Stress related neural activity: potential mechanism contributing to benefit of physical activity on heart failure incidence

Background

We previously observed that: 1) physical activity (PA) decreases stress-related neural activity (SNA) in a dose-dependent manner and 2) resting SNA predicts incident heart failure (HF). Thus, we hypothesize that PA may decrease the risk of incident HF in part through its beneficial effect on SNA.

Methods

Self-reported PA data were acquired from Partners Biobank participants. Metabolic equivalent of task-minutes per week (MET*minutes/week) were calculated and participants were categorized according to their adherence to current PA recommendations (<500 and ≥500 MET-min/wk). In a subset with clinically indicated 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) imaging (n=744), SNA was measured as a ratio of amygdalar to ventromedial prefrontal cortical activity. Subsequent HF and baseline CVD risk factor data were obtained using International Classification of Diseases Codes. Kaplan-Meier and Cox regression analyses were used to evaluate the relationship between PA vs HF risk. Mediation analysis was used to test whether PA's impact on SNA significantly mediated the association between PA and HF.

Results

Among 50,363 participants (median age 59 years [IQR 42, 70], 40% male), 13,850 (11.6%) were diagnosed with HF over a median period of 3.1 years. PA within recommendations was associated with lower risk of subsequent HF (hazard ratio (HR) [95% confidence interval (CI)]: 0.849 [0.760, 0.948], p=0.004, Figure 1) in a model adjusted for CVD risk factors and history of myocardial infarction. As noted previously, higher SNA associated with higher risk of subsequent HF (HR [95% CI]: 1.34 [1.08, 1.66], p=0.007); and PA within guideline recommendations (vs.below) associated with lower SNA (Zscore ± SD: −0.14±0.92 vs 0.02±1, p=0.037). Moreover, mediation analysis shows that exercise reduces HF risk in part via reductions in SNA, accounting for 9% of the total relationship between exercise and HF risk (p<0.05, Figure 2).

Conclusion

PA associates with reductions in: A) stress-associated neural activity and B) subsequent HF risk. Moreover, the mechanism by which PA reduces HF risk may involve salutary effects on stress-related neural pathways.

Funding Acknowledgement

Type of funding sources: None.

Genetic sensitivity to stress modifies the relationship between socioeconomic status and major adverse cardiovascular events

Background

Lower socioeconomic status (SES) associates with major adverse cardiovascular events (MACE), in part through stress-related neural pathways that elicit inflammation. However, it is unknown whether genes that heighten stress sensitivity modify the association between lower SES and MACE.

Purpose

To assess whether genetic predisposition to stress sensitivity would modify the link between low SES and MACE

Methods

13,154 participants (median age 60 yrs, 41% male) from the Mass General Brigham Biobank were studied. A polygenic risk score for neuroticism (nPRS) was used as a measure of genetic predisposition to stress sensitivity (GSS). Using home addresses, SES was evaluated as median income and area deprivation index (ADI). Stress-related neural activity (SNA) was assessed (N=978) using validated FDG PET/CT imaging methods. MACE, and cardiovascular disease (CVD) risk factors were evaluated. Mediation analyses were employed.

Results

Over median (IQR) 4.9 (4.1–5.9) years of follow-up, 1,030 (7.8%) individuals had MACE. Lower SES (as low income, or alternatively as high ADI) associated with incident MACE among individuals with higher GSS (nPRS ≥ median) but not lower GSS (Fig. 1A and 1B). Similarly, lower SES associated with SNA among individuals with higher but not lower GSS (Fig. 1C). SNA mediated the relationship between income and MACE (P<0.05**) among those with higher GSS.

Conclusions

Genetic predisposition to stress sensitivity appears to heighten CVD risk associated with lower SES. This relationship may result from differential activation of stress-related neural pathways.

Funding Acknowledgement

Type of funding sources: None.

Anxiety and depression associate with heightened risk of deep venous thrombosis: mediation through neural pathways

Background

Anxiety disorders and depression associate with an increased risk of deep venous thrombosis (DVT). However, it is unclear if this association persists after robust adjustment for confounders. Further, the mechanism mediating this potential link remains unknown. Prior studies show that anxiety and depression associate with heightened stress-associated neural activity (notably in the amygdala: AmygA), which in turn promotes chronic inflammation, a driver of thrombosis syndromes.

