Molecular and Epidemiologic Analysis of Reemergent Salmonella enterica Serovar Napoli, Italy, 2011–2015

Human infections with Salmonella enterica serovar Napoli are uncommon in Europe. However, these infections represented 5.9% of salmonellosis cases in Italy during 2014–2015. The source of infection is unknown. We analyzed surveillance data and compared strain genetic similarities and found that contaminated vegetables and surface water are probable sources of human infection.

S almonella enterica serovar Napoli has attracted interest since 1982, when chocolate contaminated with this serovar and imported from Italy caused 2 foodborne outbreaks (1,2). During 2000-2006, a 140% increase in Salmonella Napoli cases was reported in Europe (3), most (87%) in France, Italy and Switzerland. Since 2004, the Rapid Alert System for Food and Feed (https://webgate.ec.europa. eu/rasff-window/portal/) has received 8 reports of Salmonella Napoli infection, which involved 6 countries and vegetables from Italy as the source of infection.
In Italy, sporadic cases of infection with Salmonella Napoli were reported before and after the first outbreak was reported in 1984 (4). The number of cases increased by 28% during 2000-2011 (5). Recently, Salmonella Napoli has caused waterborne (6) and foodborne (7) outbreaks. Although studies in other countries (3) and Italy (8)(9)(10) investigated the mode of infection transmission, the zoonotic reservoir is still unknown.
To update epidemiologic trends for Salmonella Napoli in Italy, we identified the population at greatest risk for infection and defined putative vehicles of infection, which could assist possible control measures. We also analyzed 2011-2015 surveillance data for human, animal, environmental, food, and feed Salmonella Napoli isolates; evaluated associated factors; and compared genetic profiles.

The Study
We extracted data on February 16, 2017, from the passive, voluntary, laboratory-based, surveillance system Enter-Net (http://www.iss.it/ente) for Salmonella spp. isolates from 2011-2015; patient identity was anonymous. Enter-Net is coordinated by the Istituto Superiore di Sanità (Rome, Italy) and contains information on strains from human and environmental samples. We excluded samples from nonhuman and nonenvironmental sources (n = 43), unspecified sample types (n = 21), and undefined Salmonella serovars (n = 1,066) from analysis. We obtained data (aggregated for 2011-2012 and 2013-2014) for Salmonella spp. isolated from food, animals, and feed by using the passive and voluntary veterinary surveillance system Enter-Vet (http://www. izsvenezie.it/temi/malattie-patogeni/salmonella/enter-vet/).
We evaluated potential epidemiologic risks associated with human and environmental Salmonella Napoli isolates and the 5 most frequent nontyphoidal Salmonella (NTS) serovars other than Napoli (monophasic variant of Typhimurium, Typhimurium, Derby, Enteritidis, and Infantis) by using simple 2-way frequency tables. Univariate analysis (p<0.05 by χ 2 test) was based on comparing categories for each variable and excluding missing values. We developed a multiple logistic regression model (Stata version 12; StataCorp LLC, College Station, TX, USA) to evaluate potential confounding among variables. We also used data for patients hospitalized with Salmonella Napoli infections or bacteremia as outcomes adjusted for other factors. The top 5 other NTS represented 73% of human Salmonella serovars isolated during 2011-2015. Results were consistent when we compared Salmonella Napoli with all reported NTS.
We serotyped Salmonella spp. isolates by using the White-Kauffmann-Le Minor scheme (11) and evaluated genetic relatedness among Salmonella Napoli isolates by using pulsed-field gel electrophoresis according to the PulseNet protocol (https://www.cdc.gov/pulsenet/ pathogens/protocols.html). We performed dendrogram and cluster analyses by using BioNumerics 7.5 (Applied Maths, Sint-Martens-Latem, Belgium). We scored similarity between chromosomal fingerprints by using the Dice coefficient and used the unweighted pair-group method with arithmetic means, 1.00% tolerance limit, 1.00% optimization, and coefficient of similarity >80% to define clonal relationships between strains.

Molecular and Epidemiologic Analysis of Reemergent Salmonella enterica Serovar Napoli, Italy, 2011-2015
The monophasic variant of Typhimurium (37.3%), Typhimurium (21.6%), Enteritidis (9.3%), Napoli (4.7%), Derby (3.1%), and Infantis (1.9%) were the most frequent of 331 Salmonella serovars identified for 21,486 human records. The proportion of Salmonella Napoli cases increased from 4.3% in 2011 to 5.8% in 2015 (p<0.001). Salmonella Napoli infection was more common in men, infants, children, and the elderly in northern and southern Italy during July-September than other reported serovars (Table 1). For Salmonella Napoli-infected patients with recent travel histories, those who traveled within Italy were more likely infected with this serovar (adjusted odds ratio 7.82, 95% CI 1.00-61.25; p = 0.050) than patients who traveled abroad. Although patients infected with Salmonella Napoli were hospitalized at rates similar to those infected with other serovars (Table 1), they had a greater risk for development of bacteremia (adjusted odds ratio 4.25, 95% CI 3.07-5.88; p<0.001).
Salmonella Napoli ranked fifth (3.7%) among environmental Salmonella spp. and showed a significant increase from 1.6% in 2011 to 2.9% in 2015 (p<0.005). A total of 85% of Napoli isolates were from surface water. Multivariate analysis (Table 2) confirmed that Napoli strains were more commonly isolated during April-September (p<0.001) and less commonly in northern and southern Italy than in central Italy.
Among Salmonella spp. animal isolates, the frequency of Salmonella Napoli increased from 0.7% in 2011 to 1.3% in 2015. A total of 50% of Napoli strains were obtained in the wild, most (70%) from wild boars in northern Italy. The low proportion (0.4%) from food remained stable during 2011-2015. Of 21 Napoli samples, 7 were from shellfish. One alfalfa sample contaminated with Napoli was reported in southern Italy in 2014.
We performed pulsed-field gel electrophoresis for 182 Salmonella Napoli strains isolated from different sources. We detected 121 unique profiles; 32 identical profiles contained 2-13 isolates. Despite high genetic variability among strains, we identified 4 main clusters that had genetic homologies >80% and matched the 3 areas in Italy: cluster A in northern Italy, cluster B in central Italy, and clusters C and D in southern Italy (p<0.0001) (Figure). Most human and environmental isolates from northern Italy belonged to cluster A, and most in central Italy belonged to cluster B (p<0.0001). In    Marked seasonality of cases matching environmental reports might be caused by increased irrigation and consumption of ready-to-eat vegetables during summer and fall, and more frequent human contact with contaminated surface water during summer and fall. We also reported frequent isolation of Salmonella Napoli from wild boars, which suggests a role for dissemination through the environment (12). Human and environmental strains of Salmonella Napoli belonging mainly to cluster A in northern Italy and cluster B in central Italy suggest direct infection. Different clusters for human and environmental strains from southern Italy might indicate wider spread in the environment.

Outbreaks of Salmonella
Heterogeneous sampling of nonhuman isolates through Italy was a limitation of this study. Travel history within Italy, rather than abroad, of patients infected with Salmonella Napoli increases public health concern in Italy. Uncontrolled spread of Salmonella Napoli through the environment and associated foodborne and waterborne outbreaks, the increasing proportion of young patients, and high risks for development of bacteremia make infections with Salmonella Napoli a serious public health concern in Italy.