What's new?

Lead‑related infective endocarditis (LRIE) is the most dangerous infection of a cardiac implantable electronic device and one that is associated with high mortality rates. It requires prompt detection and management. Diagnosing LRIE is often difficult due to nonspecific symptoms, and early detection is important, so that the patient can be sooner qualified for transvenous lead extraction and begin adequate antibiotic therapy. This study is the first to assess the impact of expanded diagnostic criteria on the diagnosis and prognosis of patients with LRIE. The study included a large patient group and prolonged follow‑up. A comparison of clinical data and risk factors for death in patients with definite and possible LRIE showed that detecting the infection at an earlier stage through the use of extended diagnostic criteria may improve the long‑term survival of patients receiving appropriate treatment.

Introduction

Infections in patients with cardiac implantable electronic devices (CIEDs) are a serious complication associated with an unfavorable prognosis. The prevalence of infectious complications showed by retrospective studies was 0.5%–4.4%,^1,2 while current prospective studies put it at 0.3%–2.3%.^3-5 The most dangerous type of infection in people with CIED is lead‑related infective endocarditis (LRIE). LRIE may occur with or without pocket infection, lead, or tricuspid valve‑related vegetations. Diagnosing LRIE, especially without pocket infection, often poses a challenge due to uncharacteristic symptoms that make it difficult to quickly introduce appropriate treatment. Diagnostic problems have contributed to a search for new LRIE‑specific criteria. Subsequent guidelines of the European Society of Cardiology (ESC)^6,7 and the European Heart Rhythm Association (EHRA)⁸ suggested expanding the Duke criteria to include local pocket infection, septic pulmonary embolism (SPE), and positive lead cultures, as well as results of new imaging testing (¹⁸F‑fluorodeoxyglucose‑positron emission tomography / computed tomography [¹⁸F‑FDG‑PET/CT], white blood cell single‑photon emission CT / CT [WBC SPECT/CT]). Moreover, in the ESC guidelines^6,7 infective endocarditis was divided into definite and possible. The introduction of the term “possible LRIE” should facilitate detection of this insidious disease and increase referrals for transvenous lead extraction (TLE). So far, the literature does not include any analyses of patients diagnosed with possible infective endocarditis. This study’s aim was to compare diagnostics, treatment, and long‑term survival in patients undergoing TLE due to definite and possible LRIE.

Patients and methods

Results

Ultimately, 838 patients with LRIE were analyzed, including 622 individuals with definitive LRIE diagnosis and 216 with possible LRIE. The analysis of clinical data for patients with definite and possible LRIE showed that renal failure was more frequent in the definite LRIE group, and antiplatelet therapy was used more often in these patients. Vegetations and positive blood cultures, especially of Staphylococcus aureus, were also more common in the patients with definite LRIE. The patients with possible LRIE had higher prevalence of pocket infection (64.81% vs 51.29%; P <⁠0.001). A comparative analysis of minor Duke criteria showed higher prevalence of SPE, fever, and shivers in the patients with definite LRIE. Similarly, all laboratory inflammatory parameters showed higher values in these patients. The first symptoms of LRIE were most oftenly nonspecific and included recurrent fevers of initially unknown etiology and recurrent pneumonia—a manifestation of SPE. For this reason, the time from the onset of the first symptoms to diagnosis was long, that is, on average 3 months for definite LRIE and 2 months for possible LRIE. A detailed comparative analysis of the time from the onset of the first symptoms to diagnosis of definite and possible LRIE is presented in the Supplementary material, Table S1.

In some patients, diagnostic difficulties are also indicated by the number of hospitalizations related to atypical manifestations (fever of unknown origin, recurrent pneumonia, anemia). Our analysis showed that more than 1 hospitalization was recorded in as many as 232 patients (37.3%) with definite LRIE and 64 patients (29.63%) with possible LRIE, and 3 or more hospitalizations in 76 individuals (12.06%) with definite LRIE and 18 cases (8.33%) with possible LRIE. Also, the number of hospitalizations due to LRIE symptoms (per 1 patient) in retrospective assessment was 1.6 in the definite LRIE group and 1.38 in the possible LRIE group. The median (IQR) in both groups was 1 (1–2), but the analysis demonstrated that in 25% of patients with definitive and possible LRIE (155 patients with definitive LRIE and 54 pateints with possible LRIE) it took, respectively, at least 6 and 5 months from the onset of the first symptoms of LRIE to TLE (Table 1).

Intracardiac abrasion of the lead was observed more often in the patients with definite LRIE (29.42% vs 18.52%; P <⁠0.001). Analysis of the remaining CIED‑related factors did not show any significant differences between the possible and definite LRIE groups (Table 2).

