OUP user menu

Disseminated malignancies masquerading as cardiovascular implantable electronic devices infections

Pipin Kojodjojo, Roy M. John, Laurence M. Epstein
DOI: http://dx.doi.org/10.1093/europace/eur040 821-824 First published online: 28 February 2011

Abstract

Aims Disruption of a previously well-healed cardiovascular implantable electronic device (CIED) pocket is usually presumed to be secondary to infection and current guidelines recommend the removal of the generator and all leads. We present our experience of CIED pocket disruptions thought to be due to infection and referred to our institution for lead extraction but instead proved to be the first manifestations of disseminated malignancies.

Methods and results Out of 1001 consecutive patients referred to our institution for transvenous lead extraction, two patients were found to have CIED pocket disruptions due to metastatic deposits. In both cases, subjects presented with increased swelling and discomfort of CIED pockets during device follow-up without any signs or symptoms of sepsis and negative cultures. The cause was presumed to be due to infection as it is not uncommon for CIED pocket infections to be diagnosed based purely on clinical appearances of the pocket and negative cultures. Both CIED systems were successfully extracted. At the time of surgery, the scar tissue within the pockets was atypical and histological specimens were sent for analysis. This confirmed diffuse large-cell lymphoma and metastatic lung adenocarcinoma as the causes of CIED pocket disruption.

Conclusions Malignancy should be considered as a differential diagnosis in CIED pocket disruptions, particularly those associated with negative cultures and histological analysis of tissue specimens removed from the pocket should be considered at the time of CIED extraction.

  • Cardiac implantable electronic devices
  • Lead extraction
  • Malignancy

Introduction

The predominant indication for device and lead extractions is infections.1 Of these, 49% of patients have only infections confined to the generator pocket without any clinically evident involvement of the transvenous portion of the lead system. Inflammatory changes over a previously well-healed cardiovascular implantable electronic device (CIED) pocket is usually presumed to be secondary to infection and current guidelines recommend the removal of the generator and all leads.2 The diagnosis is based on clinical assessment of the pocket and it is not uncommon for cultures to be negative due to localized nature of the infection. In this report, we present our experience with CIED pocket disruptions that were assumed to be secondary to infections but instead were the first manifestations of disseminated malignancies.

Methods and results

Between January 2000 and May 2010, 1001 consecutive patients were referred to our institution for transvenous lead extraction. Two unusual cases were identified in which implantable cardioverter-defibrillator (ICD) pocket disruptions were the first presentation of disseminated malignancies.

Case 1

A 60-year-old male with ischaemic cardiomyopathy presented to his local hospital in 2001 with ventricular tachycardia, 2 years following an inferior myocardial infarction. A secondary-prevention dual-chamber ICD (Gem III, Medtronic, Minneapolis, MN, USA) was implanted in the left subdeltoid region. The generator was replaced in 2007 (EnTrust, Medtronic, USA) for battery depletion.

In April 2010, he developed cellulitis of his right arm that resolved with oral antibiotics. One month later, he developed a diffuse swelling of the ICD generator pocket with redness and mild tenderness of the previously well-healed incision site, associated with left cervical lymphadenopathy. The latter was thought to be reactive to localized infection. Otherwise, he was afebrile without any systemic symptoms or signs of sepsis. His past medical history was notable for paroxysmal atrial fibrillation that had precipitated inappropriate ICD therapies and hyperlipidaemia. His routine laboratory tests were as follows: haemoglobin of 14.2 g/dL, white cell count of 7260/μL, platelets of 303 000/μL and normal serum biochemistry. Blood cultures were negative. No intracardiac vegetations were visible to transthoracic echocardiography. On ICD interrogation, pacing and sensing parameters were within normal limits with no ICD therapies delivered since his generator change in 2007. He was not pacing dependent.

A clinical diagnosis of ICD pocket infection was made and he was admitted for ICD system extraction under general anaesthesia. On visual inspection, the ICD pocket looked free from overt infection. In the process of dissecting the ICD lead free from the pocket capsule, a 5 ×4 cm area with friable, lipomatous tissue which accounted for the new prominence over the ICD pocket was found. Most of this tissue was resected and sent for pathological examination. The active fixation right atrial (5076, Medtronic, USA) and right ventricular (6949, Medtronic, USA) leads were extracted using locking stylets and a 16-French laser sheath (Spectranetics, Colorado Springs, CO, USA). The pocket was closed with nylon sutures and allowed to heal by secondary intention. Although blood, wound tissue and lead cultures were negative, empirical cephalosporins were commenced.

