Author + information
- Received October 30, 2015
- Revision received December 28, 2015
- Accepted January 21, 2016
- Published online August 1, 2016.
- Ayman A. Hussein, MD,
- Yacoub Baghdy, MD,
- Oussama M. Wazni, MD,
- Michael P. Brunner, MD,
- Ghazal Kabbach, MD,
- Mingyuan Shao, MS,
- Steven Gordon, MD,
- Walid I. Saliba, MD,
- Bruce L. Wilkoff, MD and
- Khaldoun G. Tarakji, MD, MPH∗ ()
- ↵∗Reprint requests and correspondence:
Dr. Khaldoun G. Tarakji, Cardiac Pacing and Electrophysiology, Department of Cardiovascular Medicine/J2-2, Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195.
Objectives This study reports a high-volume tertiary care center experience with the microbiology of cardiac implantable electronic devices (CIED) infections with assessment of temporal trends and profiles of late versus early infections.
Background The rates of CIED infections have been increasing. With changing demographics, patient and device characteristics, prophylactic measures, and the wide use of broad-spectrum antibiotics, there is need for updated contemporary data on the microbiology of CIED infections.
Methods The study included 816 consecutive patients with confirmed CIED infections who underwent transvenous lead extraction at our institution between the years 2000 and 2011. Blood cultures were obtained in addition of pocket swabs, pocket capsule, and leads.
Results Staphylococcal species remained the most common pathogens in CIED infections (68.4%), especially coagulase-negative species (37.6%). Methicillin-resistant staphylococci were the pathogens in 33.8% of all CIED infections and accounted for 49.4% of all staphylococcal infections. Gram-negative pathogens were identified in 8.9% of cases, whereas 13.2% were with negative cultures. CIED infections related to streptococci (2.5%), enterococci (4.2%), anaerobes (1.6%), fungi (0.9%), and mycobacteria species (0.2%) were less common. Of pocket infections, 49.5% occurred more than 1 year after pocket manipulation, and 53.6% of these were related to coagulase-negative staphylococci. In contrast, most endovascular infections were related to Staphylococcus aureus. The proportions of culture negative infections have increased (p < 0.0001).
Conclusions The study provides contemporary data on the microbiology of CIED infections. The rates of methicillin resistance seem to be greater than those reported from the preceding decade.
- cardiac implantable electronic devices
- transvenous lead extraction
The use of cardiac implantable electronic devices (CIED) has increased over the course of the past decade (1). In parallel, there has been an increase in CIED infections at a rate that seems to have followed a faster disproportionate trend to the rate of increase of newly implanted devices (2,3).
Despite increasing awareness of the seriousness of CIED infections, the institution of infection control practices, the administration of prophylactic antibiotics at the time of implants or system revisions, as well as improvement in CIED and lead design, CIED infections continue to occur and are life threatening (4,5).
Importantly, the demographics and risk factors of patients receiving CIED implants seem to have changed over time, which could explain the trends in CIED infection rates (4). CIED implant recipients are increasingly older and have multiple coexisting illnesses (6–8). Similarly, the implants of devices that are at higher risk of infection due to hardware burden or the inherent characteristics of their recipients, such as dual chamber pacemakers and defibrillators or cardiac resynchronization therapy devices, have increased over time (7,9). Importantly, a significant and increasing proportion of such devices are implanted in patients who are older than 70 or 80 years of age (10,11).
With changing demographic, patient, and device characteristics, the institution of measures to prevent CIED infections and the wide use of broad-spectrum antibiotics, it remains unknown whether there has been a parallel change in the epidemiology of micro-organisms in CIED infections over time. Similarly, although many infections are thought to be related to the index implant procedures or system revisions, a significant number of pocket or endovascular infections occur more than 1 year after device-related interventions, and it remains unknown whether there are microbiological differences in early versus late CIED infections.
This study reports a 12-year experience with the microbiology of CIED infections from a high-volume tertiary care center and assesses temporal trends of pathogens and the microbiological profiles of late versus early infections.
