study [49]. Animal-ownership is not high in overcrowded urban areas, but we 383 nevertheless sampled 110 household-associated animals from 59% of households, yielding 3 384 STs from 4 animals of 3 species (1 chicken, 1 cat and 2 geckos). Bootsock sampling provided 385 53% of the non-invasive disease isolates in our study. This excellent yield from bootsocks for 386 detecting Salmonella in other contexts has been reported in other contexts [50], and we 387 recommend that similar studies carried out in the future also use this method. 388 389 Our methodology was thus able to resolve microbiological connections between people and 390 their household environments at both the individual household level and the community 391 level. The presence of some STs in multiple sample types was reassuring, indicating that we 392 were able to resolve a degree of connectivity between the different sample source types in 393 this setting, which might be expected as a result of faecal contamination of the household in a 394 setting with poor sanitation infrastructure. However, the dominant picture was a diversity of 395 non-invasive disease Salmonella, and unique distribution of different STs by sample-type. 6/9 396 STs found in the environment were found in neither animals nor healthy humans. Three STs 397 were only found in one sample type, and even when there were multiple isolates of a single 398 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint There was no definite broader discernible pattern of STs/serovars observed between different 408 categories of households. There was, however, higher household-positivity rates in typhoid-409 linked case and control household pairs, compared to iNTS-linked case and control 410 household pairs. Case and matched control houses were 100m apart, and this was chosen as a 411 distance that meant households were geographically close, but would not be likely to share 412 latrines or experience direct surface-water cross-contamination, since at that distance there 413 would usually be several intervening plots in this setting. Poor sanitation is known to be a 414 risk factor for typhoid because of the long-cycle transmission cycle, while the main 415 epidemiological risks described for iNTS disease are found in the host; malaria, malnutrition, 416 anaemia and HIV. It is possible that a higher rate of household contamination with NTS 417 serovars in case/control household pairs in proximity to typhoid cases reflects worse general 418 water and sanitation, and possibly higher poverty in that neighbourhood, compared to iNTS 419 case/control household pairs, but this would require further investigation. 420 421 While antimicrobial resistance is a major concern in our setting, only 12% of our household 422 isolates had one or more genes known to be associated with antimicrobial resistance, in 423 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted February 2, 2022. ; https://doi.org/10.1101/2022.01.31.22270114 doi: medRxiv preprint