Upon pathological examination, the excised tissue was found to be largely necrotic with a dense lymphocytic infiltrate, composed of large cells with round to irregular nuclei, vesicular chromatin, prominent nucleoli, and scant amounts of cytoplasm (Figure 1). Immunohistochemical studies confirm these cells to be B cells and expressed CD20, CD30, BCL-2, and MUM-1. A whole-body positron emission tomography combined with non-contrast helical computed tomography (CT) imaging revealed a large left subpectoral centrally hypointense mass consistent with a large necrotic lymph node measuring 6.7 ×4.5 cm. In addition, there were multiple enlarged, contiguous lymph nodes along the left cervical chain but otherwise no suspicious intrathoracic, abdominal or pelvic lymphadenopathy (Figure 2). These findings were consistent with stage II diffuse large-cell lymphoma, which is generally associated with a good prognosis. Prior to discharge, a new dual-chamber ICD (Maximo II, Medtronic, USA) was re-implanted in the right subpectoral region. He was referred to the oncology service and received combination chemotherapy with good clinical response. However, 8 weeks after discharge, he did require further surgical debridement of the old left-sided ICD pocket due to poor wound healing at his local hospital.

Figure 1

Left panel—histological specimen from Case 1 showing dense lymphocytic infiltrate, composed of large cells with irregular nuclei, prominent nucleoli, and scant amounts of cytoplasm, consistent with diffuse large-cell lymphoma. Right panel—histological specimen from Case 2 showing moderately differentiated adenocarcinoma with characteristic glandular formations (both at ×40 magnification).

Figure 2

Positron emission tomography combined with computed tomography imaging revealed a large left subpectoral centrally hypointense mass (left panel) consistent with a large necrotic lymph node measuring 6.7 ×4.5 cm. In addition, there were multiple enlarged lymph nodes along the cervical chain (C5 level shown in the right panel).

Case 2

A 62-year-old male with a history of non-ischaemic cardiomyopathy with reduced ejection fraction of 25%, type II diabetes and hypertension, was successfully resuscitated from ventricular fibrillation in 1999. This led to the implantation of a secondary prevention left-sided single-chamber ICD (CPI Ventak Mini III, Boston Scientific, Natick, MA, USA) A few months later, a left lower lobe mass was found incidentally and subsequent biopsies confirmed the diagnosis of adenocarcinoma. No evidence of metastasis was found and therefore he underwent potentially curative left lower lobectomy.

During follow-up, pacing and sensing parameters were satisfactory. He had received several inappropriate shocks due to sinus tachycardia but was otherwise free from ventricular arrhythmias or pacing. Two years later, he presented to his usual cardiologist complaining of increased prominence of generator site and 5 kg weight loss. Routine blood tests were initially unremarkable. Chest radiograph revealed the presence of the single-chamber ICD. Apart from changes consistent with a previously left lower lobectomy, lung fields, and mediastinum were reported to be within normal limits. Over the next few weeks, the patient described progressive enlargement of generator site associated with worsening pain and erythema. There were no other localizing signs and symptoms of sepsis. At his referring hospital, blood cultures were negative but he was empirically started on oral cephalosporins. On arrival to our institution, he was afebrile. The site of the generator was erythematous, swollen with some sanguinous leakage, although the skin overlying remained grossly intact. White cell count was moderately elevated at 12 100/μL. A clinical diagnosis of a pocket infection was made.

The following day, the ICD pocket was opened in the operating room, the generator removed and the ICD lead (CPI Endotak, Boston Scientific, USA) extracted with the aid of locking stylets and 16-French laser sheath. Within the pocket, there was exuberant, thick, possibly chronically infected scar tissue, predominantly between the inner capsule of the pocket and the underlying pectoral muscle. Extensive debridement was performed to remove the entire mass of scar tissue which was sent for analysis. The pocket was closed with a Penrose drain in place with the aim of the pocket healing by secondary intention. He was transferred back to his referring hospital with the recommendations to continue antibiotic therapy for another 2 weeks before re-implantation of a new ICD on the right side.

