Dynamic Handgrip Exercise: Feasibility and Physiologic Stress Response of a Needle-Free Cmr Stress Test

Purpose: CMR pharmacological stress-testing is well-established to detect myocardial ischemia. Despite stressor and contrast agents appear rather save, contraindications and side effects have to be considered. Substantial costs are further limiting its applicability. Dynamic handgrip exercise(DHE) may have the potential to address these shortcomings as a physiological stressor. We therefore evaluated the feasibility and physiologic stress response of DHE in relation to pharmacological dobutamine-stimulation by cardiac magnetic resonance(CMR). Methods: Two subgroups were prospectively enrolled: (i)volunteers without relevant disease and (ii)patients with known CAD referred for stress-testing. A both-handed, metronome-guided DHE was performed over 2 minutes continuously with 80 contractions/minute by all participants, whereas dobutamine stress-testing was only performed in group(ii). Short axis strain by fast-Strain-ENCoded imaging was acquired at rest, immediately after DHE and during dobutamine infusion. Results: Eighty middle-aged individuals(age 56±17years, 48males) were enrolled. DHE triggered signicant positive chronotropic(HR rest :68±10bpm, HR DHE :91±13bpm, p<0.001) and inotropic stress response(GLS rest :-19.4±1.9%, GLS DHE :-20.6±2.1%, p<0.001). Exercise-induced increase of longitudinal strain was present in healthy volunteers and CAD patients to the same extent, but in general pronounced from basal to apical layers(p<0.01). DHE was aborted by a minor portion(7%) due to peripheral fatigue. The inotropic effect of DHE appears to be non-inferior to intermediate dobutamine-stimulation(GLS DHE =-19.5±2.3%, GLS Dob =-19.1±3.1%, p=n.s.), whereas its chronotropic effect was superior (HR DHE =89±14bpm, HR DOB =78±15bpm, p<0.001). Conclusions: DHE causes positive ino- and chronotropic effects superior to intermediate dobutamine-stimulation, suggesting a relevant increase of myocardial oxygen demand. DHE appears safe and timesaving with broad applicability. The data encourages further studies to determine its potential to detect obstructive CAD.


Introduction
Cardiac magnetic resonance (CMR) stress testing to quantify myocardial ischemia provides an excellent prognostic value that is noninferior to invasive fractional ow reserve (FFR) measurements and is therefore suggested by current guidelines to direct revascularisation therapy in chronic coronary syndrome [1][2][3][4][5]. Current CMR protocols, however, have certain shortcomings restricting their applicability in a relevant number of patients [6][7][8][9]. On one hand, there are safety aspects due to the dependency on pharmacological agents. Myocardial perfusion imaging relies on vasodilating stressors (e.g., adenosine or regadenosone) and gadolinium-based contrast agents. Less commonly used are adrenergic stressors (e.g., Dobutamine) that increase the myocardial oxygen demand and allow to detect coronary insu ciencies by inducible wall motion abnormalities. Although, side effects of these agents are rare, contraindications and complicating risk factors such as hemodialysis, bronchial asthma or ventricular arrhythmogenicity have to be critically assessed. Furthermore, long-term implications of gadolinium depositions in the brain remain unclear. On the other hand, the time consuming CMR stress tests cause tremendous running costs for CMR scanners, personnel, and pharmacological agents paired with incomplete reimbursement. These arguments taken together, have led to a hesitant adoption of CMR stress testing in the majority of health care providers worldwide despite its proven bene ts. Various physiological exercises were sought to replace these downsides of pharmacological stressors, but neither MR-conditional ergometers, steppers nor treadmills could be established since these protocols were found to be time-consuming and exercise-related body motion severely affected image quality [10].
We aimed to address these shortcomings by dynamic handgrip exercise (DHE) as a modi ed needle-free physiological stress test. Unlike previously assessed static handgrip maneuvers, we expected a "dobutamine"-equivalent, positive ino-and chronotropic effect in response to repetitive isotonic bothhanded contractions without relevant body motion [11][12][13][14]. The goal of this study was to assess the feasibility and hemodynamic effect of DHE in healthy volunteers and in relation to varying doses of continuous dobutamine infusion.

