Kevin O’Connor, MD
About one fifth of acute ischemic strokes and transient ischemic attacks stem from cardioembolism. Although cardiac ultrasound is generally recommended as a reasonable diagnostic study, there is ongoing debate on the utility of transthoracic echocardiography (TTE) versus transesophageal echocardiography (TEE). In practice, TTE may be preferred as it is less invasive and easier to obtain but at the expense of decreased sensitivity when evaluating for some pathologies involving the aorta and left atrium. Thomalla et al. designed the Comparative Effectiveness Study of Transthoracic and Transesophageal Echocardiography in Stroke (CONTEST) study to compare the diagnostic yield of treatment-relevant findings (i.e., findings sufficient to justify a change in medication, intervention, or surgery) of TTE and TEE in patients with acute ischemic stroke, transient ischemic attack, or retinal ischemia of undetermined cause. Despite early study termination due to funding cessation, 494 patients were enrolled with 454 undergoing both TTE and TEE.
TTE identified treatment-relevant findings in 64 patients (14.1%) compared to 86 patients (18.9%; P<0.001) with TEE. Of these, 7 patients only had treatment-relevant findings with TTE and 29 only had treatment-relevant findings with TEE. Patent foramen ovale (PFO; n=23) comprised the majority of findings detected by TEE but not TTE while regional wall motion abnormalities (n=5) comprised the majority of findings detected by TTE but not TEE. Findings from TEE (but not TTE) that resulted in an immediate change in management occurred in 13 patients (2.6% of patients in intention-to-treat population [494] and 2.9% of per-protocol population [454]) and included PFO (n=10) as well as endocarditis (n=2) and left atrial thrombus (n=1). Considering protocol patients ≤60 years (n=191), 27 patients had treatment-relevant findings with TEE but not TTE (14.1%).
After adjudication, TEE yielded treatment-relevant findings in 27 out of 454 patients yielded a number needed to diagnose of 16. Limiting analysis to patients ≤60 years with treatment-relevant findings (n=27), the number needed to diagnose decreased to 7, predominantly driven by the presence of PFO and evidence-based guidance on the benefit of closure. When considering the 11 (2.4%) per-protocol patients in whom TEE negated potentially treatment-relevant findings from TTE, TEE yielded treatment-relevant information in 40 (8.8%) patients.
Although the data show that TEE can provide treatment-relevant information beyond TTE in patients with acute ischemic stroke and TIA, the choice of TEE or TTE will likely be decided on a per-patient basis in the absence of consensus guidance. Further studies may reveal if TEE has greater utility early in the course of a workup (as in this study) or as a subsequent diagnostic tool if a patient has an embolic stroke of unknown source.