Two days later, pocket tissue cultures were negative for bacterial growth but histology of the debrided scar tissue revealed moderately differentiated adenocarcinoma (Figure 1). Immunophenotyping suggests the adenocarcinoma to be consistent with being pulmonary in origin. Remarkably, his staging CT scan showed no other sites of metastases except for lymphadenopathy in the left subpectoral region (Figure 3). Two weeks later, he underwent implantation of a right-sided single-chamber ICD at his local hospital as life expectancy was estimated to be >1 year. However, despite commencing chemotherapy, he succumbed to rapidly progressive disseminated lung adenocarinona 8 months later.

Figure 3

Computed tomography of the chest showing significant lymphadenopathy overlying the left pectoralis muscle (white arrow). No other metastases were seen.

Discussion

Out of 1001 patients referred to a high-volume centre for lead extraction, we report 2 cases whereby ICD pocket disruptions were caused by metastatic deposits from disseminated malignancies, masquerading as pocket infections. Isolated case reports have described several cases of tumours such as breast carcinoma and melanoma with local invasion into the device pocket.3,4 Two cases of non-Hodgkin's lymphoma developing in pacemaker pockets have also been reported.5,6

Out of 1449 patients undergoing lead extraction in 13 US centres, 402 patients (27.7%) were undergoing intervention for pocket infection or erosions. Typically, these cases do not have bacteraemia or systemic signs of sepsis and have minimal inflammatory response such as leucocytosis or a raised C-reactive protein.7 The diagnosis is made clinically based on the appearance of the CIED generator site and findings such as purulent discharges, chronically draining sinuses, device erosions, and fat necrosis are highly confirmatory. Both cases described presented with swelling, erythema, and pain over their previously well-healed generator site. Positive tissue, pocket swab, and lead cultures would also support the diagnosis of pocket infection. In a previously reported cohort of 498 patients undergoing lead extraction at our institution, the indication for lead extraction was infection in 301 patients (60.3%) but wound cultures were only positive in 64.1% of infections.8 In keeping with our findings, in a study of 71 patients undergoing lead extraction at another high-volume unit, 69% of patients diagnosed with CIED infection had positive tissue or pocket swab cultures. However, 42% of patients (n= 15) without evident clinical infection also had positive cultures and of these patients, only one patient subsequently developed clinically evident pocket infection during follow-up.9 These observations may be affected by the empirical administration of antibiotics prior to transfer to tertiary lead extraction centres and the possibility of contamination when obtaining cultures. The diagnostic accuracy may be improved by cultures of lead fragments.10

It has been suggested that the presence of chronic low-level inflammation within the CIED pocket could promote seeding and proliferation of dormant neoplastic cells, a process termed inflammatory oncotaxis.4,6 Numerous case reports exist of metastasis from various tumour types developing in mechanically injured tissues such as sites of recent blunt trauma, around implants such as joint prostheses and percutaneous feeding tubes.1113 This concept is supported by experimental models showing that sites of induced tissue injury have an increased predisposition to seeding and implantation of circulating tumour cells.13 The link between tissue trauma and metastasis can occur via two postulated mechanisms. First, physical trauma could induce local inflammation that creates favourable conditions such as increased capillary permeability to attract metastatic cells from distant sites. Secondly, it is possible that micro-metastatic foci are already present at the site and subsequent tissue injury promotes proliferation and development of tumour cells into metastatic deposits. Although this concept is thought provoking, we also cannot exclude the possibility that the relationship between CIED implantation and metastases is coincidental.

Although uncommon, it would be prudent to consider malignancy as a differential diagnosis in all patients presenting with CIED pocket disruptions, particularly in patients with negative wound/lead cultures and an unusual appearance of the pocket tissue. Histological analysis of tissue specimens removed from the pocket should be considered at the time of CIED extraction.

Conflict of interest: none declared.

Funding

P.K. was funded by a British Heart Foundation Travel Fellowship (FS/09/047).

References

View Abstract