Study population and design
Participants were prospectively enrolled at our Department between December 2019 and March 2020 after individual signed consent in two subgroups of (i) volunteers without relevant history of disease and (ii) patients with known CAD that were referred for CMR stress testing and underwent dobutamine stress (predominantly due to contraindications to adenosine, e.g. bronchial asthma). All participants answered a speci c questionnaire for symptoms (pre-and post-stress), risk factors and relevant preexisting illnesses. Group (i) of healthy volunteers excluded individuals with history, signs or symptoms of a cardiac disease -except mild arterial hypertension or other existing, isolated cardiovascular risk factors. Participants unable to perform DHE, with impaired LV ejection fraction (EF) < 50% patients, evidence of stress-induced perfusion de cit or previous myocardial infarction were excluded. The study was approved by the institution's ethics committee and was in accordance to the Declaration of Helsinki. CMR acquisition protocol CMR was performed on a 1.5 Tesla or 3 Tesla clinical scanner (Ingenia and Ingenia CX®, Philips Healthcare, Best, The Netherlands) with a dedicated 32-element cardiac phased array receiver coil. Rwave triggered SSFP cine sequences were acquired in long-(2-, 3-, 4-chamber views) and short axis (apical, midventricular and basal) views with 35 phases per cardiac cycle. As previously described, fSENC was performed as a single heartbeat acquisition [15]. fSENC sequences were acquired at rest and after two minutes of DHE at apical, midventricular and basal short axis layers. The speci c study protocols for healthy individuals (group i) performing DHE alone was extended in CAD patients (group ii) with a dobutamine stress test as demonstrated in Figure 1a. In dobutamine stress, infusion rate started at 10 µg/kg body weight/minute with increments of 10 µg/kg body weight/minute every 3 minutes and a maximum dose of 40 µg/kg body weight/minute. Additionally, fractions of 0.25 mg atropine were substituted to achieve the target heart rate (85% x (220-age)). Three short axis and three long axis cine sequences were acquired at each stress level [16].
Dynamic handgrip exercise DHE was performed with both-sided, metronome-guided rhythmic hand contractions for two minutes ( Figure 1b). Commercially available, CMR-capable rubber handgrip rings in three different strengths (30lb, 50lb, 70lb) were offered to the subjects before the scan started. Maximal voluntary contraction (MVC) for each person was quanti ed by a dynamic handgrip trainer. The handgrip ring, closest to 50% of maximal voluntary contraction (MVC), was chosen for DHE which was performed at a frequency of 80/min, acoustically indicated by a metronome beat over CMR voice communication. In case of premature physical exhaustion, subjects were asked to indicate that by pressing the alert bell and fast Strain-ENcoded magnetic resonance imaging (fSENC) acquisition was immediately initiated. An adequate execution of DHE was supervised by the attending technician via visual control and an adequate heart rate (HR) response, which was controlled continuously by electrocardiogram (ECG) monitoring. DHE was stated as insu cient when hand movement rate continuously dropped below 80/min. Shortly before nishing the two minutes of handgrip exercise, the subjects were advised to hold their breath after expiration to start fSENC sequence manually.

Image analysis
Analysis of ventricular volumes, LV myocardial mass and LV ejection fraction (LVEF) were derived from short-and long axis slices on commercially available workstations (IntelliSpace Portal®, Philips Healthcare) and a dedicated post-processing software (cvi 42 ™ v5.5, Circle Cardiovascular Imaging, Calgary, Canada) from CMR-experienced physicians. Dobutamine stress was analysed according to current CMR interpretation guidelines [17].
For the interpretation of fSENC sequences and measurements of longitudinal strain at rest, after DHE and during dobutamine stress, a dedicated software (MyoStrain 5.2.1 Myocardial Solutions, Inc., Morrisville, North Carolina, USA) was used ( Figure 2): endo-and epicardial borders were drawn manually at endsystole for each short-axis slice resulting in segmental and global longitudinal strain values. The examiners underwent speci c training for the MyoStrain® software. For intra-and interobserver variability, 10 randomly selected subjects were analyzed twice. As reported before [18,19], GLS response after DHE was classi ed as stable (ΔGLS ≥ -0.5% and ≤ 0.5%), increase (ΔGLS < -0.5%) and decrease (ΔGLS > 0.5%).

