Integrated Diagnostic Algorithm for Acute Aortic Syndromes

Background and rationale. Acute aortic syndromes (AASs) are deadly and difficult to diagnose cardiovascular emergencies. AASs include acute aortic dissection, intramural aortic hematoma, penetrating aortic ulcer and spontaneous aortic rupture. A conclusive diagnosis of AASs requires an advanced imaging exam, such as computed tomographic angiography (CTA), transoesophageal echocardiography (TEE) or magnetic resonance imaging. These exams are time-consuming, costly and frequently require transfer of patients to specialized clinical centres. CTA, the most frequently used advanced imaging exam for suspected AASs in the Emergency Department (ED), exposes patients to significant ionizing radiations and to contrast medium, carrying inherent risks of anaphylaxis and contrast nephropathy. Selection of patients necessitating an advanced aortic imaging exam is therefore worrisome. There is a need for standardized diagnostic management of AAS workup, and in particular for a validated diagnostic algorithm incorporating evidence-based decision rules allowing:

1. rapid identification of patients necessitating urgent CTA;
2. safe rule-out of patients not requiring CTA. Based on available guidelines and further evidence, standardized decision rules for rule-in and rule-out of AASs can effectively based on integration of: (1) standardized pre-test probability assessment, (2) bedside imaging of the aorta with focus cardiac ultrasound (FoCUS), and (3) biomarker testing (D-dimer assay).

Aim of the study. The aim of the present study is to evaluate the performance of integrated diagnostic algorithms for standardized diagnostic workup (rule-in/out) of AASs. Primary outcome will be the occurrence of AASs in patients at low clinical pre-test probability presenting a negative D-dimer test result.

Setting. The study will be performed in several international EDs. Participating EDs will satisfy the following criteria for 24 hours / 7 days, in their routine clinical practice: (1) experienced use of bedside FoCUS, (2) local access to urgent D-dimer (DD) assay, (3) local access to urgent CTA. Both hub and spoke centers will be allowed to participate in the study, if local census is >30.000 visits/years.

Study design. This will be an observational, prospective, multicentre, international, non-profit, investigator-driven, outcome and diagnostic accuracy study. The study protocol does not imply any modification in the routine diagnostic and therapeutic approach to patients with suspected AASs. The only difference from standard care will be:

1. registration of selected demographic and clinical data of study patients, obtained during the index ED visit;
2. follow-up (by telephone or clinical visit) of patients after 1 month from the index ED visit, for collection of selected clinical data.

Study population. The study will be performed on adult outpatients evaluated in the ED for AAS-compatible symptoms (chest/abdominal/back pain, syncope, neurological deficit, limb ischemia), in whom AASs are considered in differential diagnosis by the attending physician(s). Eligibility will be defined during or after medical evaluation in the ED and always before establishment of a final diagnosis. In each centre, patient recruitment will start after obtaining approval from the local Ethic Committee.

Study methods and organization. Patients will be evaluated in the ED by an authorized physician practicing Emergency Medicine and will be managed in accordance with international guidelines. The recommended standard of care for diagnosis and treatment of patients with suspected AAS will be as indicated by the European Society of Cardiology 2014 guidelines (see diagnostic management protocol section). However, attending physicians will be allowed to derogate from guideline indications based on the clinical scenario, irrespective of patient's participation to the present study. After medical evaluation, a clinical suspicion of AAS will prompt to possible enrolment in the study. The attending physician or an authorized study research investigator or collaborator will then acquire patient (or next of kin, if deemed necessary) consent to participate to the study and will fill a standardized clinical form registering the relevant clinical data.

Pre-test probability assessment. During the index visit, physicians will collect relevant clinical data and evaluate the presence/absence of pre-specified clinical variables. The pre-test probability of AAS will be assessed using the aortic detection dissection risk score (ADDs). The ADDs, recommended by the AHA 2010 and the ESC 2014 guidelines, is based on 12 variables organized in 3 categories. Patients are classified at low probability if ADDs is ≤1 (i.e. satisfying criteria in 0 or 1 risk category), or at high probability if ADDs is ≥2 (i.e. satisfying criteria 2 or 3 risk categories). During data analysis, the pre-test probability of AAS will be assessed also with the AORTA score (AORTAs). The AORTAs, developed and validated by the Study Coordinators, is a simplified score based on only 6 ADDs variables. Patients are classified at low probability if AORTAs is ≤1, or at high probability if AORTAs is ≥2. AORTAs based classification will be used in data analysis, and not for clinical decision.

D-dimer. D-dimer levels will be analysed in the hospital laboratory on venous samples with an automated test. Diagnostic tests will be site-specific. In the study centers, the D-dimer test results will be interpreted by clinicians using a fixed cutoff of 500 ng/mL (Fibrinogen Equivalent Unit, FEU). Data analysis will also evaluate the performance of an age-adjusted cutoff, calculated in each patient as: age (years) x 10, with a minimum of 500 ng/mL.

