Description and sources of contamination by Campylobacter spp. of river water destined for human consumption in Brittany, France.

Presence or absence of Campylobacter spp. in water of five rivers upstream from an intake point for drinking water production was investigated, and isolates genetically compared with human, pig and poultry isolates in order to determine their source. River water and drinking water obtained from these rivers were sampled one time per month, over a period of one year, and tested for Campylobacter. Isolates were typed by PFGE. Campylobacter was not detected in treated drinking water, but 50% of the river samples were contaminated. Contamination was observed on the four seasons. In total, 297 Campylobacter isolates were collected and generated 46 PFGE profiles. Campylobacter jejuni was the most frequently detected species in samples (74.1% of the isolates), followed by Campylobacter coli (17.8%) and Campylobacter lari (8.1%). Forty-two of the 46 PFGE profiles were unique. Only one genotype was detected three times in a river during the year and four genotypes in two different rivers. When compared to animal and human Campylobacter PFGE profiles, 14, 11 and one Campylobacter genotypes from water were genetically closed to human, poultry, and pig Campylobacter genotypes, respectively. The Campylobacter population displayed a high level of genetic diversity, suggesting that contamination originated from various origins. Human, poultry and pig were sources of contamination of the river by Campylobacter. Finally, no Campylobacter were detected in drinking water, indicating that the risk of outbreaks due to consumption of drinking water is low.

Comparison of genetic profiles of Campylobacter strains isolated from poultry, pig and Campylobacter human infections in Brittany, France.

Five hundred eighty-two Campylobacter isolates (177 from humans, 319 from poultry and 86 from pig) collected in Brittany, France, in 2003 and 2004 were typed by pulsed-field gel electrophoresis. The number of human cases increased during the hot season, particularly for C. jejuni. Twelve genetic groups out of 27 contained human isolates collected over the two years. These groups had 21.3 and 17.0% of the isolates obtained in 2003 and 2004, respectively. In four cases, isolates from 2003 have the same Pulsed-field gel electrophoresis (PFGE) profile as isolates from 2004. Six PFGE profiles common to poultry and human isolates were identified. Poultry isolates were found in 47 clusters containing human isolates. Caeca from farms and slaughterhouses accounted for 66% of these isolates, with chicken legs obtained from supermarkets accounting for the other 34%. Pig isolates never clustered with poultry and human isolates. In conclusion, the analysis of the genetic profiles of Campylobacter resulting from human cases showed that there were few identical or genetically close isolates between the human cases declared in 2003 and those declared in 2004. This highlighted a great genetic diversity in the isolates and indicated that it should be difficult to bind the human infections with groups of Campylobacter isolates presenting particular genetic profiles. The Campylobacter isolates obtained from the two animal production systems had different genotypes, and isolates from pigs differed genetically from isolates obtained from humans. We found that 44.6% of human Campylobacter isolates were genetically related to genotypes found in poultry and a part of these campylobacteriosis are due to contact with poultry. This is not particularly surprising in Brittany, a farming area with many animal-rearing farms and slaughterhouses. This work highlights the implication of the poultry in the French human cases and that handling of poultry is also an important risk for Campylobacter infection in humans.