All 816 consecutive patients with confirmed CIED infections who underwent device and transvenous lead extraction or removal at the Cleveland Clinic between 2000 and 2011 were included. The clinical features, characteristics, and presentation of device infection were entered into a prospectively maintained data registry. All patients were evaluated and followed by an electrophysiologist and an infectious disease specialist from the infective endocarditis and cardiac device infection service. In our practice, we have established a multidisciplinary center for the management of CIED infections that includes, but is not limited to, cardiac electrophysiologists, infectious disease specialists, cardiac imaging specialists, radiologists, and cardiac surgeons.
The microbiological profiles and temporal trends were assessed in the overall population, which was then categorized into 2 groups based on the initial clinical presentation for the comparison of microbiology in early and late infections. The first group included patients who presented with signs and symptoms of device pocket infection with or without systemic symptoms. The second group included patients with endovascular infections who had systemic signs and symptoms of infection and a clinical history supported by microbiology and in most patients by echocardiographic imaging. In patients with clinical features of endovascular infection, transesophageal echocardiographs were obtained. In all patients, a clinical consensus was reached between the managing electrophysiologist and the infectious disease specialist regarding the need for device and lead extraction.
Blood cultures were obtained from all patients before the extraction procedures and before the initiation of antibiotic therapy at our institution. For patients who were referred from other institutions on antibiotic therapy, every effort was made to obtain all culture data from the referring institutions, and these were updated in our clinical records. At the time of the extraction procedure, device pocket swab cultures were sent when there was evidence of purulent drainage in the pocket. The fibrotic capsule was excised fully in all patients and tissue material was also cultured. Cultures were also obtained from all extracted lead tips and any attached fibrotic tissue.
For coagulase-negative staphylococcal species, which are well recognized as common microbiological specimen contaminants, repeated isolation or isolation from another source of the same coagulase-negative organism with an identical antibiotic susceptibility profile was required to be adjudicated as culprit organism. The study was approved by the Cleveland Clinic Institutional Review Board.
All statistical analyses were performed by using SAS version 9.2 (SAS Institute, Cary, North Carolina). Continuous variables are presented as mean ± SD or median (interquartile range), as appropriate. The chi-square test was used for comparison of proportions. For continuous variables, the Student t test or a nonparametric were used as appropriate. The Cochran-Armitage trend test was used to assess trends of culprit micro-organisms over the years. A 2-sided p < 0.05 was considered significant.
Between 2000 and 2011, 816 consecutive patients with confirmed CIED infections underwent device and transvenous lead extraction or removal at our institution and were included in the current study. Their clinical characteristics are summarized in Table 1. The mean age was 69.3 ± 15.0 years, most of them were men (73.6%), and the majority were Caucasian (88.3%). Their comorbid conditions included hypertension (53.1%), clinical heart failure (48.2%), coronary disease (53.1%), atrial fibrillation (44.3%), diabetes mellitus (31.9%), valvular heart disease (10.2%), chronic obstructive pulmonary disease (16.0%), end-stage renal disease (7.9%), and a prior history of stroke (10.2%). A history of prior coronary bypass surgery was present in 29.2%, and 10.0% had prior valve surgery. Only a minority had a history of prior endocarditis (3.3%).
Most patients had defibrillator leads in place (51.8%), and the remaining (48.2%) had pacemaker leads only. Of note, 20.5% of patients were pacemaker dependent, and 15.2% had coronary sinus leads. In all, the median number of leads in place was 2 (interquartile range, 2 to 3).
Pathogens were identified in the vast majority of patients (86.8%). The source, from which these micro-organisms were identified, were as follows: lead or lead material cultures (63.9%), blood cultures (54.5%), pocket tissue cultures (52.9%), and pocket swab cultures (44.2%). The remaining 13.2% of patients had no bacterial or other micro-organism growth from any of these cultures.
