In Lombardy, the largest region of Italy with approximately 10 million inhabitants, during 2006-2010 the number of cases of listeriosis reported by the laboratory based surveillance system increased 161% . The detection, identification and reporting procedures in use by the regional laboratories network remained substantially unchanged through the entire period. Consequently, an artifact of surveillance does not appear to explain this increase. Moreover, our findings are consistent with reports of listeriosis in Lombardy by the mandatory notification system of infectious diseases in the same period. Indeed, invasive listeriosis cases notified to the local health authorities increased from 35 in 2006 (incidence rate 0.37 per 100,000 population) to 70 in 2010 (incidence rate 0.74 per 100,000 population) . This increase mainly affected non-pregnant subjects, in accordance with previous reports from other European countries [6–11]. In Italy, in the same period the number of confirmed cases of listeriosis fluctuated between 51 in 2006 and 118 in 2008, but with an incidence rate always < 0.2 cases per 100,000 population [10, 20]. Moreover, in 2010 the overall European notification rate was 0.35 cases per 100,000 population with a 3.2% decrease compared with 2009 . The epidemiological trend in Lombardy appears to be counter to that of most European countries in during 2006-2010 .
In Lombardy, no common-source outbreaks were notified during the study period to the health authorities. However, the detection of a listeriosis outbreak may be very challenging. Indeed, many concurrent factors, such as the contamination of foods with long shelf lives, the long incubation period, and infrequent infections vs. presumably frequent exposures, may allow a listeriosis outbreak to occur as a succession of apparently unrelated cases .
The proportions of patients >65 years and without underlying medical conditions among non pregnancy associated cases were comparable to the rate reported in Europe in 2010 [8–10]. Age, socio-economic factors, differences in diet or food consumption habits and missed epidemic clusters might have confounded our observations. However, it is likely that some strains of L. monocytogenes with enhanced virulence may play a role as the etiologic agent in previously healthy persons . This supports the need to refine the field data collection tools for use in epidemiological investigations of listeriosis.
During the study period, the most common L. monocytogenes serotypes were 1/2a and 4b. However, unlike serotype 4b, which showed a steady annual prevalence, except for a peak in 2010, the number of serotype 1/2a isolates increased through the entire period. These results agree with previous reports indicating that serotype 1/2a is presumably replacing serotype 4b worldwide as the leading serotype causing human infections [20, 27–29].
In our experience, as many as 87% of listeriosis isolates were included in PFGE clusters, a proxy of presumably undetected single source outbreaks. These unrecognized events could have contributed to the increased incidence of listeriosis in Lombardy. A combined use of molecular subtyping and epidemiological investigations, as well as the application of space-time cluster analysis or Geographic Information Systems (GIS), could improve effectiveness of listeriosis outbreaks identification .
MLST provided information about distribution of MLST-defined clones in our country. CC1 (4b), CC2 (4b) and CC3 (1/2b), which have been described as prevalent worldwide, were also common in our region . Conversely, CC9 being ranked third in Europe , proved to be relatively infrequent among our isolates. In contrast, CC101 (1/2a) was overrepresented in our collection of isolates. In particular, 1/2a/ST38 isolates were recovered from 31 cases of listeriosis occurring through the entire study period, peaking in 2009 and 2010. ST38 is a very infrequent finding in the Listeria MLST database at the Pasteur Institute, France, where only three isolates are present from France, 1997, Australia, 2009 and Germany, 2011, respectively.
Moreover, a national database at the Istituto Zooprofilattico Sperimentale of Apulia and Basilicata, including >1000 isolates from human, food and animal source, contains only two isolates belonging to ST38. Of interest, the first one had been isolated in 2005 from a sample of “gorgonzola” cheese, mainly produced in Piedmont and Lombardy. Unfortunately, epidemiological investigations performed by the health authorities were unable to trace a food source. Detection of isolates with identical molecular types in food and clinical cases does not prove per se a causal association, unless consistent epidemiological data are available. However, our data strongly suggest the likely occurrence of a prolonged outbreak of foodborne listeriosis which could have been associated with the consumption of regional food products. The likelihood of a listeriosis outbreak lasting many years cannot be disregarded, because of the relationships of L. monocytogenes with food and food handling environments, such as its ability to survive in food processing plants and multiply over extended periods of time under harsh chemical and physical conditions . Recently, Knabel et al. have reported a 1/2a/CC8 strain as causing listeriosis throughout Canada for at least two decades, with most cases in elderly or immunosuppressed patients and not in pregnant women, as seen in our cluster of 1/2a/ST38 cases.
In a large proportion of our isolates, markers of ECs I to V have been detected. These ECs have been reported to cause multiple outbreaks worldwide and, more recently, to be associated with prevalent STs/CCs . In particular, ECII deserves attention because strains of this clonal group have been responsible for two large multistate outbreaks in 1998-1999 and in 2002 in the United States, but never in Europe . The present findings confirm our previous observations on a more limited sample of isolates from Lombardy and Tuscany regions . Moreover, ECIII has been identified only in two unrelated strains of serotype 1/2a, belonging to STs 11 and 14, respectively. ECIII has been previously associated with ST11 only . ECIV has been detected by MVLST in 11 ST2 isolates. An ECIV strain was previously reported from Italy as the causative agent of an outbreak of febrile gastroenteritis involving 1,566 immunocompetent subjects associated with consumption of a corn and tuna salad in 1997 in Piedmont [32, 33]. Eventually, the recently recognized ECV was also identified in 10 isolates into two clusters, comprising eight ST8 and two ST120 isolates, respectively . The frequent detection of EC markers is of concern, as these clones have been associated with nationwide and international outbreaks [20, 31]. It has been hypothesized that these epidemic clones have a greater potential for being the cause of foodborne events, perhaps because of their enhanced transmissibility, virulence, and/or persistence in food processing environments . The large proportion of isolates belonging to ECs could inherently confirm the poor sensitivity of traditional surveillance systems in detecting listeriosis outbreaks.
In our study, no specific molecular types could be conclusively associated with maternal-fetal cases, gender, age group or presence/absence of underlying conditions, except for the seven isolates 4b/CC6/ECII belonging to cluster 7, which were all recovered from non pregnant patients younger than 65 years, and the isolates grouped in cluster 11, belonging to serotype 1/2a/CC101, which were more prevalent in females and elderly persons and consistently associated with an underlying condition.
Our study has some limitations. Because of the restricted geographical source of L. monocytogenes strains, any generalization of the results is questionable, and an overestimate of clustering has to be considered. The Lombardy notification rate of listeriosis is higher than the overall Italian rate, as previously described. This difference could be explained by a higher sensitivity of the regional surveillance system, as well as by a more frequent exposure to hazardous food sources, such as some soft-ripened cheeses.