Extraction of Infected Pacemaker/ICD Leads

Abstract & Commentary

By John P. DiMarco, MD, PhD

Source: Grammes JA, et al. Percutaneous pacemaker and implantable cardioverter-defibrillator lead extraction in 100 patients with intracardiac vegetations defined by transesophageal echocardiogram. J Am Coll Cardiol. 2010;55:886-894.

The number of patients with implantable pacemakers and implantable defibrillators has increased markedly in recent years. Unfortunately, cardiac rhythm device (CRM)-related infections have also increased, and management of these infections is often problematic.

In this paper, Grammes et al from Hahnemann University Hospital in Philadelphia describe their experience with device-related infections that were associated with intracardiac vegetations. The authors maintained a longitudinal database of all patients undergoing pacemaker or ICD lead extraction at their institution. They identified patients in the database who had device-related endocarditis, as defined by the modified Duke criteria. In most patients, preoperative transesophageal echocardiography (TEE) was used to risk stratify patients. Intracardiac vegetations were defined as discrete echogenic masses found by transthoracic or transesophageal echocardiography on a valve, lead, or endocardial surface. Lead extractions were performed either with manual traction or with powered electrosurgical or Excimer laser extraction sheaths. Occasional patients required a femoral snare or basket extraction technique for complete lead removal. When necessary, patients were maintained with a temporary pacing system with subsequent device reimplantation based on sterility of blood cultures and resolution of vegetations.

Over a 16-year period, 984 patients underwent CRM device extraction; 1,838 leads were removed. In 480 patients (49%), the reason for extraction was systemic or localized infection. Of these 480 patients, 100 patients (21%) had echocardiographic evidence of intracardiac vegetations. The size of the vegetations ranged from 0.2 to 0.4 cm, with a mean diameter of 1.6 cm. In these 100 patients with vegetations, 215 leads were removed using standard percutaneous techniques. The most common infectious organisms were methicillin-resistant Staphylococcal aureus and methicillin-sensitive Staphylococcal aureus. Other organisms identified included coagulase-negative Staphylococcal species, Enterococcus faecalis, Streptococcus, vancomycin-resistant Enterococcus, Citrobacter, and Candida species. In 16 patients, blood cultures were negative despite clinical evidence of infection and echocardiographic evidence of vegetations. In some, this may have been due to prior antiarrhythmic therapy. The median implant duration was 32.5 months, with a range of 1 to 300 months. The median time for extraction of the first lead was three minutes, but the range was broad, from 1 to 187 minutes. Extraction times were similar to those in a reference population without endocarditis. In 54 patients, new CRM devices were reimplanted during the hospitalization, with a median time to reimplantation of seven days. Forty-six patients did not receive implants during the hospitalization. In 18 of these 46 patients, the devices were not reimplanted because of persistent systemic infection, despite complete device removal. Among the 54 patients who underwent device reimplantation during the original hospitalization, 51 were discharged in stable condition. None of these 51 patients had relapsing infections that required a second extraction.

Long-term follow-up was available in only 71 patients, since many patients were followed at other institutions. During follow-up, 19 of these patients died (27%). Ten patients died during the initial hospitalization and seven after discharge. Persistent septicemia was the cause of death in 11 patients. Seven patients died of unknown causes, without evidence for recurrent infection, and one patient had sudden death. Five patients had complications during lead extraction. These complications included embolization of the vegetation to the lungs in two patients, embolization of a lead fragment, severe tricuspid regurgitation with a posterior flail tricuspid leaflet, and prolonged hypotension in single patients. All of these five patients recovered and did well during follow-up.

The authors argue that percutaneous lead extraction is possible in patients with endocarditis and vegetations. Mortality is primarily related to persistent sepsis and multi-organ failure rather than complications of the lead extraction itself. The data from this series compare favorably with other results in the medical and surgical literature that included patients with and without vegetations.


Device-related infections are increasing in frequency as the number of implants increase and patients are able to live longer with these devices. The early infection risk after a CRM device implant is about 1%-1.5%. Over time, chronic leads may become infected in the setting of sepsis from another primary source, and devices may erode through the skin if the pocket was not created properly. The pocket infection rate with generator changes or device upgrades is higher than that seen with initial implants.

Successful treatment of a CRM device-related infection usually requires complete removal of all infected hardware. Up until about six months after implant, virtually all new leads can be removed with just manual traction, and special tools and skills are not usually required. After six months, the development of fibrosis may fix the lead to other leads, the blood vessel wall, or the endocardium, and manual traction may not be effective. If excess force is applied, the lead itself may break down, making eventual removal even more difficult. Removal of these leads often requires either special transvenous sheaths to break up the fibrotic adhesions or direct, open-chest surgical removal. Some authors have argued in the presence of visible vegetations on the lead, direct surgical extraction should be preferred to minimize the risk of embolization of the infected material.

In this paper, Grammes et al, from an experienced extraction referral center, show that most infected leads with vegetations may be safely removed with transvenous approaches. Complications directly related to the vegetations were uncommon. Although some patients died with persistent sepsis and multi-organ failure, these conditions were present before the extraction. Are there vegetations that are too big to remove transvenously? In this series, vegetations up to 4 cm in diameter were seen. Larger vegetations are rare but may be seen in some cases, especially with fungal endocarditis. Surgical management of such very large vegetations may still be required.

My current approach is similar to the one described by the authors. If there is a hemodynamic reason for a direct operation to replace an infected valve, the leads may be removed surgically at the same time. In most other cases, a transvenous extraction approach in a center with experienced operators is likely to be a better option.