Salmonella enterica serovars Panama and Arechavaleta: Risk Factors for Invasive Non-Typhoidal Salmonella Disease in Guadeloupe, French West Indies

A retrospective study was conducted to identify the risk factors associated with Salmonella enterica bacteremia in infants and children in Guadeloupe, French West Indies. The 171 patients with S. enterica infection seen between 2010 and 2014 included 155 (90.6%) with acute gastroenteritis, of whom 42 (27.1%) had concomitant bacteremia, and 16 (9.4%) with primary bacteraemia. Most cases (97.7%) were in infants and children with no underlying health condition. Two subspecies were recovered: enterica (n=161, 94.2%) and houtenae (n=10, 5.8%). All but one (serovar Typhi) were non-typhoidal Salmonella . The most common serovars were Panama (n=57, 33.3% of isolates) and Arechavaleta (n=28, 16.4%). Univariate analysis showed a strong association only between age > 6 months and infection with the Panama or Arechavaleta serovar ( P = 0.002). The rate of resistance to all classes of antibiotics during the study period was low (< 15%); however, the detection of one extended-spectrum beta-lactamase-producing S. enterica strain highlights the need for continued monitoring of antimicrobial drug susceptibility. Infection with Panama ( P < 0.001) or Arechavaleta ( P < 0.001) serovar was significantly associated with bacteremia in a multivariariate analysis. These serovars are probably poorly adapted to humans or are more virulent. A delay between onset of symptoms and hospital admission > 5 days ( P = 0.01), vomiting ( P = 0.001) and increased respiratory rate ( P = 0.001) contributed independently to bacteremia in the multivariariate analysis. Thus, if non-typhoidal infection is suspected, blood should be cultured and antibiotic treatment initiated in all patients who meet these criteria.


Abstract
A retrospective study was conducted to identify the risk factors associated with Salmonella enterica bacteremia in infants and children in Guadeloupe, French West Indies.The 171 patients with S. enterica infection seen between 2010 and 2014 included 155 (90.6%) with acute gastroenteritis, of whom 42 (27.1%)had concomitant bacteremia, and 16 (9.4%)with primary bacteraemia.Most cases (97.7%) were in infants and children with no underlying health condition.Two subspecies were recovered: enterica (n=161, 94.2%) and houtenae (n=10, 5.8%).
All but one (serovar Typhi) were non-typhoidal Salmonella.The most common serovars were Panama (n=57, 33.3% of isolates) and Arechavaleta (n=28, 16.4%).Univariate analysis showed a strong association only between age > 6 months and infection with the Panama or Arechavaleta serovar (P = 0.002).The rate of resistance to all classes of antibiotics during the study period was low (< 15%); however, the detection of one extended-spectrum beta-lactamase-producing S. enterica strain highlights the need for continued monitoring of antimicrobial drug susceptibility.
Infection with Panama (P < 0.001) or Arechavaleta (P < 0.001) serovar was significantly associated with bacteremia in a multivariariate analysis.These serovars are probably poorly adapted to humans or are more virulent.A delay between onset of symptoms and hospital admission > 5 days (P = 0.01), vomiting (P = 0.001) and increased respiratory rate (P = 0.001) contributed independently to bacteremia in the multivariariate analysis.Thus, if non-typhoidal infection is suspected, blood should be cultured and antibiotic treatment initiated in all patients who meet these criteria.

