Detecting DEC relies greatly on PCR techniques and many reports are available on procedures used to identify all categories of DEC [13–18, 26]. Most of these techniques however, are based on standard PCR. We have developed a novel multiplex real-time PCR which, with three reactions run parallel to one another, identifies the 5 major categories of DEC simultaneously. To our knowledge this is the first reported use of a TaqMan PCR, which provides both a convenient diagnostic tool and avoids the risk of cross contamination due to post amplification handling.
By rinsing all the colonies from the agar plate with saline rather than picking representative colonies as has been described in the literature, we were able to increase the sensitivity of the PCR. This method is quick, simple and inexpensive. The real-time PCR was able to detect 1,5 genome equivalents per assay.
As reported by the Robert Koch Institute, Berlin, Germany, of all reportable intestinal infections in 2006 in Germany, DEC, excepting EHEC, was the fifth largest group of pathogens after Norovirus, Rotavirus, Salmonella and Campylobacter spp. In our study reported here, the prevalence of EPEC (8,9%) and EAEC (4,8%) exceeded those for salmonella and campylobacter, each 2,4%. It is possible that the incidence of EPEC and EAEC throughout Germany is far higher than that reported. This discrepancy may very well be due to the low level of routine requests for DEC as we observed in the study period in this hospital. Unfortunately no data can be provided here to compare the incidence of DEC to that of viral pathogens since in this study stool specimens were not systematically tested for viral pathogens. However, in the paediatric population (the age group under 18 years in this study), many specimens were additionally tested for viruses and out of 40 EPEC positive samples additionally tested for viruses, 16 were additionally positive for rotavirus, norovirus or adenovirus. 4 specimens positive for EPEC were additionally positive for Salmonella spp. (1) or Campylobacter spp. (3). This leaves EPEC as the sole identified pathogen in 50% of all cases. EAEC was the only pathogen detected in 45% of the samples in this age group. In nine cases the specimen was positive for rotavirus, norovirus or adenovirus as well, whereas in three samples Campylobacter spp. was found as a co-infection and one specimen was tested positive for Salmonella spp. Even though this study provides no systematic data on viral testing we must assume that viral infections outnumber the cases of diarrhoea due to DEC in children. It is possible that more stool samples were submitted for viral detection only and thus the prevalence for EPEC and EAEC found in this study may still be an underestimation in this age group.
A strong association between EPEC and diarrhoea in children has been reported , yet we also found a relatively high prevalence for EPEC in older patients.
A large proportion of EPEC (and EAEC) infections may well be associated with travel, since a high number of positive specimens within the group between 18 and 65 years were obtained from the department for tropical diseases. Concerning the association of atypical EPEC with diarrhoea findings are contradictory [27–29], but it continues to be the most prevalent pathotype of EPEC found in industrialised countries [27, 29, 30], which we also noted in our study. A subset of the heterogeneous group of atypical EPEC has been found in patients with bloody diarrhoea, (in some cases leading to HUS). They resembled EHEC according to their serotype, virulence profile and multilocus sequence types, leading to the assumption that these strains might be EHEC that have at some stage lost the shiga toxin gene during infection . Unusual in this study is the detection of EAF without the eae gene as most often eae is found alone (atypical EPEC) or with EAF. It is questionable whether strains positive for EAF but not for eae should be classified as EPEC and whether they actually are pathogenic. Further investigation is required as to whether this is due to a sequence variation within the intimin gene resulting in a failure of amplification or whether perhaps other bacterial strains have acquired the EAF plasmid.
One limitation of this study is the inability of the method to distinguish between those patients with a double infection with two or more DEC and those patients who are infected with a single bacterium carrying more than one virulence gene. Because we use a pool of bacteria from an overnight culture, it is not possible to trace the result back to a single bacterial clone. In this study we found 17 specimens which were positive for both EPEC and EAEC. In no case was the distinction between a single bacterium carrying both factors or a double infection clinically relevant. Nevertheless, it would be interesting to determine the frequency of multiple gene carriage by a single clone and we are currently prospectively attempting to trace back to the gene-carrying clone(s).