The distribution of pathogens in CIED infections is summarized in Figure 1. Staphylococcal species were identified in the majority of CIED infections in this cohort (68.4%). Of these, coagulase-negative staphylococci were more commonly observed (37.6%) than Staphylococcus aureus (30.8%). Methicillin-resistant staphylococci were the pathogens in 33.8% of CIED infections and accounted for 49.4% of all staphylococcal infections.
Gram-negative bacteria were identified in 8.9% of CIED infections in this cohort. The remaining infections were related to enterococci (4.2%; vancomycin sensitive 2.8%, vancomycin resistant 1.4%), streptococci (2.5%), anaerobes (1.6%), fungi (0.9%), and mycobacteria species (0.2%).
Trends in micro-organisms in CIED infections
The trends in micro-organisms in CIED infections are summarized in Figure 2. Between 2000 and 2011, the proportions of CIED infections related to coagulase-negative Staphylococcus, Staphylococcus aureus, or Enterococcus species did not seem to have changed over time (p = NS). There was a decreasing trend in the proportion of CIED infections related to methicillin-resistant coagulase-negative Staphylococcus (p < 0.0001) and methicillin-sensitive Staphylococcus aureus (p = 0.02). The proportions of CIED infections related to methicillin-resistant Staphylococcus aureus, methicillin-sensitive coagulase-negative Staphylococcus, enterococci, or other bacteria did not seem to have changed between 2000 and 2001 (p = NS). Importantly, the proportions of culture-negative CIED infections increased over the same time period (p < 0.0001).
Early versus late CIED infections
Of all CIED infections, 430 patients (52.6%) had pocket infections; the remaining patients had endovascular infections.
In the group of patients with CIED pocket infection, 50.5% had early infections defined as occurring within 1 year after device implantation or last pocket intervention. The remaining 49.5% of pocket infections occurred more than 1 year after device pocket manipulation. Coagulase-negative staphylococci were isolated in most early (40.0%) and late (53.6%) pocket infections, but were more likely to be observed in late pocket infections (Table 2). Staphylococcus aureus was more likely to be the identified pathogen in early infections (30.2% vs. 16.3%). In contrast, negative cultures were more likely in late pocket infections (23.9% vs. 16.7%). The staphylococcal resistance profiles were different between early and late pocket infections with methicillin-resistant species being more common in late infections (34.4% vs. 29.8%) and methicillin-sensitive staphylococci being more common in early pocket infections (40.5% vs. 35.4%) (Table 2).
In the group of patients with CIED endovascular infection (n = 386), 29.8% had early infections within 1 year after implantation or last pocket intervention. The remaining 70.2% had late CIED endovascular infections that occurred more than 1 year after pocket manipulation. In endovascular infections, Staphylococcus aureus accounted for most early (51.7%) and late (44.5%) infections. The overall distribution of organisms was not different in early or late CIED endovascular infections (p = NS) (Table 2). This was also true for staphylococcal resistance profiles (p = NS) (Table 2). Concomitant infections in endovascular infections were identified in 85 patients (22.0% of all endovascular infections). These included abscesses elsewhere (n = 39), osteomyelitis (n = 16), septic arthritis (n = 14), pneumonia (n = 7), cellulitis (n = 3), discitis (n = 2), meningitis (n = 2), urosepsis (n = 2), and tooth infection (n = 1).
The current study provides an updated contemporary epidemiology of the microbiology of CIED infections requiring extraction. It is the largest report to date on this topic and used multiple sources for microorganism cultures including lead or lead material, blood, pocket tissue, and pocket swab cultures. There were multiple observations with direct implications for clinical practice and patient management.
This topic is becoming increasingly relevant in clinical practice due to an increase in the number of CIED implants but, most important, due to increase in CIED infections (2,3). In fact, the longevity of patients with cardiac disease has increased and the number of system revisions or upgrades that a patient would require in a lifetime will increase in parallel. The risk of a CIED infection, a time-dependent variable, would likely follow a similar trend and the rates of CIED infections have indeed shown a disproportionate trend to increasing CIED implants (2,3). CIED infections carry not only a significant risk of morbidity but a risk of death up to 66% if left untreated, and this is decreased to about 18% with antibiotics and extraction (1,12–14). The eradication of infection requires complete removal of the devices and all lead material (1,12–14), with inherent risks to surgical or transvenous extractions.
