Risk stratification and outcomes in hemodynamically stable patients with acute pulmonary embolism: a prospective, multicentre, cohort study with three months of follow-up

Summary. Background: The role of risk stratification innormotensive patients with acute pulmonary embolism (PE)is still unclear. Objectives: We evaluated, in these patients, theusefulness of six prognostic markers for predicting in-hospitaladverse events related to PE and 3-month mortality. Patients/Methods: Two hundred and one consecutive patients withconfirmed acute PE and normal blood pressure, who wereadministered conventional anticoagulation, were recruited in amulticentre prospective cohort study with 3 months of follow-up. At baseline, they received a comprehensive risk-evaluationincluding echocardiographic assessment of right ventriculardysfunction, determination of troponin I, brain natriureticpeptide and D-dimer, arterial blood gas analysis and a clinicalscore.PrimaryoutcomeofthestudywasPE-relatedin-hospitaldeath or clinical deterioration. Secondary outcomes were in-hospitaland3-monthall-causemortality. Results: Theprimaryoutcome occurred in one patient (0.5%), who died from PEduring hospitalization. The in-hospital and 3-month all-causemortality were 2% and 9%, respectively. None of theprognostic markers was predictive of the primary outcome.Clinical score, troponin I and hypoxemia predicted in-hospitalall-cause mortality (P = 0.02, 0.01 and < 0.01, respectively).Clinical score (HR, 4.7; 95% CI, 1.9–12.0), D-dimer (4.8;1.4–16.3), hypoxemia (5.7; 2.1–15.1) and troponin I (7.5; 2.5–22.7) were predictors of 3-month all-cause mortality onunivariate analysis. On multivariate analysis clinical score andtroponinIremainedindependentlypredictive. Conclusions: Wedid not find prognostic markers useful as predictors ofin-hospital PE-related adverse events. Clinical score, troponinI and hypoxemia predicted in-hospital all-cause mortality.Clinical score and troponin I independently predicted 3-monthall-cause mortality.Keywords: hemodynamic stable, prognosis, pulmonaryembolism.IntroductionPatients with acute pulmonary embolism (PE) and cardiogenicshockorpersistenthypotensionhaveapoorprognosis,andarelikely to benefit from a more aggressive treatment thananticoagulation alone [1].Althoughlower thaninpatients withhemodynamicinstabil-ity, the in-hospital mortality rateof normotensive patients withPE reached values of about 10% in three large studies [2–5]. Intwo of these works, PE was reported to cause a relevantproportion of deaths [3,5]. Furthermore, in the ICOPERregistry and in one randomized trial, normotensive patientswithrightventriculardysfunction(RVD)facedmorefrequentlyadverse outcomes [3,6]. Therefore, risk-stratification toolsaimed at identifying hemodynamically stable patients at higherriskofunfavourableeventshavebeenadvocated,inviewalsooftheirpossibleuseasaguideformoreaggressivetreatments[7,8].Right ventricular dysfunction, as assessed by echocardio-graphy or spiral computed tomography, is the most commonlyused method for risk stratification in acute PE [9]. It has beenassociated with a 2-fold increase in short-term mortality rate[3,10], and has been used as a guide for thrombolysis inhemodynamically stable patients [11,12].In recent years several studies have showed that, in patientswith acute PE, abnormal levels of cardiac biomarkers such astroponins T and I, brain natriuretic peptide (BNP) and N-terminal-pro BNP, correlate with RVD and predict anincreased risk of in-hospital adverse outcomes [7].