Introduction
Salmonella enterica serotypes Typhi, Paratyphi A, Paratyphi B, and Paratyphi C are grouped as typhoidal Salmonella, and other serovars are described as non-typhoidal Salmonella (NTS).Typhoidal Salmonella are human host-restricted bacteria that cause typhoid fever, a systemic disease, and paratyphoid fever.NTS strains may be host-generalists, capable of infecting or colonizing a broad range of vertebrate animals species, or host-specialists, adapted or restricted to particular non-human animal species. 1 Although most salmonellosis due to NTS results in self-limited acute gastroenteritis, NTS have emerged as an important cause of bloodstream infection. 2In 2010, they were estimated to have caused approximately 3.4 million invasive infections and 681 000 deaths worldwide.Europe (102 cases/100 000 population) and the Americas (23 cases/100 000 population) had the second and third most cases of invasive NTS, respectively, but the numbers were substantially lower than that for Africa (227 cases/100 000 population).Young children, the elderly, malaria-infected and malnourished children, and immunocompromised people are at particular risk for invasive disease, which explains at least partially the discrepancies in incidence among regions. 3adeloupe, a French overseas territory located in the Caribbean, is a very high resource country according to the Human Development Index in 2013.Although few data are available on the epidemiology of Salmonella in humans in the Caribbean, it appears to be specific.In Martinique and Guadeloupe, S. enterica Panama was a major serovar, representing 35% of all isolates between 1990 and 1994 (first rank) 4 and 15% between 1992 and 1995 (second rank). 5rprisingly, this serovar has been rarely encountered in metropolitan France or in other regions of the world. 6At the University Hospital of Pointe-à-Pitre, around 35% (268/805) of the Salmonella isolates from humans belonged to serovars Panama and Arechavaleta during the period 2005-2014 (personal communication).The Panama serovar appears to have a propensity to cause bloodstream infection and severe human disease. 4,5,7 W therefore conducted a retrospective study to identify the risk factors associated with NTS bacteraemia in infants and children in Guadeloupe and to determine the pathogenicity of Salmonella serovars by analysing their contribution to Salmonella bacteremia.Antimicrobial susceptibility was also evaluated.

Microbiological analysis and serotyping
Strains were serotyped on the basis of somatic O and both phase 1 and phase 2 flagellar antigens by agglutination tests with antisera (Bio-Rad, Marnes-La-Coquette, France), as specified in the White-Kauffmann-Le Minor scheme. 8Antibiotic susceptibility to amoxicillin, amoxicillin-clavulanic acid, ticarcillin, cefalotin, cefoxitin, cefotaxime, ceftazidime, amikacin, tobramycin, gentamicin, nalidixic acid, ciprofloxacin, chloramphenicol, sulfonamides, cotrimoxazole, and tetracycline was determined by the disk diffusion method on Mueller-Hinton agar (Bio-Rad) according to the 2017 guidelines of the French Society for Microbiology/EUCAST (http://www.sfm-microbiologie.org).The minimum inibitory concentrations (MICs) for ceftriaxone and ceftazidime, and for ciprofloxacin for strains resistant to this antibiotic, were determined by the disk diffusion method on E-test strips (BioMerieux, Marcy L'Etoile, France).Salmonella strains that showed resistance to cefotaxime and/or ceftazidime were tested for the presence for extended-spectrum beta-lactamase (ESBL) by the double disk synergy method.
If more than one isolate with the same serotype and antimicrobial resistance phenotype was recovered from the same patient, only the first was retained for the analysis.

Statistical analysis
Children and infants were divided into those with bacteremia with or without gastroenteritis and those without bacteremia.Acute gastroenteritis was defined as diarrhea with a stool culture positive for Salmonella.Bacteremia was established when Salmonella was isolated from a blood culture.
Microsoft Access 2003 was used for data entry and Stata Version 10 for statistical analysis.In univariate analyses, the chi-squared test (or Fisher's exact test when appropriate) and Student's t test were used to compare categorical and continuous variables, respectively.Factors with P values < 0.20 in the univariate analysis were retained for the multivariate analysis.We considered P values < 0.05 to be significant.

