Hyperventilation/Breath-Hold Maneuver to Detect Myocardial Ischemia by Strain-Encoded CMR: Diagnostic Accuracy of a Needle-Free Stress Protocol

Objectives: The purpose of this study was to evaluate the diagnostic accuracy of a fast, needle-free test for myocardial ischemia using fast Strain-ENCoded (fSENC) cardiovascular MR (CMR) after a hyperventilation/breath-hold maneuver (HVBH).

Background: Myocardial stress testing is one of the most frequent diagnostic tests performed. Recent data indicate that CMR first-pass perfusion outperforms other modalities. Its use, however, is limited by the need for both, a vasodilatory stress and the intravenous application of gadolinium. Both are associated with added cost, safety concerns, and patient inconvenience. The combination of 2 novel CMR approaches, fSENC, an ultrafast technique to visualize myocardial strain, and HVBH, a physiological vasodilator, may overcome these limitations.

Methods: Patients referred for CMR stress testing underwent an extended protocol to evaluate 3 different tests: 1) adenosine-perfusion; 2) adenosine-strain; and 3) HVBH-strain. Diagnostic accuracy was assessed using quantitative coronary angiography as reference.

Results: A total of 122 patients (age 66 ± 11years; 80% men) suspected of obstructive coronary artery disease were enrolled. All participants completed the protocol without significant adverse events. Adenosine-strain and HVBH-strain provided significantly better diagnostic accuracy than adenosine-perfusion, both on a patient level (adenosine-strain: sensitivity 82%, specificity 83%; HVBH-strain: sensitivity 81%, specificity 86% vs. adenosine-perfusion: sensitivity 67%, specificity 92%; p < 0.05) and territory level (adenosine-strain: sensitivity 67%, specificity 93%; HVBH-strain: sensitivity 63%, specificity 95% vs. adenosine-perfusion: sensitivity 49%, specificity 96%; p < 0.05). However, these differences in diagnostic accuracy disappear by excluding patients with history of coronary artery bypass graft or previous myocardial infarction. The response of longitudinal strain differs significantly between ischemic and nonischemic segments to adenosine (ΔLS_ischemic = 0.6 ± 5.4%, ΔLS_nonischemic = -0.9 ± 2.7%; p < 0.05) and HVBH (ΔLS_ischemic = 1.3% ± 3.8%, ΔLS_nonischemic = -0.3 ± 1.8%; p = 0.002). Test duration of HVBH-strain (t = 64 ± 2 s) was significantly shorter compared with adenosine-strain (t = 184 ± 59 s; p < 0.0001) and adenosine-perfusion (t = adenosine-perfusion: 172 ± 59 s; p < 0.0001).

Conclusions: HVBH-strain has a high diagnostic accuracy in detecting significant coronary artery stenosis. It is not only significantly faster than any other method but also neither requires contrast agents nor pharmacological stressors.