Purpose

To evaluate whether the association between anxiety/depression and DVT risk: A) persists after robustly accounting for potential confounders, and B) is mediated by upregulated stress-associated neural activity (namely AmygA).

Methods

Data were obtained from the Mass General Brigham Biobank, which included detailed health information on 118 871 adult participants. A subset of 1520 study subjects underwent clinical 18F-fluorodeoxyglucose positron emission tomography imaging during the follow up period, from which AmygA was measured as the ratio of amygdalar to regulatory (ventromedial pre-frontal cortex) activity. International Classification of Disease (ICD) codes in the medical record were used to define anxiety disorders, depression, and lower extremity DVT. Individuals on anticoagulant therapies for atrial fibrillation prior to enrolment and/or imaging were excluded. Cox analysis were performed wherein patients who developed DVT prior to Biobank enrolment (2599 subjects) were excluded. Mediation analysis was used to examine the role of AmygA in mediating the link between anxiety/depression and DVT.

Results

The median age of the study population was 57 years [interquartile range (IQR) 28] and 58.4% were female. DVT occurred in 1231 participants (1.2%) over a median follow up period of 3.3 years (IQR 3.0). Cox regression analysis showed that anxiety disorders and depression were independent predictors of incident DVT after controlling for confounders (Table 1; p<0.0001 for all analyses). In the subset of 1383 participants who underwent brain imaging, anxiety disorders and depression associated with increased AmygA activity after controlling for risk factors including age, sex, Charlson index, hormone use and cancer history (standardized β [95% CI]: 0.124 [0.017–0.232], p=0.023 and 0.151 [0.041–0.260], p=0.007 respectively). Further, AmygA associated with DVT (odds ratio (OR) [95% CI]: 1.248 [1.064–1.465], p=0.007). Path analysis demonstrated that increased AmgyA mediated the effect of both anxiety and depression on DVT (log OR [95% CI]: 0.0256 (0.0009–0.0592), p<0.05 and 0.0314 (0.0033 to 0.0718), p<0.05 respectively, Figure 1).

Conclusion

Anxiety disorders and depression associate with an increased risk of DVT via a mechanism that includes heightened stress-related neurobiological activity. Future studies should evaluate whether modulating this neural pathway could reduce the incidence of DVT.

Funding Acknowledgement

Type of funding sources: None.

Bone marrow and splenic metabolic activity by 18F-FDG PET/CT are associated with noncalcified coronary burden and lipid-rich necrotic core in psoriasis

Background/Introduction

Psoriasis is an immune-mediated inflammatory skin condition with an increased risk of myocardial infarction (MI). Elevated bone marrow (BM) and splenic hematopoiesis occurs after MI. In stable patients without chronic inflammation, higher splenic hematopoiesis predicts major adverse cardiovascular events (MACE). Nevertheless, studies in humans investigating these relationships in states of chronic inflammation on coronary artery disease features associated with MACE are limited.

Purpose

To investigate the relationships between bone marrow and splenic metabolic activity by [18]-fluorodeoxyglucose (FDG) PET/CT and subclinical cardiovascular disease in psoriasis.

Methods

Healthy participants (N=30) and psoriasis participants (N=210) were age and sex matched. All participants underwent 18FDG PET/CT and CT angiography (Toshiba 320 slice). Coronary artery plaque characteristics were assessed using QAngio CT (Medis, The Netherlands) and lipid rich necrotic core (LRNC) was assessed using vascuCAP (Elucid Bioimaging, Boston, MA). For tissue metabolic activities target-to-background ratio (TBR) was calculated as the ratio of arterial and venous standardized uptake values (SUV).