Patient characteristics
A total of 171 patients with S. enterica infection were included retrospectively over the 36month study period.The sex ratio (M/F) was 0.94, and the median age was 18.3 months (mean, 30.3 months; 25th percentile, 6.4 months; 75th percentile, 43.0 months).No specific medical history was reported, except for sickle-cell anemia in three children (one SS homozygous, one AS heterozygote, and one S/β heterozygote) and a brain tumor in one.A total of 155 (90.6%) patients presented with acute gastroenteritis, of whom 42 (27.1%)had concomitant bacteremia and 16 (9.4%)primary bacteremia.Of the 171 cases, 151 (88.3%) were hospitalized: 56 (96.6%) with bacteremia and 95 (84.0%) without (P = 0.016).
The characteristics of the 171 patients are presented in Table 1.Five children were in septic shock, but no deaths were reported.
Only four patients were admitted to the emergency room with a probabilistic antibiotic (amoxicillin).Most of patients received antibiotic treatment on admission (86.0%, 147/171), consisting mainly of third-generation cephalosporins (76.6%, 131/171).Other commonly prescribed antibiotics were amoxicillin (n=8), amoxicillin-clavulanic (n=7), and cotrimoxazole (n=1).The treatment was adequate in all cases except one (third-generation cephalosporin for a strain resistant to this antibiotic).Antibiotic treatment was prescribed significantly more to patients with bacteremia than to those without (95.0%vs 81.4%, P < 0.016).

Risk factors for Salmonella bacteraemia
A delay between onset of symptoms and hospital admission > 5 days (P = 0.002), age > 6 months (P = 0.002), infection with Panama or Arechavaleta serovar (P < 0.001), vomiting (P = 0.005), and increased respiratory rate (P = 0.004) were significantly associated with bacteremia in the univariate analysis.A delay between onset of symptoms and hospital admission > 5 days, infection with Panama or Arechavaleta serovar, vomiting and increased respiratory rate were the main independent contributors to bacteremia in the multivariariate analysis (Table 1).
Univariate analysis showed that only age > 6 months was associated with infection with Panama or Arechavaleta serovar (P = 0.002).

