To the Editor: Reptiles are reservoirs of a wide variety of Salmonella serotypes, including all Salmonella enterica subspecies and S. bongori. In reptiles born in captivity or kept as pets, S. enterica subsp. enterica is frequently isolated (1). Salmonella strains are well adapted to reptiles, and they usually cause asymptomatic infections in such animals, while retaining pathogenicity for warm-blooded animals. For several years, reptiles have been recognized as a source of human salmonellosis. In North America, reptile-associated salmonellosis (RAS) has been reported, particularly in children, the elderly, or immunocompromised persons; severe and fatal infections are described occasionally (2). In contrast, only a limited amount of information on RAS is available in Europe. We report a case of RAS that occurred in an adult man in Italy.

A 32-year-old man had symptoms of enteritis. For 2 weeks, he had experienced intermittent watery diarrhea, mild fever, and abdominal pain. He was then treated with ciprofloxacin, and after 15 days of treatment, he recovered from enteritis. A stool sample, collected before treatment, underwent bacteriologic analysis, and Salmonella spp. were identified biochemically (api 20E, bioMérieux, Marcy l'Etoile, France) and by a polymerase chain reaction assay specific for the invA gene of Salmonella spp (3). Since the man was a reptile owner, RAS, rather than a foodborne infection, was initially suspected. He owned several cold-blooded animals; all had been tested for Salmonella spp. (at least 3 times at 2- to 3-week intervals), and results were negative. Three weeks before the onset of enteric symptoms, he acquired a boa (Boa imperator) that was subjected to routine analysis for Salmonella spp. in our laboratories (1). Salmonella spp. were isolated from a cloacal swab of the snake. Subsequently, both the human and reptile Salmonella isolates were characterized as S. enterica serovar Paratyphi B. In addition, both strains were found to be d-tartrate–fermenting (dT+) biovars (4), susceptible to ampicillin, amoxicillin-clavulanic acid, cephalothin, ceftazidime, gentamicin, streptomycin, chloramphenicol, tetracycline, neomycin, nalidixic acid, norfloxacin, and ciprofloxacin and resistant to sulfamethoxazole and co-trimoxazole.

By pulsed-field gel electrophoresis analysis of DNA, the strains displayed the same pattern, which suggests a clonal origin (4). The isolates were also assayed for virulence-associated genes. The SopE1 gene was detected in both isolates, and the avrA gene was not detected, which is consistent with an invasive pathovar of S. Paratyphi B (4). Conversely, the spvC, pef, and sef genes were not detected (5).

In recent years, a general increase in RAS detection has been observed, which may be the result of the increasing diffusion of reptiles as pets and a better awareness of RAS risk. In the United States, annual reports of RAS cases are published by the Centers for Diseases Control and Prevention (2). In Europe, studies on free-living and captive reptiles have shown a high prevalence of Salmonella spp (1). Nevertheless, national surveillance systems for RAS do not exist, and epidemiologic data are incomplete.

Notably, since Sweden became a member of the European Union in 1995, and the import restriction rules for reptiles were removed, a marked increase in RAS was observed in that country (6). As the deregulation of the trade in reptiles is applied, in agreement with the European Union laws, a similar scenario may be projected in other European countries. As is the case for nontyphoid salmonellosis, RAS may be underestimated, especially if patients are not hospitalized. Although a few cases of RAS have been previously reported in children in Italy (7,8), this report provides the first description of RAS in adults. S. Paratyphi B dT+, also known as S. enterica serovar Java, has been isolated in reptiles and tropical fish and has been associated with epidemics of human salmonellosis acquired from food, such as goat milk or chicken (9). The evidence shows that salmonellosis by S. Paratyphi B dT+ apparently occurs more frequently in adults (10), while so-called exotic reptile strains seem to be more prone to causing salmonellosis in children (7,8), which has led to the proposition that S. Paratyphi B dT+ strains may be highly pathogenic. By screening virulence-associated genes, both our isolates were found to be SopE1+ and avrA–, a pattern usually observed in the systemic pathovars of S. Paratyphi B (4) and associated with invasiveness, which suggests a high pathogenic potential. Accordingly, strict preventive sanitation measures should be adopted when handling reptiles (2), and reptiles should be always regarded as a potential source of pathogenic Salmonella strains for humans.