Dynamic handgrip exercise for the detection of myocardial ischemia using fast Strain-ENCoded cardiovascular magnetic resonance

Abstract

Background: Previous data suggest dynamic handgrip exercise (DHE) as a potential physiological, needle-free stressor feasible for cardiovascular magnetic resonance (CMR) examinations. DHE-fast Strain-ENCoded imaging (fSENC) is potentially cost-saving, ultra-fast and avoids pharmacological side effects thereby targeting the drawbacks of conventional pharmacological stress CMR.

Objectives: To assess the diagnostic accuracy of DHE-fSENC for detecting ischemia-related wall motion abnormalities in suspected obstructive coronary artery disease (CAD).

Methods: Patients with known or suspected obstructive CAD referred for CMR stress testing were prospectively enrolled. Diagnostic accuracy was assessed in comparison to pharmacological stress CMR and in a subgroup, compared to invasive coronary angiography (ICA). The CMR protocol was extended by both-handed DHE with 80 repetitions per minute over 2 min followed by fSENC short-axis acquisition before pharmacological stress testing. Stress-induced impairment of regional longitudinal strain was graded suspicious for obstructive CAD.

Results: Two-hundred sixty individuals with cardiovascular high-risk profile (64±13years, 75% male) were enrolled. DHE-fSENC provided a sensitivity of 79% (95% CI: 64-89) and specificity of 87% (95% CI 82-91) compared to pharmacological stress CMR. In a subgroup of 105 patients with recent ICA, high diagnostic accuracy was found for the detection of obstructive CAD (sensitivity 82% [95% CI: 67-92], specificity 89% [95% CI: 78-95]). Exam duration of DHE-fSENC was significantly reduced compared to conventional CMR stress protocols (DHE-fSENC 207±69 s vs. adenosine-perfusion 287±82 s vs. dobutamine-cine 1132±294 s, all p<0.001).

Conclusion: DHE-fSENC allows for a reliable and fast detection of obstructive CAD, thereby expanding the applicability of needle-free CMR stress testing.

Keywords: CMR; FSENC; Handgrip; Ischemia; Longitudinal strain.

DHE and fSENC imaging. Schematic visualization of the both-sided, DHE at approx. Fifty percentage of MVC and a frequency of 80/min including fSENC acquisition at rest and during DHE. The comparison between fSENC at rest and after DHE of the apical slice in this example shows an impairment of LS after DHE of the inferior apical wall. DHE and fSENC acquisition was integrated into CMR routine protocol and was performed before pharmacological stress. DHE-fSENC: response of LS. Comparison of LS at rest and during DHE. In ischemic segments, LS was significantly impaired during DHE. In contrast, LS was more pronounced during DHE in non-ischemic segments. DHE-fSENC: test accuracy compared to pharmacological stress CMR. Test accuracy of DHE-fSENC compared to pharmacological stress CMR was excellent with a sensitivity of 79% (64–89) and a specificity of 87% (82–91). CMR cardiac magnetic resonance imaging, DHE dynamic handgrip exercise, fSENC fast Strain-ENCoded imaging, LS longitudinal strain, MVC maximum voluntary contraction, NPV negative predictive value, PPV positive predictive value, SEM standard error of the mean.

Fig. 1

DHE and fSENC imaging. Schematic visualization of the both-sided, DHE at approx. Fifty precentage of MVC and a frequency of 80/min including fSENC acquisition at rest and during DHE. The comparison between fSENC at rest and after DHE of the apical slice in this example shows an impairment of LS after DHE of the inferior apical wall. DHE and fSENC acquisition was integrated into CMR routine protocol and was performed before pharmacological stress. DHE dynamic handgrip exercise, fSENC fast Strain-ENCoded imaging, MVC maximum voluntary contraction, LS longitudinal strain

Fig. 2

Overview of left ventricular strain directions and their corresponding CMR planes. In contrast to commonly used CMR post-processing tools like feature tracking or speckle-tracking echocardiography, short-axis fSENC slices are required for the assessment of LS, long-axis fSENC slices are required for CS. RS cannot be assessed using fSENC. In this DHE-fSENC study, only LS was assessed. LS longitudinal strain, CS circumferential strain, RS radial strain, LAX long axis, SAX short axis, fSENC fast Strain-ENCoded imaging, DHE dynamic handgrip exercise, CMR cardiovascular magnetic resonance

Fig. 3

Flowchart of patient recruitment. Three hundred thirty nine patients agreed to the handgrip study. In total, 79 patients had to be excluded mainly because of technical reasons including bad CMR image quality or failed triggering/bad quality of fSENC sequence leading to a final study population of 260 patients. CMR cardiac magnetic resonance imaging, DHE dynamic handgrip exercise, fSENC fast strain-encoded imaging, ICA invasive coronary angiography

Fig. 4

Stress test duration of DHE-fSENC. The duration of DHE-fSENC was significantly reduced compared to standard pharmacological stress tests (p<0.001). DHE dynamic handgrip exercise, fSENC fast Strain-ENCoded imaging

Fig. 5

(B) DHE-fSENC: longitudinal strain response. Comparison of LS at rest and during DHE. In ischemic segments, longitudinal strain was significantly impaired during DHE. In contrast, LS was more pronounced during DHE in non-ischemic segments. DHE dynamic handgrip exercise, LS longitudinal strain, SEM standard error of the mean

Fig. 6

Examples. (A) Negative stress test. 61-year-old, male patient. Progressive dyspnea without typical angina. ICA revealed proximal LAD stenosis with 50% lumen narrowing. Adenosine-perfusion was rated negative. During DHE, the HR increased from 60/min to 98/min, GLS was more pronounced after DHE, from −16.7% to −20.2%. DHE-fSENC was rated negative for ischemia. (B) Positive stress test. 75-year-old, male patient. No dyspnea or typical angina. Former resuscitation during non-ST-elevation myocardial infarction due to LAD stenosis 6 months before. Residual stenosis of LCx (75%). Adenosine-perfusion was rated positive lateral midventricular (2 segments). During DHE HR increased from 54/min to 83/min, GLS slightly increased from −19.0% to −19.5%. In the qualitative analysis of DHE-fSENC a worsening of LS lateral midventricular was observed, similar to adenosine-perfusion – DHE-fSENC was rated positive. (C) “False-positive” stress test. 57-years-old, female patient. Three months before - after progressive, typical angina –PCI including the insertion of a drug-eluting stent into the distal RCA was performed. The patient was referred to stress CMR for residual LCx and LAD stenoses (both 50–75% lumen narrowing). Adenosine-perfusion revealed no significant perfusion deficit. During DHE (HR increase from 75/min to 115/min, GLS increase from −19.6% to −22.5%), a new subendocardial worsening of LS of the inferior wall (2 segments) was detected. Two weeks after CMR the patient presented at chest pain unit with typical angina and ST-segments elevation (II, III, and aVF). In ICA, a subtotal occlusion of RCA, independent of the first lesion, was found and a drug-eluting stent was successfully inserted. CMR cardiac magnetic resonance imaging, DHE dynamic handgrip exercise, fSENC fast Strain-ENCoded imaging, GLS global longitudinal strain, HR heart rate, ICA invasive coronary angiography, LAD left anterior descending artery, LCx left circumflex artery, NSTEMI non-ST-elevation infarction, PCI percutaneous coronary intervention, RCA right coronary artery