Predictors of long-term survival free from relapses after extraction of infected CIED

Aims We explored the possible predictors of long-term prognosis after transvenous lead extraction (TLE) for a cardiac implantable device related infection (CIEDI), including the modified Duke score result. Methods and results We performed a single centre prospective observational study in a population of consecutive patients referred for TLE to a teaching hospital to treat a CIEDI without associated valve-endocarditis. 121 patients were enrolled between January 2012 and March 2016. According to the modified Duke criteria, the presence of CIED-related endocarditis was rejected in 54.5%, possible in 21.5%, and definite in 24.0%. 20/121 patients died after a mean follow-up of 46.0 ± 2.5 months, while 7 patients reported hospitalization for CIEDI recurrence/relapse in the same period. Modified Duke score was significantly associated with a poor prognosis at univariate Cox regression analysis (HR 1.847, 95% CI 1.160-2.941; P = 0.010). However, the three factors independently associated with death and/or CIEDI relapse/recurrence were: a 'closed' CIED pocket (HR 2.720; 95% CI 1.135-6.520), presence of ghost at post-TLE transoesophageal echocardiography (HR 3.469; 95% CI 1.420-8.878), and a GFR <60 (HR 4.565; 95% CI 1.668-12.493). Conclusion CIEDI has a poor long-term prognosis despite an effective TLE. Renal failure, presence of 'ghosts' at post-TLE transoesophageal echocardiography and a closed CIED pocket are associated with a worse prognosis.