Statistical methods
A dedicated software, MedCalc™ v17.7.2 (MedCalc software, Mariakerke, Belgium) was used for statistical analysis. Normal distribution was assessed using Shapiro-Wilk test. Continuous parameters were expressed as mean ± standard deviation for parametric and as median with interquartile range (IQR) for nonparametric variables. For the comparison of continuous variables between two groups, Student's ttest and Mann Whitney U test were used as applicable. Not normal distributed continuous variables were tested for differences using the nonparametric Wilcoxon test. The intra-and interobserver variability was described using the intra-class correlation coe cient (ICC with 95% CI) with a two-way random model with absolute agreement. A p-value of < 0.05 was regarded as statistically signi cant.

Results
Study population. Eighty middle-aged individuals including two subgroups of healthy individuals (i) and CAD patients (ii) (mean age 56 ± 17 years; 48 males) were enrolled ( Table 1). All subjects were in sinus rhythm. CMR revealed regular biventricular function and morphological dimensions in all individuals (LVEF 62 ± 6%, LVEDV 145 ± 33ml, LV mass 103 ± 28g). Fifty-two subjects (65%) underwent the study examination on 1.5 Tesla MR scanner, the others on a 3 Tesla clinical MR scanner. Mean systolic blood pressure at rest was 124 ± 10mmHg, diastolic blood pressure at rest was 78 ± 7mmHg.
Dynamic handgrip exercise. Seventy-four persons (93%) fully completed DHE, a minor portion of six persons (7%) had to abort prematurely due to peripheral fatigue ( Table 1). The lightest handgrip ring resistance (30lb) was used by the majority of subjects (74%), whereas 25% relied on medium resistance (50 lb) and a single person utilized the highest resistance (70lb).
Mean resting HR was 68 ± 10 bpm. DHE induced a signi cant increase of HR to 91 ± 13 bpm (p < 0.001, Figure 4A). Figure 3 shows a representative course of HR during DHE in a healthy subject. HR increased steadily as DHE progressed. A HR plateau was not evident after two minutes of DHE. After the end of DHE, HR fell rapidly towards the resting HR.
GLS rest was -19.4 ± 1.9%. GLS signi cantly increased to -20.6 ± 2.1% (p < 0.001) following DHE, according to a relative increase of 7 ± 7%. The majority of our study population (70%) responded with a relevant increase of GLS (DGLS < -0.5%) on a paired comparison, whereas GLS remained unchanged in 25% of individuals and decreased in 5% ( Figure 4B). On a segmental level ( Figure 4C), DHE induced a signi cant increase of LS in every segment of apical and midventricular layer (p < 0.01). However, in most basal segments no signi cant changes of LS could be observed.
Divided at the median age (53.7 years), two age-dependent subgroups were created: younger (n=25, mean age = 36.8 ± 10.4 years) and older adults (n=25, mean age = 65 ± 7 years). Except HR DHE (young HR DHE = 96 ± 12bpm vs. old HR DHE = 88 ± 12bpm; p < 0.05), no signi cant differences were observed between younger and older adults related to DHE study. Neither age, gender nor handgrip ring strength were signi cant confounders for heart rate or GLS stress response.