Discussion
The serovar most often recovered in our study was Panama, as found in Martinique (38.5% of all strains investigated) and French Guiana (11.7%). 4,6Panama was also the major serovar in humans in Colombia and Chile. 9,10 his high prevalence is in contrast to that in other regions of the world, including mainland France (0.5% in 2011), 11 highlighting the specific epidemiology of these regions.In industrialized countries, NTS is transmitted predominantly through commercially produced food contaminated with animal feces. 1 The reservoir appears to be different for the Panama serovar, as suggested by the four recently described cases of Salmonella Panama meningitis in exclusively breastfed infants in French Guiana. 12In addition, no Salmonella Panama isolate was found among 275 Salmonella spp.isolated from 1636 samples from bovine, porcine, and avian food products and from poultry and the poultry environment between 2010 and 2014 at the Food Testing Laboratory at the Institut Pasteur in Guadeloupe (unpublished data).Contact with animals, such as reptiles, is an important non-food source of NTS infection. 13,14 n Guadeloupe, wild reptiles and amphibians (e.g.lizards, geckos, frogs) are commonly found in and around houses.Thus, a reptilian reservoir for the Panama serovar is plausible, as it is found in frogs, toads, turtles, lizards, and snakes, 6,15,16 and the host range might be much larger due to its presence in wild birds, swine, poultry, and Indian mongooses. 17,18,19,20,21 In ur study, Arechavaleta was the second most important serovar.To the best of our knowledge, Arechavaleta has been identified only in cane toads, dogs, and Indian mongooses. 21,22,23 Te major serovar 43:z4,z32:-(houtenae subspecies) in our study is also exotic.It was recovered in one case of osteomyelitis in a Taylor's cantil pit viper 24 but never in humans.The other serovars have commonly been associated with human infections.Further investigations are needed to identify the reservoir of the exotic Panama, Arechavaleta, and 43:z4,z32:-Salmonella serovars in Guadeloupe, and a reptilian source of contamination should be investigated.
Although we had a small sample, we detected a tangible trend of association between age > 6 months and the Panama and Arechavaleta serovars in our study.Guadeloupe has a high rate of breastfeeding (around 90% in 2013), 25 which protects neonates against infection through specific and nonspecific immune factors, 26 and exclusive breastfeeding also avoids exposure to contaminated water or food.
NTS Salmonella bacteremia usually occurs in 6-11% of children with gastroenteritis, 7,27,28 which is less than the 27% observed in our study.Several factors are known to be associated with an increased risk for invasive disease among children with Salmonella infection, including HIV infection, sickle cell disease SS, and specific serovars of Salmonella.HIV status was not systematically investigated in our study, but none of the parents spontaneously declared their child to be seropositive, and we assumed that HIV infection was low in our population.Only 4 (2.3%) of the 171 patients with NTS bacteremia had underlying disease, three with sickle-cell anemia, indicating that most of the cases of Salmonella bacteremia occurred in children with no underlying health condition.
Infection with Panama and Arechavaleta serovars was significantly associated with the occurrence of bacteremia.The Panama serovar was first isolated during an investigation of food poisoning among United States soldiers stationed in Panama in 1934, 29 and this serovar has repeatedly been described as causing invasive disease, such as bacteremia and meningitis, in children. 30,31,32,33 Tothe best of our knowledge, the Arechavaleta serovar has not been reported to cause severe infections.The way in which these two serovars cause invasive disease is unknown; however, both evolutionary theory and empirical comparisons predict that chronic pathogens such as Helicobacter pylori will become less virulent over time because of coevolution with their hosts. 34The disruption of coevolved hosts and Panama and Arechavaleta serovars might explain the severity of infection, as these serovars are probably poorly adapted to humans, as indicated by the fact that reptile-related salmonellosis has been associated with young age, a high rate of hospitalization, and invasive disease. 35,36 he greater virulence of these serovars might also be an explanation.Further studies are needed to clarify their pathogenicity.
Patients whose clinical symptoms persisted for > 5 days and who had an increased respiratory rate on admission were significantly more likely to have bacteremia, in agreement with previous reports. 37,38 he association between vomiting and bacteremia has not been described, but vomiting was found to be a predictive factor for NTS bacteremia in children < 5 years of age also infected with Plasmodium falciparum at a low parasite count. 39No deaths were reported in our study, indicating prompt, effective management.In contrast to the worldwide situation, 40 the rates of resistance to all classes of antibiotics were low, and there was no increase in the prevalence of resistance to first-line antibiotics during the study period, which accounts for the high rate of adequate antibiotic treatment.One therapeutic failure due to the only ESBLproducing Salmonella isolate was observed.Although the rate of ESBL-producing Salmonella isolates in our study was low, consistent with studies elsewhere in the world (0-2.4%), 8,41,42,43 itis important to continue monitoring antimicrobial susceptibility in Salmonella isolates from humans, as effective antimicrobial therapy reduces mortality and complications and shortens the illness.
In conclusion, the data reported here add to understanding of the epidemiology of Salmonella in the Caribbean.NTS bacteraemia should be recognized in healthy infants and children of all ages.Panama and Arechavaleta were the two serovars most often recovered in our study, with a propensity to cause bloodstream infection.If NTS infection is suspected, blood should be cultured and antibiotics initiated in all infants and children ill enough to be admitted to hospital with clinical symptoms for > 5 days, vomiting, or an increased respiratory rate.

Between
January 2010 and December 2014, 171 infants and children aged ≤ 15 years admitted to the emergency room at the University Hospital in Pointe-à-Pitre with Salmonella infection confirmed by stool and blood culture were included.A specific, standard anonymized medical questionnaire was completed to collect demographic, clinical, and biological data and information on associated pathologies.Empirical antibiotic therapy was considered appropriate if the treatment regimen included at least one antibiotic that was active in vitro against the infecting microorganisms.The study protocols were approved by the French Advisory Committee on Information Processing in Material Research in the Field of Health (CCTIRS 11-40).