Focused cardiac ultrasound. The attending physician, a cardiologist or an authorized study research investigator expert in bedside emergency ultrasonography will perform a FoCUS on study patients during ED evaluation, before advanced aortic imaging tests or surgery. Evaluation of the heart and aorta will be performed with the patient in the supine or left lateral decubitus positions, using the following views: (1) left parasternal (long-axis) and (2) suprasternal notch view. Additional views including apical, subcostal, right parasternal, abdominal and neck for carotid arteries will be decided by the attending physician.

The following will be considered as direct sonographic signs of AASs:

1. presence of an intimal flap separating two aortic lumens;
2. presence of an intramural aortic thickening (circular or crescentic thickening of the aortic wall >5 mm);
3. presence of a crater-like outpouching with jagged edges in the aortic wall.

The following signs will also be researched, as potential indirect sonographic signs of AASs:

1. thoracic aorta dilatation (diameter ≥4 cm);
2. pericardial effusion or tamponade
3. aortic valve regurgitation at colour Doppler (at least moderate), based on subjective evaluation.

Integrated diagnostic strategies. In study centers, AAS will be considered to be ruled out without advanced imaging by a D-dimer level <500 ng/mL, in patients with a low clinical pre-test probability assessed with the ADDs. The study will also assess and compare alternative diagnostic strategies integrating pre-test probability assessment, FoCUS and D-dimer (DD). Data analysis will evaluate all possible integrations of these tools. Based on available evidence, key rule-out strategies will be: ADDs/DD 500 ng/mL, ADDs/DD age-adjusted, AORTAs/DD 500 ng/mL, AORTAs/DD age-adjusted. FoCUS results will be variably combined to optimize diagnostic performance.

Advanced Aortic imaging. The advanced aortic imaging exam used to confirm or refuse a diagnosis of AASs will be CTA, TEE or magnetic resonance angiography, performed by physicians not involved in the present study, according to local procedures. Patient enrolment in the study will not change standard patient management and will not affect medical decision to perform aortic imaging. Instruments used for imaging will be site-specific. These exams will be performed and interpreted by specialized radiologists, cardiologists or surgeons not involved in the present study.

Diagnostic management protocol. In participating centers, patient care will be managed according to the European Society of Cardiology (ESC 2014) algorithm. Clinically stable patients at low clinical pre-test probability (defined by ADD score ≤1) and a D-dimer level <500 ng/mL FEU, will be considered suitable for diagnostic rule-out (i.e. exclusion of AAS without perfoming advanced aortic imaging). All other patients, including patients at high clinical pre-test probability (defined by ADD score ≥1) and patients at low clinical pre-test probability with D-dimer ≥500 ng/mL FEU, will be subjected to advanced aortic imaging. This protocol will be considered as standard of care in the participating centers. However, attending physicians will be allowed to derogate from protocol indications based on the clinical scenario, and irrespective of patient's participation to the present study.

Follow-up data. AASs are difficult to diagnose disease entities. Given the severity of AASs in untreated patients, investigators assume that individuals with an undiagnosed AAS would experience major clinical events leading to conclusive diagnosis within 30 days from the ED visit. Therefore, all study patients will enter a 30-day clinical follow-up to allow accurate case adjudication.

Final diagnosis. The final diagnosis will be established in each centre dichotomically (presence or absence of an AAS) by two senior investigators, who will review all available clinical data (including aortic imaging studies, medical, surgical, autopsy records if applicable). Reviewers will be blinded to pre-test probability assessment, FoCUS findings and D-dimer levels. The reviewers will judge on presence or absence of an AAS, and define the AAS type (acute aortic dissection, aortic intramural hematoma, penetrating aortic ulcer or spontaneous aortic rupture), and Stanford classification. In patients without AAS, a prevalent alternative diagnosis will be defined.

Analysis and statistical methods. Dichotomous data will be expressed as proportions with their 95% confident interval (95% CI) and continuous data as mean and standard deviation (SD). The chi square test will be used for the comparison of dichotomous data, and the unpaired Student's t-test will be used to compare normally distributed data. The diagnostic performance of diagnostic tests and strategies will be assessed by computing sensitivity, specificity, negative/positive predictive values and negative/positive likelihood ratios. The failure rate (or false negative proportion) will be calculated as the number of patients with a final diagnosis of AAS divided by the number of patients satisfying rule-out criteria. The rule-out efficiency will be calculated as the number of patients satisfying the rule-out criteria divided by the number of all included patients. The rule-in efficiency will be calculated as the number of patients satisfying the rule-in criteria (i.e. proceeding to advanced aortic imaging) divided by the number of all included patients. P-values will be two-sided, and a P-value lower than 0.05 will be considered as statistically significant. Analysis will be performed with the Statistical Package for the Social Sciences (SPSS) statistical package (version 17.0, SPSS Inc., Chicago, Illinois).

Sample size. The study will be sized to test the null hypothesis that the failure rate of the integrated rule-out diagnostic strategy (low probability/DD-) exceeds 2%. Investigators assume, based on literature data, that the punctual failure rate is 0.5%. With a type I error (alpha) of 0.05 and a type II error (beta) of 0.2, approximately 496 patients at low probability with a negative DD will have to be enrolled, to reject the null hypothesis. Assuming that such patients are 30% of enrolled subjects, approximately 1653 patients should be analysed. Assuming a 10% rate of patients lost at follow-up, at least 1837 patients should be enrolled.