Results

The psoriasis cohort was middle aged 49.2 (±SD 11.9) years and predominantly male (64%). Those with psoriasis vs. healthy participants had higher BM (1.58 (IQR 1.35–1.89) vs. 1.23 (IQR 1.14–1.35); p<0.001) and splenic (1.40 (IQR 1.21–1.66) vs.1.17 (IQR 1.11–1.26); p<0.001) metabolic activity. After adjustment for cardiovascular risk factors bone marrow metabolic activity was associated with total burden, non-calcified burden (NCB) and LRNC (β=0.36, β=0.39, β=0.26; all p<0.001) respectively. Similar findings were observed for splenic activity (β=0.33, β=0.36, β=0.36; all p<0.001). In ROC analysis, when comparing area under the curve, BM activity better incrementally predicted non-calcified burden and lipid rich necrotic core compared to splenic activity (Figure).

Conclusions

BM and splenic metabolic activity are increased in psoriasis. Both are associated with coronary artery disease but there was a slightly stronger association with BM activity compared to splenic activity, These findings warrant further study to understand immune mechanisms underlying these observations.

Funding Acknowledgement

Type of funding sources: Public grant(s) – National budget only. Main funding source(s): National Heart, Lung and Blood Institute Intramural Research Program in Bethesda, Maryland Figure 1. Median values of NCB and LRNC were used to convert these continuous variables into dichotomous variables such that values ≤ median were designated as 0 and values >median were designated 1. Bone marrow model compared to base model and splenic model added incremental value in predicting NCB (p<0.0001) and LRNC (p=0.0003). Base model: Framingham risk score, lipid treatment, biologic therapy, homeostasis model assessment as an index of insulin resistance (HOMA-IR), low density lipoprotein (LDL).

Amygdalar activity among patients with head and neck cancer predicts cerebrovascular events: a retrospective FDG PET/CT Cohort Study

Background

Radiation therapy for head and neck cancer (HNC) associates with an increase in cerebrovascular events. Amygdalar metabolic activity (AmygA), a measure of stress-associated neurobiological activity, associates with cardiovascular events in patients without cancer. In this study, we tested whether AmygA was associated with cerebrovascular events among patients treated with radiation therapy.

Methods

The associations between AmygA (at cancer staging) and subsequent cerebrovascular events as well as inflammatory biomarkers and arterial inflammation, were tested among consecutive patients with HNC treated with radiation therapy who underwent 18F-FDG-PET/CT.

Results

Among 240 patients (59±13 years; 30% female, 48% had stage 4 disease), there were 26 events over a median follow-up period of 3 years (IQR: 1.7–5.1). AmygA at cancer staging was associated with subsequent cerebrovascular events, an effect that remained significant after adjustment for age, sex, and cerebrovascular risk factors (adjusted hazard ratio 1.47, 95% CI 1.04–2.09, P=0.031). The association persisted when the analysis was restricted to patients with no prior history of CV disease and even among those with lower predicted CV risk (P<0.001). Higher AmygA at cancer staging also associated with serum inflammatory markers and arterial inflammation (P<0.05 each).

Conclusions

Amygdalar activity, a quantifiable biomarker of stress-associated brain activity, associated with subsequent cerebrovascular events among patients with cancer treated with radiation therapy.

Funding Acknowledgement

Type of funding source: None

Effect of macitentan on left ventricular (LV) function in pulmonary arterial hypertension (PAH): results from REPAIR

Introduction

PAH impacts right ventricular (RV) structure and function but also leads to changes in the LV due to RV/LV interaction and underfilling. REPAIR, the first PAH study to use a primary endpoint assessed by cardiac MRI (cMRI), reported that RV stroke volume (RVSV) increased by 12 mL and pulmonary vascular resistance (PVR) decreased by 38% from baseline (BL) to Week 26 with macitentan.

Purpose

To assess the effect of macitentan on LV function in patients with PAH.

Methods

REPAIR (NCT02310672) was a 52-week, multicentre, open-label, single-arm, phase 4 study assessing the effect of macitentan primarily on RV structure and function, determined by cMRI and right heart catheterisation. Macitentan 10 mg was initiated in treatment-naïve patients, in patients receiving stable background phosphodiesterase type-5 inhibitor (PDE5i) at BL, or in initial combination with PDE5i. Exploratory LV endpoints were assessed by cMRI at Weeks 26 and 52. Safety was assessed up to end of study treatment +30 days in all patients who received ≥1 dose of macitentan (N=87). Patients with BL and Week 26 assessments for both PVR and RVSV were included in the modified Full Analysis Set (mFAS; N=71).

