Tese de Doutoramento Medicina, Investigação Clínica 2022 Faculdade de Ciências Médicas, Universidade NOVA de Lisboa

Background:Hypertrophic cardiomyopathy (HCM) is defined by the presence of left ventricular(LV)hypertrophy that cannot be solely explained by an increased afterload. The pathophysiologic mechanisms of this disease include coronary microvascular dysfunction and ischemia. The microvascular dysfunction is multifactorial, including reduced capillary density and vascular remodelling, fibrosis, myocyte disarray, and extravascular compression. Chronic and recurrent myocardial ischemia leads to fibrosis, which may culminate in myocardial dysfunction. Despite the recognition of microvascular dysfunction in HCM, further investigation is needed to delineate the interrelationships between microvascular dysfunction,fibrosis and prognosis.Objectives:The present research aimed to study the coronary microvascular dysfunctionin HCM using a multimodality approach, in order to identify predictors of microvascular dysfunction and to assess the association between microvascular dysfunction withtissue abnormalities and clinical manifestations. Methods:This prospective study enrolled83adult patients with HCM,withoutobstructiveepicardial coronary artery disease, submitted to the assessment of coronary microcirculation by:Echocardiography to evaluatecoronary flow velocity reserve (CFVR)(normal cut-off value ≥2.0), during adenosine-induced hyperemia;CMRto assess the ischemic burden by perfusion imaging during regadenoson-induced hyperemia;Cardiac catheterization to determine index of microcirculatory resistance(IMR)(normal cut-off value ≤22.0) and coronaryflow reserve (CFR) (normal cut-off value ≥2.0), during adenosine-induced hyperemia.Echocardiographic protocol included the assessment of myocardial deformation by two-dimensional(2D)longitudinal strain and three-dimensional(3D)longitudinal, circumferential and radial strain, area strain, torsion and twist.CMR protocol also included parametric mapping(to assess native T1, extracellular volume (ECV) and T2), late gadolinium enhancement (LGE) and three-dimensional longitudinal, circumferential and radial strains analysis.12-lead electrocardiogram, 24 hours Holter recording and cardiopulmonary exercise testing (CPET) were performed to assess arrhythmias and functional capacity. Results:EchocardiographyEighty-three patients underwent echocardiographic study. Mean age 55.0(14.4)years,50(60%)patientswere male; 59(71%) had nonobstructive HCM.CFVRin the left anterior descending artery(LAD)was 1.81(0.49) and CFVRin the posterior descending(PD)was 1.73(0.55); CFVR LAD was <2.0in 49(59%)patients and CFVR PD was <2.0in 43(52%).Greater LV maximum wall thickness (MWT) (β-estimate: -0.040, 95%CI: -0.071;-0.010, p=0.010) and female gender (β-estimate: -0.379, 95%CI: -0.640;-0.118, p=0.005)were independently associated with impaired CFVR.Lower CFVR PD was associated with impaired global longitudinal strain (GLS)2D (β-estimate: -3.240,95%CI: -4.634;-1.846, p<0.001), GLS 3D (β-estimate:-2.559,95%CI:-3.932;-1.186, p<0.001) and area strain (β-estimate: -3.044, 95%CI: -5.373;-0.716, p=0.011).Lower values of CFVR PD were relatedtoworsemyocardialglobal work index(β-estimate:267.824,95%CI: 75.964;459.683,p=0.007);global constructive work(β-estimate:217.300,95%CI: 38.750;395.850,p=0.018) and global work efficiency(β-estimate:5.656,95%CI:2.229;9.084, p=0.002).Impaired CFVR LAD (β-estimate:2.826, 95%CI:0.913;4.739,p=0.004) and CFVR PD (β-estimate:2.801,95%CI:0.657;4.945,p=0.011) were found to be associated with lowertricuspid annular plane systolic excursion. Lowervalues of CFVRLAD (β-estimate:2.580, 95%CI:0.169;4.991,p=0.036)and CFVR PD (β-estimate:3.163, 95%CI:0.721;5.606,p=0.012)were associated with worsepeak oxygen uptakein CPET.Cardiovascular magnetic resonanceSeventy-five patients underwent CMR,meanage 54.6(14.8) years, 47(63%) males, 51(68%)patients had nonobstructive HCM, MWTwas 20.2(4.6)mm andLV ejection fraction 71.6(8.3)%. Perfusion defect in at least in one segment was noted in 68 (91%) patients and ischemic burden was 22.5(16.9)% of LV. Greater MWT was associated with the severity of ischemia (β-estimate:1.353,95%CI:0.182,2.523,p<0.024). Ischemic burden wasassociated with higher values of native T1 (β-estimate:9.018,95%CI:4.721,13.315,p<0.001). The association between ischemia and LGE was significant in subgroup analysis: MWT 15-20mm (β-estimate:1.941,95%CI:0.738,3.143,p=0.002), nonobstructive HCM (β-estimate:1.471, 95%CI:0.258,2.683,p=0.019), females (β-estimate:1.957,95%CI:0.423,3.492,p=0.015)and age <40 years (β-estimate:4.874,95%CI:1.155,8.594,p=0.016). Ischemia in ≥21% of LV was associated with LGE>15% (area under the curve0.766,sensitivity 0.724, specificity 0.659). Ischemia was also associated with atrial fibrillation/flutter (AF/AFL) (OR:1.481,95%CI:1.020,2.152,p=0.039), but no association was seen for nonsustained ventricular tachycardia. Ischemia was associated with shorter time to anaerobic threshold in CPET (β-estimate:-0.442,95%CI:-0.860,-0.023,p=0.039).Cardiac catheterizationFourteen patientsunderwent cardiac catheterizationwith a mean age of62.8(6.2)years, 8 (57.1%) males, 9 (64.3%) of whom had obstructive HCM. Among 4 patientswith an IMR >22.0, all had nonobstructive HCM and 2 had angina. CFR<2.0was reported in 8(57%)patients. Among 4 patients with IMR>22.0, perfusion defects were found in 2 of the 3 patients who underwent stress CMR.Increased ECV (>28%) was documented in2 of the patients with IMR>22 and in 3 of the patients with IMR≤22.0.LGE was >15% in 2 of the patients with IMR>22 and in 4 with IMR≤22.0. Conclusions: Coronary microvascular dysfunction isafrequentpathophysiological finding in HCM, and its evaluation has clinical relevance. In our cohort, greater MWT was linkedto depressed CFVR, and blunted CFVR associated with impaired biventricular systolic function and worse functional capacity.Ischemic burden, secondary to microvascular dysfunction, wasrelated to the severity of LV hypertrophy and impacts on various pathological and clinical features, includingtissue abnormalitiesandarrhythmic events.IMR assessment inHCMisfeasible and safe. Patients with abnormal IMR seemed to havemore significant tissue abnormalities in CMR.Our findings highlight the potential additional role of the evaluation ofcoronary microvascular dysfunction in patients with HCM,which may allow more accuraterisk stratificationfor arrhythmic events and progression to heart failure.