Despite improvements in the diagnostic approach and management strategy, infective endocarditis (IE) remains associated with high rates of in-hospital and long-term mortality, and significant complications [1‐6]. Therefore, rapid identification of patients at high-risk of death could help clinical decision-making with respect to the timing of surgery and intensity of in-hospital care in order to improve prognosis.

Systemic embolization (especially stroke) is a severe complication of IE and is associated with increased morbidity and mortality [7]. Early identification of coagulation activation and thrombus formation with associated biomarkers such as D-dimer may be of prognostic value. D-dimer is a fibrin-degradation product that is released when a blood clot disintegrates, indicating thrombosis or fibrinolysis [8]. It is a valuable blood marker for the diagnosis and evaluation of a vast array of thrombosis-related clinical conditions such as venous thromboembolism, pulmonary thromboembolism, and myocardial infarction [9‐11]. Turak et al. reported that increased D-dimer level at admission was associated with high in-hospital mortality in patients with IE [12]. However, the sample size of their study was small (n = 157), and the impact of D-dimer on long-term outcomes in patients with IE was not discussed. Therefore, in this study, we aimed to evaluate the association of D-dimer at admission with in-hospital and six-month outcomes in a relatively large series of IE patients.

This study investigated the association of D-dimer at admission with the in-hospital adverse events and six-month mortality in a large cohort of patients with IE. Elevation of at-admission D-dimer level was associated with increased rate of in-hospital adverse events and six-month death. D-dimer was an independent predictor for both in-hospital adverse events and six-month mortality.

Early and accurate identification of patients at high risk and timely surgical intervention have been found to improve prognosis in patients with IE [14, 15]. However, the clinical history of IE is highly variable, rendering more focus on identifying predictors of adverse outcomes. Systemic embolization and in particular, CNS embolization, is one of the determinants of adverse outcome in patients with IE [2, 7, 16]. Therefore, early detection of activation of the coagulation cascade could play a pivotal role in recognizing excessive thrombus formation. Previous studies have proposed some clinical and laboratory predictors, including pro-inflammatory and hemodynamic biomarkers [17‐20]. However, they are not directly related to thromboembolism. Identification of novel prognostic biomarkers on the basis of coagulation may further stratify patients with IE based on risk, providing guidance for clinical decision-making.

Elevated D-dimer level in plasma is indicative of acute thrombus formation and fibrinolysis, which is likely a valuable tool to diagnose a vast array of thrombosis-related clinical conditions [21]. Understanding the pathophysiology of IE might be helpful to completely elucidate the underlying mechanism of the prognostic role of D-dimer in IE. The characteristic endocardial lesion in IE—a vegetation—is an aggregation of platelets, fibrin, microorganisms, and inflammatory cells [2, 22]. Development of non-bacterial thrombotic endocarditis which is composed of a platelet–fibrin network is the nidus for bacterial adhesion and invasion. The bacteremia and colonization of bacteria on heart valves further promote platelet aggregation, coverage of the bacteria by a platelet–fibrin meshwork, and formation of mature vegetation [23]. This process involves recruitment of inflammatory cells, release of pro-inflammatory cytokines, and activation of the coagulation cascade. Pro-inflammatory cytokines and other mediators are capable of activating the coagulation system and down-regulating physiologic anticoagulant pathways and fibrinolysis, which reversely modulate the inflammatory process through protease-activated cell receptors and activation of platelets. Hence, down regulation of anticoagulant pathways not only promotes thrombosis but also amplifies the inflammatory process. This interplay of inflammation and coagulation, contributing significantly to the outcome, is one of the most prominent features of sepsis [24]. When the inflammation-coagulation interactions overwhelm the natural defense systems, catastrophic events such as those manifested in sepsis and IE occur. Furthermore, the continued proliferation of bacteria and deposition of platelets and fibrin result in vegetation that can embolize peripherally and cause embolic phenomena. As showed in our study, patients with elevated levels of on-admission D-dimer had a significantly higher rate of in-hospital thromboembolic events than those that did not. Embolism was the most common adverse events in our study cohort, attributing to 70% of all in-hospital events.

The prognosis of IE could be influenced by many factors. Previous studies have identified several predictors for in-hospital and long-term mortality in patients with IE, such as prosthetic heart valve, staphylococcus infection, LVEF, surgical therapy, vegetation size > 10 mm, perivalvular abscess and the presence of complications (stroke, heart failure, renal failure) [1, 5, 7, 12]. After adjusting these risk factors, D-dimer remains an independent predictor for in-hospital and 6-month all-cause mortality in our study. Besides, the presence of NYHA III-IV heart failure, negative blood culture and absence of surgical therapy were also identified as independent risk factors, which is consistent with previous studies. This finding supports the application of D-dimer in clinical practice to acquire additional prognostic information apart from traditional risk factors, especially when D-dimer > 3.5 mg/L.

Currently, many studies have been conducted to investigate the prognostic value of D-dimer in several thrombosis-related conditions. On-admission plasma D-dimer level has been reported to be a valuable marker in predicting short- and long-term outcomes in acute aortic dissection [25]. Similar results were found in patients with ST-elevation myocardial infarction undergoing primary percutaneous coronary intervention [26]. Musicant et al. reported that baseline D-dimer level was significantly associated with the occurrence of myocardial infarction in subjects with symptomatic peripheral artery disease [27]. To date, however, only four studies have been performed to evaluate the role of D-dimer in outcome prediction in IE. In a study including 42 patients, Bakal et al. showed that plasma D-dimer levels were increased in patients with IE who suffered from clinically significant systemic embolism; D-dimer level > 425 ng/dL could predict clinical embolism with a sensitivity of 77% and specificity of 62% [28]. However, the association of D-dimer level with in-hospital mortality in patients with IE was not studied. Turak et al. included 157 patients with IE and found that on-admission D-dimer ≥ 4.2 mg/L was independently associated with IE-related in-hospital death [12]. Recently, Baris et al. enrolled 79 patients with IE, and using multiple logistic regression analysis, showed that D-dimer was a strong parameter for predicting in-hospital mortality and embolic events [29]. Nevertheless, the small sample size and lack of long-term follow-up limit the significance of their results. Our study was conducted in a relatively large sample of patients, and the long-term follow-up demonstrated that increased D-dimer level on admission was independently associated with adverse in-hospital and long-term outcome in patients with IE. Additionally, this association was still significant after adjusting for cardiac function and surgical therapy, which have been shown as strong predictors for poor outcome in the previous studies and the current guideline. Therefore, we believe our study results provide further evidence for supporting D-dimer as a reliable biomarker to predict increased risk of complications and in-hospital and long-term death in patients with IE. Patients diagnosed with IE with increased D-dimer level should be closely monitored for embolization and carefully evaluated for early surgical intervention.

Our study has some limitations. First, it is a single center study and clinical data were retrospectively collected from electronic medical records. Our results should be interpreted with caution and require validation by prospective, multi-center studies. Second, reasons of death cannot be clearly identified by telephone follow-up. Third, Third, some peri-operative parameters that may be associated with patient outcome were not available in our study and not included into the analysis. Fourth, we only recorded on-admission D-dimer level without serial measurement. Measurement of D-dimer levels after treatment or upon discharge would be helpful to further analyze the correlation with disease progression and evaluate long-term outcome.

Prognostic evaluation of patients with IE is of utmost clinical interest, which often guides clinicians to develop an algorithm for risk stratification and decision-making. Our study suggested that increased on-admission D-dimer was a reliable prognostic biomarker that associated with high risk of in-hospital adverse events and six-month mortality in patients with IE.