In‑hospital mortality of patients with definite and possible LRIE was comparable (P = 0.17). Thirty‑day, 1‑year, 3‑year, and long‑term survival was better in the patients with probable LRIE (median [IQR] follow‑up, 5.06 [2.07–8.75] years) than for definite LRIE (median [IQR] follow‑up, 4.61 [1.04–9.4] years; P <⁠0.001) (Table 3, Figure 1).

Multivariable analysis identified the following prognostic factors of long‑term mortality in the patients with definite LRIE: age at TLE, decreased LVEF, higher Charlson comorbidity index, SPE, positive blood culture, and presence of abandoned lead (Table 4).

According to a multivariable analysis, independent factors for long‑term survival of patients with possible LRIE were: age at TLE, low LVEF, and comorbidities (Table 5).

Comparison of long‑term survival of patients with definite and probable LRIE depending on the time from the onset of the first symptoms to diagnosis showed no difference in mortality after TLE (Supplementary material, Figure S1 and S2).

Discussion

Despite constant developments in diagnostics and therapy, CIED infections, especially LRIE, continue to be associated with high mortality. In 19 studies that included 100 or more patients with follow‑up up to 5.5 years, all‑cause mortality after CIED infections was as high as 35%.^12-21 All studies that included only patients with LRIE reported high mortality of 24.5%–29%, with a maximum of a year of follow‑up.^13,20,22 Moreover, 8 studies reporting in‑hospital or 30‑day mortality showed that mortality increased in subsequent observation periods and ranged from 2%–15%^13-21 to 24%–29% at 6 months,^13,22-24 9%–35% at 1 year,^{14,19,20,25-28} and 6%–37% at 2 years or longer.^14-16,29-33 In our study, the 1‑year and 3‑year mortality equaled 8.1% and 22.3% for patients with possible LRIE, and 22.3% and 33.1% for definite LRIE, and was comparable to literature data. Long‑term follow‑up (5.55 years and 6.05 years for definite and possible LRIE, respectively) showed very high mortality of 61.14% for definite LRIE and 49.29% for possible LRIE. This high mortality rate was probably due to the advanced age of the patients (median age at TLE was 71 years [IQR, 61–58]) and high Charlson index (median, 4 [IQR, 3–9]). There are few publications on long‑term survival of patients with CIED‑related infections, and they are based on small study cohorts. According to a recent study,³¹ the mortality rate of 191 patients undergoing TLE (50% infectious indications) during 6.5‑year follow‑up was 33%. Another study, which analyzed a cohort of 125 patients with CIED infections, showed a mortality rate of 24% after 41 months of follow‑up,³² and 1 more study with mean follow‑up of 4.1 years demonstrated a 37% mortality rate in 82 patients with CIED‑related infections undergoing TLE.³³ The risk factors for death in the long‑term follow‑up in the above studies^31-33 were age at TLE, anemia, atrial fibrillation, presence of vegetation, positive blood cultures, and high comorbidity index. Our analysis identified similar factors. The main predictors of long‑term mortality in patients with definite LRIE were age at TLE, decreased LVEF, high Charlson comorbidity index, positive blood cultures, SPE, and presence of an abandoned lead. However, in patients with possible LRIE, only the influence of age, low LVEF, and comorbidities was confirmed. Our analysis showed that patients with possible LRIE had lower long‑term mortality than those with definite LRIE. This is probably related to the use of extended diagnostic criteria, enabling initiation of treatment at an earlier stage (before systemic expansion of the infection), and the more frequent co‑occurrence of pocket infections in this group. The course of infection in our patients with definite LRIE was more severe and featured vegetations, more often more dangerous pathogens, and high values of inflammatory parameters. In this group of patients, abandoned leads and intracardiac abrasion of the leads were also observed more frequently, which indicates a chronic intracardiac infection.

The multivariable analysis showed that the poorer survival of patients with definite LRIE is probably related to the greater extent of the infectious process than in the patients with possible LRIE. Previous studies have mainly confirmed the influence of age and comorbidities on the long‑term prognosis of patients with LRIE, but emphasized that early diagnosis limits formation of vegetation and systemic infection spread thanks to earlier TLE and initiation of appropriate antibiotic therapy.^31-34

The often insidious course of the disease, manifested by relapses of fever or recurrent pneumonia in the absence of vegetations visible on echocardiography, significantly delays a correct diagnosis. Ineffective antibiotic therapy also adversely affects long‑term treatment results.³⁶ We noticed a significant delay (up to several months) in diagnosing LRIE 10 years ago, when the diagnostic criteria for LRIE were more strictly adhered to, and valuable time was wasted searching for other causes of fever. This study did not demonstrate a direct impact of delayed diagnosis on the long‑term survival of patients after TLE, and this is probably related to the introduction of less stringent criteria for diagnosing possible LRIE.