Bacterial Colonization of Implantable Cardiovascular Electronic Devices

Rohacek M, Weisser M, Kobza R, et al. Bacterial colonization and infection of electrophysiological cardiac devices detected with sonication and swab culture. Circulation 2010; 121:1691–1697. The implantation of electronic cardiovascular implantable devices for management of cardiac rhythm disturbances is a booming business. While the incidence of clinical infection of these devices, estimated at 0.7% to 1.6%, is low, the absolute number of infections is considerable given the large number of individuals at risk. Although infection may result from bacteremic seeding of the device, it is believed to most frequently occur at the time of implantation, although the onset of clinical manifestations may be remarkably delayed. This last observation indicates that device infection, which may involve leads (and, often, the endocardium), the generator pocket, or both, may be asymptomatic for long periods of time and, perhaps, indefinitely. In fact, culture of devices removed in the absence of suspicion of infection have yielded bacterial growth in approximately one-third of cases [1]. Rohacek and colleagues in Bern, Switzerland, have taken this observation a step further by performing not only swab cultures of removed devices, but also culture of sonicates, designed to allow sampling of disrupted biofilm, of the removed devices. Over a 14-month period, 121 devices were removed: 68 pacemakers and 53 implantable cardioverter/defibrillators (ICDs). Most were removed for battery replacement and only 6 were removed because of infection. A dose of antibiotic (usually cefazolin) was administered before the procedure. Sonicate cultures were incubated for 7 days while swab cultures were only incubated for 2 days (aerobic) or 5 days (anaerobic). Of the 115 devices removed for reasons other than suspected infection, sonication fluid was culture positive for 44 (38%), including 25% of pacemakers and 56% of ICDs, whereas conventional swab cultures were positive in 27% of the 112 devices examined by this method. Propionobacterium acnes and coagulase negative staphylococci were the most frequent isolates with both methods. The results of the 2 methods were concordant in 68%; this rose to 83% if thioglycolate enrichment cultures were excluded from consideration. In only 2 (3.6%) of the 56 instances in which either the sonicate or swab culture (or both) results were positive did a clinical infection subsequently develop during the median follow-up of 17.9 months—but the investigators do not indicate whether the infections were caused by the same organism that had previously been recovered. The incidence of clinically apparent infection was 4.5% among those with a positive sonicate culture overall, but was 9.5% in those in which the sonicate yielded >10 CFU/mL. The greater sensitivity of sonicate cultures compared to swab cultures is not unexpected. At least 2 additional factors in the current study could, however, have contributed to this difference. All patients received a dose of preoperative antibiotic which likely would be more effective against organisms recovered with swabs than those within biofilm. There was, in addition, a discrepancy in the lengths of incubation which would have favored organism detection with sonicate cultures. The use of increasingly sensitive techniques, such as culture of sonicates, inevitably lead to the greater likelihood of detection of microorganisms, as has been described with orthopedic device infections. The sensitivity of detection of organisms in association with such devices has been taken even further by use of polymerase chain reaction for amplification of bacterial DNA in sonicate fluid [2]. One problem with the use of increasingly sensitive techniques is the accompanying concern that organisms detected could, in some cases, represent contamination, rather than true infection. The results of an ever-increasing number of studies examining this issue, however, suggest that this is not true in the majority of cases. In the case of the cardiac devices examined in this study for which there was no clinical evidence of infection, the significance of this finding remains open to question since only a tiny proportion of those with positive cultures went on to develop clinically evident infection during the period of observation. This study provides strong evidence for the frequent presence of bacterial biofilm on cardiac devices in the absence of clinical infection. It also demonstrates, despite a mean duration of follow-up in excess of 1.5 years, the frequency of subsequent development of infection after exchange of the device at the same site is very low. It nonetheless provides impetus to the search for implantable devices of all sorts that are capable of preventing biofilm growth on their surfaces. The findings reviewed here also provide support for a recent statement of the American Heart Association with regard to the avoidance of microbiological studies at the time of removal of noninfected implantable electronic devices: Noninfectious device-related complications or device malfunction may occur that requires CIED removal with new device placement. At the time of device removal, specimens from the generator pocket or the explanted device should not be routinely sent for microbiological studies unless there are intraoperative findings to suggest concurrent infection. As evidenced in a previous investigation, intraoperative culture specimens frequently yield bacteria, although there is no other evidence of infection. In these cases, it is presumed that contamination during device explantation and processing of specimens likely account for the majority of positive cultures. In these cases, no antimicrobial therapy has been administered, and the rate of subsequent device infection has been no greater than expected. Thus, cultures should not be obtained routinely, because if positive, they could be misinterpreted as being clinically significant and lead to the inappropriate administration of antimicrobials, or worse, removal of the newly implanted device. [3, p 472]