Results

In the mFAS, 57 (80%) patients were female. At BL, median age was 45 years; median (Q1, Q3) six-minute walk distance was 395 (323, 483) m; 48%/51% of patients were WHO functional class II/III; 59% had idiopathic PAH. Compared to BL, at Weeks 26 and 52 there were significant changes in LV cMRI parameters (table). The most common AEs were peripheral oedema (22%), headache (21%) and dizziness (14%).

Conclusions

Macitentan led to improvements in LV mass, volume and function, including clinically-relevant increases in LV stroke volume, at both 26 and 52 weeks in patients with PAH. Safety was consistent with other macitentan clinical trial data.

Funding Acknowledgement

Type of funding source: Other. Main funding source(s): Actelion Pharmaceuticals Ltd

Detection of inflamed atherosclerotic lesions with diadenosine-5',5'''-P1,P4-tetraphosphate (Ap4A) and positron-emission tomography

Diadenosine-5',5'''-P(1),P(4)-tetraphosphate (Ap(4)A) and its analog P(2),P(3)-monochloromethylene diadenosine-5',5'''-P(1),P(4)-tetraphosphate (AppCHClppA) are competitive inhibitors of adenosine diphosphate-induced platelet aggregation, which plays a central role in arterial thrombosis and plaque formation. In this study, we evaluate the imaging capabilities of positron-emission tomography (PET) with P(2),P(3)-[(18)F]monofluoromethylene diadenosine-5',5'''-P(1),P(4)-tetraphosphate ([(18)F]AppCHFppA) to detect atherosclerotic lesions in male New Zealand White rabbits. Three to six months after balloon injury to the aorta, the rabbits were injected with [(18)F]AppCHFppA, and microPET imaging showed rapid accumulation of this radiopharmaceutical in the atherosclerotic abdominal aorta, with lesions clearly visible 30 min after injection. Computed tomographic images were coregistered with PET images to improve delineation of aortoiliac tracer activity. Plaque macrophage density, quantified by immunostaining with RAM11 against rabbit macrophages, correlated with PET measurements of [(18)F]AppCHFppA uptake (r = 0.87, P < 0.0001), whereas smooth-muscle cell density, quantified by immunostaining with 1A4 against smooth muscle actin, did not. Biodistribution studies of [(18)F]AppCHFppA in normal rats indicated typical adenosine dinucleotide behavior with insignificant myocardial uptake and fast kidney clearance. The accumulation of [(18)F]AppCHFppA in macrophage-rich atherosclerotic plaques can be quantified noninvasively with PET. Hence, [(18)F]AppCHFppA holds promise for the noninvasive characterization of vascular inflammation.

Evidence of reduced resting blood flow in viable myocardial regions with chronic asynergy

OBJECTIVES

We tested the hypothesis in patients (n = 24) with ischemic heart disease that chronic contractile dysfunction occurs in myocardial regions with true reduction in rest blood flow.

BACKGROUND

Whether viable myocardial regions with chronic contractile dysfunction have true reduction in rest myocardial blood flow is controversial.

METHODS

Positron emission tomography (PET) 13N-ammonia was used to measure myocardial blood flow in combination with 18F-fluorodeoxyglucose (18FDG) to assess myocardial viability. Viability also was assessed by dobutamine echo and recovery of function after coronary artery bypass grafting (CABG). Segments (n = 252) were selected based on PET measured reduced resting blood flow and rest asynergy on echo.

RESULTS

Regional myocardial viability was present in 20 of 23 patients by PET, 13 of 23 by dobutamine echo and 10 of 11 by postrevascularization criteria. Rest blood flow in normal regions was 1.14+/-0.52 ml/min/g and by definition exceeded (p < 0.005) that in both viable (0.48+/-0.15; n = 8 patients) and nonviable (0.45+/-0.14; n = 8 patients) regions (post-CABG criteria), which did not differ. Correction of rest myocardial blood flow in viable asynergic segments, only, for fibrosis and incomplete tracer recovery raised the level to 0.67+/-0.21 (p < 0.005 vs. normal). Finally, evidence of both stunning (rest asynergy with normal flow) and hibernation was present in 15 of 23 (65%) patients.

