Endocarditis Patients with Preoperative Neurological Complications. Do Radiological Lesions Predict Neurological Outcome?

Background: Infective endocarditis (IE) is associated with high mortality (20–40%) and neurological complications (20–50%). Although radiological brain lesions are important in decision making in IE, their impact on outcome is not clear. We analyzed the influence of different types of preoperative neurological complications in our patients on outcome.

Methods: We retrospectively reviewed all charts, brain imaging, and follow-up data from patients operated for left-sided endocarditis between Jan-07 and April-13. We performed Chi-Square, multivariate, and Cox-regression analyses.

Results: A total of 308 patients (age 62.0 ± 13.9) underwent surgery for IE. The preoperative neurological complications were as follows: stroke in 91 patients (56 ischemic and 35 hemorrhagic), silent cerebral infarction in 27 patients and transient ischemic attacks in 4 patients. In-hospital mortality was higher in patients with preoperative stroke than those without it. However, the difference was not statistically significant (27.5% versus 22.6%, p = 0.38). The incidence of death because of neurological complications (massive ischemic infarction or ICB) was 2.6% (n = 8) and did not differ between patients with or without preoperative stroke (3.3% vs 2.3%, respectively, p = 0.70). Cox-regression analysis showed that preoperative stroke did not affect long-term survival (hazard ratio 0.78, Confidence interval 0.53–1.13). Newly occurring postoperative stroke (new incidence or exacerbation of preoperative stroke) was found in 10.7% (n = 33) of patients and was hemorrhagic in 57.6% and ischemic in 42.4% of cases. The presence of preoperative stroke was associated with higher incidence of newly occurring postoperative stroke (Odds ratio 2.2, 95% CI 1.1–4.6). The radiological lesions in patients with preoperative stroke were as follows: ischemic without hemorrhagic transformation (HT) in 60.4%, ischemic with HT in 18.7%, ischemic with micro bleeds (MB) in 16.5%, and space occupying intra-cerebral hemorrhage in 3.3% of cases. The preoperative combination of MB and ischemic lesions was slightly associated with more new postoperative stroke than the combination of embolic lesions with HT (20.0% vs 17.6%, p = 0.69).

Conclusions: The results suggest that the presence of preoperative ischemic or hemorrhagic stroke in IE patients does not significantly affect hospital mortality or survival. Importantly, the preoperative brain lesion characteristics do neither predict neurological outcome nor mortality.