In our cohort, staphylococcal species accounted for most CIED infections, which is consistent with previous reports (6,15–20), but the rates of coagulase-negative staphylococcal infections seemed to be more common in the European literature (21). Published data (19) from the 1990s (the decade preceding the period covered by this study) reported coagulase-negative staphylococci in 42% of CIED infection, methicillin-sensitive Staphylococcus aureus in 25%, and methicillin-resistant Staphylococcus aureus in 4%. Although the coagulase-negative rates were somewhat similar to the prior decade, an important observation in our study is that 15% of CIED infections were related to methicillin-resistant Staphylococcus aureus, an alarming rate compared with the prior decade. Despite the lack of a trend in methicillin resistance over the period covered by this study, the comparison to published data from the previous decade (19) suggests an increase in methicillin resistance. Furthermore, 1 in 3 CIED infections was caused by a methicillin-resistant staphylococcal organism, and one-half of all staphylococcal infections were methicillin resistant. Compared with the European literature (21) covering the same time period as the current study, the overall rates of methicillin resistance were significantly higher in the current report, but there seemed to be an increasing trend in the European report. Overall, these signals of increasing methicillin-resistant organisms may reflect the common inappropriate use of broad-spectrum antibiotics and suggest the acquisition of culprit organisms in health care environments in a significant proportion of patients, which has implications for empirical therapy. In general, culprit organisms in CIED infections may be acquired either from the patient's own skin or exogeneously from the health care environment. An association has been reported between pre-axillary flora and the pathogens isolated from CIED infections (17), which supports the theory of endogenous acquisition. On the other hand, low-level colonization with methicillin-resistant species has been reported in individuals with no recent health care contact or antibiotic exposure (22), but the disproportionate frequency of drug-resistant staphylococci in CIED infections suggests that the health care environment is the source of acquisition of these organisms (23,24).
Nonstaphylococcal infections such as anaerobes, Gram-negative bacilli, Candida, and nontuberculous mycobacteria account only for a minority of CIED infections (18,19,25–27). Clinicians need, however, to remain alert to the possibility of these infections due to their serious clinical course if not managed appropriately.
In this study, an assessment of the trends in culprit micro-organisms in CIED infections between 2000 and 2011 showed mostly absence of significant trends in the epidemiology of these pathogens. This suggests that the host factors, that is, changing epidemiology of patients receiving CIEDs or the type of devices being implanted (4,7–11), did not impact the epidemiology of culprit organisms, but the findings suggest increasing methicillin resistance compared with the prior decade, as discussed. Importantly, the proportions of culture-negative CIED infections seemed to have increased over time, which could reflect the wide and sometimes inappropriate use of antibiotics. This trend is consistent with rates of culture negative endocarditis of about 11% observed in Europe (21), which highlight the importance of obtaining culture material before the initiation of antibiotics, whenever possible.
Another observation in this study is that one-half of CIED pocket infections occurred more than 1 year after device implantation or last pocket intervention. Importantly, coagulase-negative staphylococci accounted for most early and late pocket infections but were more likely to be observed in late pocket infections, accounting for more than one-half of all late CIED pocket infections. It is very possible that these latent infections developed in an indolent fashion over time and were acquired at the time of pocket manipulation. In contrast, Staphylococcus aureus was more likely to be culprit in early than late CIED pocket infections, which highlights the more aggressive nature of this pathogen.
For endovascular CIED infections, most of those seemed to have occurred more than 1 year after implantation or pocket manipulation. In contrast to pocket infections, Staphylococcus aureus accounted for most early and late endovascular CIED infections, and it seemed that methicillin resistance was more common in endovascular versus pocket infections. This observation also highlights the likelihood of acquisition in health care environments. Although the original source of endovascular infections could not be identified with certainty in the current report, concomitant infections could be identified in 22% of all endovascular infections, suggesting that clinical investigation to identify concomitant infections is of importance to guide further management.