CONCLUSIONS

Reduced rest blood flow in viable myocardial regions with chronic asynergy is common and cannot be accounted for by partial volume effect. Thus, hypotheses concerning physiologic mechanisms underlying chronic contractile dysfunction should consider the role played by chronic reduction of basal myocardial blood flow.

Estradiol therapy combined with progesterone and endothelium-dependent vasodilation in postmenopausal women

BACKGROUND

Epidemiological studies indicate that estrogen replacement therapy decreases the risk of cardiovascular events in postmenopausal women. Estrogen may confer cardiovascular protection by improving endothelial function because it increases endothelium-dependent vasodilation. It is not known whether progesterone attenuates the beneficial effects of estrogen on endothelial function.

METHODS AND RESULTS

Seventeen postmenopausal women with mild hypercholesterolemia were enrolled in a placebo-controlled, crossover trial to evaluate the effect of transdermal estradiol, with and without vaginal micronized progesterone, on endothelium-dependent vasodilation in a peripheral conduit artery. Brachial artery diameter was measured with high-resolution B-mode ultrasonography. To assess endothelium-dependent vasodilation, brachial artery diameter was determined at baseline and after a flow stimulus induced by reactive hyperemia. To assess endothelium-independent vasodilation, brachial artery diameter was measured after administration of sublingual nitroglycerin. During estradiol therapy, reactive hyperemia caused an 11.1+/-1.0% change in brachial artery diameter compared with 4. 7+/-0.6% during placebo therapy (P<0.001). Progesterone did not significantly attenuate this improvement. During combined estrogen and progesterone therapy, flow-mediated vasodilation of the brachial artery was 9.6+/-0.8% (P=NS versus estradiol alone). Endothelium-independent vasodilation was not altered by estradiol therapy, either with or without progesterone, compared with placebo. There was a modest decrease in total and LDL cholesterol during treatment both with estradiol alone and when estradiol was combined with progesterone (all P<0.001 versus placebo). In a multivariate analysis that included serum estradiol, progesterone, total and LDL cholesterol concentrations, blood pressure, and heart rate, only the estradiol level was a significant predictor of endothelium-dependent vasodilation.

CONCLUSIONS

The addition of micronized progesterone does not attenuate the favorable effect of estradiol on endothelium-dependent vasodilation. The vasoprotective effect of hormone replacement therapy may extend beyond its beneficial actions on lipids.

Hyperhomocyst(e)inemia is associated with impaired endothelium-dependent vasodilation in humans

BACKGROUND

Hyperhomocyst(e)inemia is a risk factor for atherosclerosis and is prevalent in the elderly. The objective of this study was to determine whether hyperhomocyst(e)inemia is associated with impaired endothelium-dependent vasodilation in humans.

METHODS AND RESULTS

High-resolution vascular ultrasonography was used to study endothelium-dependent and -independent vasodilation in a nonatherosclerotic peripheral conduit artery of 26 elderly hyperhomocyst(e)inemic subjects and 15 age- and sex-matched subjects with normal homocysteine levels. Flow-mediated, endothelium-dependent (nitric oxide-mediated) vasodilation was assessed by measuring the percent change in brachial artery diameter during reactive hyperemia. Endothelium-independent vasodilation was assessed after the administration of 0.4 mg sublingual nitroglycerin. Endothelium-dependent vasodilation was significantly impaired in the hyperhomocyst(e)inemic subjects compared with control subjects (3.7 +/- 0.6% versus 8.1 +/- 1.2%; P = .004), whereas endothelium-independent vasodilation was not different between the two groups (10.1 +/- 1.6% versus 9.3 +/- 1.5%; P = NS). In a linear regression analysis with serum homocysteine concentration, folic acid, age, sex, cholesterol (serum total, LDL, or HDL cholesterol), mean arterial blood pressure, use of antihypertensive medication, and baseline brachial artery diameter included as covariates, serum homocysteine concentration emerged as the only significant predictor of flow-mediated vasodilation.

CONCLUSIONS

These data indicate that hyperhomocyst(e)inemia is associated with impaired endothelium-dependent vasodilation in humans and suggest that the bioavailability of nitric oxide is decreased in hyperhomocyst(e)inemic humans.