Positive cultures in asymptomatic patients during elective device replacement: a murderer hides in the darkness or an innocent person on the crime scene?

The paper of Kleeman et al . 1 is a prospective observational study on bacterial colonization of implantable cardioverter defibrillator (ICD) systems and its relevance to device infection. Out of 122 patients, without clinical signs of device infection, who underwent ICD replacement or lead revision, 33% had positive cultures. After a median follow-up of 203 days, seven patients developed a device infection (five patients with and two without prior bacterial colonization). The authors conclude that asymptomatic bacterial colonization was a common problem in ICD recipients and became clinically relevant in 7.5% after replacement/revision. This paper presents an interesting and clinically relevant study and addresses the concept of silent colonization of devices or prosthetic valves and the importance of latent infection. Latent infection is a reality. In intravascular-catheter-related infections, colonization of catheters does not induce systemic infection. However, there is a correlation between the number of micro-organisms retrieved by semiquantitative or quantitative cultures from a catheter surface and the risk of associated infection. 2 Latent infective endocarditis could be suspected by the presence of an unexpected inflammatory infiltrate in 0.9% of heart valves retrieved for other reasons than endocarditis. 3 Several well-done studies found asymptomatic bacterial colonization of pacemaker. 4 , 5 Several hypotheses could explain an equilibrium between the human host and bacteria. 6 When this balance is broken, bacteria are destroyed or infection occurs. Many factors may influence this balance ( Figure 1 ), such as the number of bacteria or the addition of a new infection, the virulence of bacteria and their ability to adapt to the unfavourable environment, and the defence capacity of the host. We have some arguments for such equilibrium in device infection. Infection is mainly due to low virulent bacteria originating from normal skin microflora. Staphylococci have some capacity of sequestration and antibody absorbing to survive in hostile conditions and some small-colony variants of Staphylococcus aureus and of coagulase-negative Staphylococci have been involved in pacemaker infections 7 as well as in recurrent infection in bone, heart valves, lung, and soft tissues. These small-colony variants grow slowly, have been found to persist inside of cultured cells (intracellular location), and can grow in anaerobic condition. Hypothesis for a balance between the human host and bacteria. Recent studies have shown the importance of genes involved in adherence and biofilm formation in pacemaker-related infection. We can speculate that some micro-organisms in certain unfavourable conditions do not induce infection that occur only after phenotypic modifications or/and changes in host conditions sometime long after the last procedure on the pacemaker site. This hypothesis explains observations of very late device infection with bacteria from the normal skin flora several years after the last procedure on the device implantation site and without evidence of haematogenous contamination. At our centre, we observed three cases of patients who developed haematoma of their pacemaker pocket without any wound, several years after the implantation due to anticoagulation in one case, rupture on the pacemaker pocket due to explosion in another case, and a car crash in the third case. A few weeks after complete resolution of the haematoma, local infection with S taphylococcus epidermidis occurred. These cases are perfect examples of changes in host condition with haematoma and the occurrence of favourable condition for bacterial growth. Re-interventions after asymptomatic bacterial colonization induce blood effusion inside the pocket, local inflammation, and sometimes contamination with new colonies increasing the bacterial charge and adding some new bacterial species that could interact and displace the equilibrium, therefore favouring infection. Secondary procedure such as pulse generator change was a risk factor of infection 8–10 and early re-intervention for haematoma or lead dislodgement was the first risk factor of infection with and an odds ratio of 15.04. Ite missa est ? No, these results raise several questions. The first one is, how is it possible to have more than one-third 4 , 5 of positive cultures from the pacemaker pocket after intervention presumably performed in strict aseptic conditions? These observations emphasize the need for very strict conditions for implantation. Perhaps some of these positive cultures could have been false positives due to contamination. There are many criticisms about swab techniques, 5 , 11 and we have to note that in the paper of Kleeman et al . when some other cultures have been studied they were negative even in patients with a positive swab culture. In the paper by Dy Chua et al ., 5 some patients had tissue cultures and swab cultures and only 8% of asymptomatic patients had both cultures positive. Pichlmaier et al . have extracted bacterial DNA from tissue biopsy and from the device in 47.2% of 108 asymptomatic patients undergoing elective device replacement. DNA was extracted from both in only 10.2% and the strain found was different from those commonly seen in series of pacemaker-related infections and different from the series of swab cultures in asymptomatic patients (a majority of pseudomonas,16%, and only 11% of Staphylococcus ). The second question is about the management of these patients. Antibiotic therapy in asymptomatic patients is probably not reasonable. Some bacteria of the skin flora like Propionibacterium or Trichosporum could be contamination and, if not, they have a very low virulence and are associated with an extremely low rate of clinical infection. Positive cultures with such bacteria should not induce an antibiotic treatment in asymptomatic patients. However, in the study of Kleeman et al ., one patient with S. aureus and 11.5% of patients with positive culture with S. epidermidis have developed infection with the same bacteria during follow-up. Further studies are warranted to evaluate the efficacy of antibiotic therapy in the case of positive culture with Staphylococcus in asymptomatic patients. In conclusion, device contamination by bacteria may not always induce an immediate infection. We can hypothesize that, in many cases, an equilibrium will be created between the host and the bacteria and infection will occur only after destabilization. The prevalence of such contamination is unclear, but previous studies have found positive cultures in more than one-third of asymptomatic patients during re-intervention. If local contamination is indeed as high as one-third of patients, we can certainly not conclude that a positive swab culture may ascertain a device infection. These results trigger many questions about our ability to reduce the contaminations and about its management. Conflict of interest: none declared.