Discussion
In our study, we examined the feasibility of DHE and its hemodynamic effects. We found a positive chrono-and inotropic response to DHE expressed by a signi cant increase of GLS and heart rate after DHE. The effect strength of DHE is non-inferior to intermediate dobutamine-stimulation. Furthermore, the reasonable low DHE abortion rate supports the practicability of the approach.
As a potential alternative to conventional stress testing, exercise-CMR allows for a needle-free protocol without pharmacological side effects. Due to its simple and favorable handling, DHE represents a costand timesaving physiologic stressor. Although, the hemodynamic impact of DHE doesn't achieve standard target heart rate criteria (≥85% of maximum heart rate), strain imaging had demonstrated in the past to allow the detection of ischemia at intermediate dobutamine-stimulation with high accuracy using Strain-ENCoded CMR [20]. As shown in our subgroup of CAD patients, DHE achieved a comparable or slightly higher increase of HR and longitudinal strain than intermediate dobutamine stress. Due to its similar chronotropic and inotropic effects, DHE represents a promising, needle-free stressor to induce ischemia. Future, prospective trials in CAD patients should determine the potential of DHE-fSENC for the detection of obstructive CAD.
In order to detect segmental stress-induced functional impairment, fSENC appears to ful ll the prerequisites of a fast and reliable strain assessment unlike other techniques as feature tracking known to struggle with segmental reproducibility or myocardial tagging that appears time consuming in both preparation and acquisition [15,18]. As shown recently by our group, measurements of segmental longitudinal strain using fSENC allowed for the detection of ischemia related wall motion abnormalities after hyperventilation/breath-hold maneuver and during adenosine stress [21] .
Remarkably, our data suggest a higher chronotropic effect using repetitive, two-handed exercise compared to other studies using different variations of handgrip exercise [19]. In contrast, previous studies had evaluated isometric one-handed handgrip exercise protocols with a broad range of exercise duration, handgrip application and devices hampering a general comparability [19,[22][23][24]. However, several comparative studies investigated the differences between dynamic and isometric (static) handgrip exercise [25][26][27][28]. Although most authors found no signi cant differences in hemodynamic response, Stebbins et al. observed a signi cantly higher increase of heart rate, blood pressure and cardiac output with increasing handgrip strengths compared to isometric handgrip protocols. We found in comparison to intermediate dobutamine stimulation, DHE to achieve even higher heart rates re ecting a good chronotropic response. Beyond an isolated increase of afterload as observed in isometric handgrip exercise, in DHE this may be attributed more to an increased production and accumulation of muscle metabolites leading to a higher exercise pressor re ex and greater activation of muscle mechanoreceptors [28].
Regarding the GLS response after handgrip exercise, several authors made different observations [19,23,29,30]. Most recently, Blum et al. examined the response of GLS after an isometric handgrip exercise using fSENC-CMR, as we did [19]. They assumed a limited diagnostic purpose for strain imaging after isometric handgrip exercise due to a heterogenous response of GLS. To our knowledge, data on the GLS response after DHE do not exist so far. In our study population, GLS signi cantly increased after DHE. The vast majority of 56 subjects (70.0%) had a relevant increase of < -0.5%. Even on a segmental level, LS increased signi cantly in most segments, which is important for its potential future use in detecting regional impairment in ischemic myocardium. In comparison to isometric handgrip exercise, our results are more homogenous and LS increased signi cantly in the vast majority of patients suggesting that DHE could be a more suitable protocol for future use with fSENC-CMR.
Limitations. In the present study we did not investigate varying durations of DHE and their hemodynamic impact. A prolongation of the stress appears of a potential bene t if tolerated by the patient, as heart rates did not reach a plateau level. Moreover, continuous heart rate and blood pressure monitoring was not available during stress testing. The non-invasive blood pressure monitoring with a common upper arm cuff did not allow for reliable measurements during the repetitive contractions. Peak exercise blood pressure would have allowed to better understand the development of left ventricular afterload.

Conclusions
DHE causes a positive inotropic and chronotropic effect comparable or slightly higher than intermediate dobutamine stress, suggesting a relevant increase of myocardial oxygen demand. In this rather small cohort, DHE appears safe with broad applicability. Even minor differences of LS can be detected fast and reliable using CMR-fSENC. Further studies which investigate the differences of isometric and dynamic handgrip exercise, as well as the in uence of one-and two-handed approaches in regard to the response of LS are needed. Nevertheless, the data encourages further studies to determine its potential to detect obstructive CAD as a potential needle-free CMR stress test.

Declarations
Funding: The study received no funding.
Con icts of interest/Competing interests: Nael Osman is the CSO and founder of Myocardial Solutions, provider of MyoStrain® software for the analysis of fSENC sequences. Christian Stehning is an employee at Philips Healthcare. All other authors declare no con ict of interest, esp. no relationships with industry according to journal policy.
Availability of data and material: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.    Example of the HR response during DHE of a 29-year-old, male subject. HR increased steadily as DHE progressed. After the end of DHE (120s), HR fell rapidly towards the resting HR within 10 seconds. DHE= dynamic handgrip exercise; HR= heart rate. a. Heart response in subgroup II after DHE (green) and during dobutamine stress (red