A large proportion of patients who undergo transvenous lead extraction for CIED infections at our institution are referred from other centers across the United States. As such, the denominator of all implants is difficult to estimate, and the study could not assess the specific incidence rates of infection. Only patients who underwent extraction of their devices were included in this registry, and the study does not include patients with CIED infections in whom extraction was not performed due to comorbidities or patient preference. Although material was sent for culture from multiple sources, the generators were generally not cultured and returned to the manufacturers for quality testing. The pocket capsules are generally excised in full and cultured in our practice, which minimizes the effect of absence of generator cultures. Another caveat is that the data included in this study included only infections of devices with leads as an endovascular component. The use of heart rhythm devices without endovascular components such as subcutaneous defibrillators or loop recorders is increasing in clinical practice, and the microbiology of infections of these devices merits investigation.
In a large population of patients with confirmed CIED infections undergoing lead extraction, staphylococcal species remain the most common pathogens in CIED infections, especially coagulase-negative species. One-third of CIED infections involve methicillin-resistant staphylococci, which are more likely to be acquired in health care environments. One-half of all pocket infections occur more than 1 year after pocket manipulation, and more than one-half of these late infections are related to coagulase-negative staphylococci and are likely to have been acquired at the time of the index pocket intervention. In contrast, most endovascular infections are related to Staphylococcus aureus. Over the course of 12 years, there did not seem to be a temporal trend in the epidemiology of culprit organisms, which suggests that the changing epidemiology of host factors did not affect the distribution of culprit pathogens over the years. However, the rates of methicillin resistance seemed to be higher than those reported in the preceding decade, which raises concerns regarding the wide use of broad-spectrum antibiotics and likelihood of acquisition in health care environments.
COMPETENCY IN MEDICAL KNOWLEDGE: The study provides contemporary micorobiological profiles of CIED infections showing primarily that staphylococcal species account for most of these infections and that the rates of culture-negative CIED infections have increased over time. The study also shows that the rates of methicillin resistance seemed to be higher than those reported in the preceding decade. The study provides important information for the clinical management of CIED infections and highlights the importance of proper use of antibiotics in CIED infections.
TRANSLATIONAL OUTLOOK: The findings have implications for clinical practice, primarily the use of initial empirical therapy in CIED infections, and highlight the importance of obtaining cultures before initiation of antibiotics in suspected CIED infections. The findings also suggest that a large proportion of CIED infections are from organisms that are acquired in health care environments and would indicate the importance of preventive measures.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviation and Acronym
- cardiac implantable electronic devices
- Received October 30, 2015.
- Revision received December 28, 2015.
- Accepted January 21, 2016.
- American College of Cardiology Foundation
- Baddour L.M.,
- Epstein A.E.,
- Erickson C.C.,
- et al.
- Tarakji K.G.,
- Wazni O.M.,
- Harb S.,
- Hsu A.,
- Saliba W.,
- Wilkoff B.L.
- Voigt A.,
- Shalaby A.,
- Saba S.
- Baddour L.M.,
- Bettmann M.A.,
- Bolger A.F.,
- et al.
- Camus C.,
- Leport C.,
- Raffi F.,
- Michelet C.,
- Cartier F.,
- Vilde J.L.
- Klug D.,
- Lacroix D.,
- Savoye C.,
- et al.
- Da Costa A.,
- Lelievre H.,
- Kirkorian G.,
- et al.
- Sohail M.R.,
- Uslan D.Z.,
- Khan A.H.,
- et al.
- Bongiorni M.G.,
- Tascini C.,
- Tagliaferri E.,
- et al.
- Kernodle D.S.,
- Barg N.L.,
- Kaiser A.B.
- Archer G.L.,
- Climo M.W.