Biomedical subjects
Publications and source records attributed to Yen-Bin Liu.
We present a patient with hypertrophic cardiomyopathy and a large scar in the apex demonstrated on 201Tl SPECT and 18F-FDG PET images. Subsequent coronary angiography revealed no significant coronary artery stenosis but a myocardial bridge with severe systolic milking in the middle segment of the left anterior descending artery.
BACKGROUND: Permanent pacing is the treatment for chronotropically incompetent hearts. However, the right ventricular (RV) apical pacing-induced asynchrony, even maintaining the atrioventricular (AV) sequential activation, has depressed left ventricular contractility. Whether RV septal pacing would less compromise the electromechanical performance of the left ventricle and the chronotropic effect on myocardial contractility, is unknown. METHODS: We prospectively studied 42 patients without structural heart diseases and with symptomatic bradycardia. There were 10 patients receiving atrial pacing (AAI) pacemakers, 18 patients having AV sequential pacing at RV apex (DDDapx) and 14 patients being AV sequentially paced at septum (DDDspt). Echocardiography was performed before and within 72 h after the pacemaker implantation. The ventricular mechanical performance and asynchrony was compared in conditions of programmed rates of 60, 80 and 100/min. RESULTS: Myocardial performance index was significantly better in DDDspt than in DDDapx patients (p=0.003). With faster programmed rate, the QRS/RR increased (p<0.05) in DDDapx patients with more inter- and intraventricular asynchrony, implicating the disadvantage of prolonged depolarization time. The DDDspt group demonstrated comparable parameters of diastolic function to AAI patients and preserved mechanical performance during accelerated pacing. CONCLUSIONS: RV septal pacing showed the advantages of shorter depolarization time, less ventricular contractile asynchrony, better mechanical performance and preserved chronotropic response on myocardial contractility in comparison with apical pacing.
BACKGROUND: Ventricular tachyarrhythmia developing in the acute stage of myocardial infarction (MI) is an important cause of sudden cardiac death. The aim of this study was to determine whether dyslipidemia is associated with the occurrence of ventricular tachycardia/fibrillation (VT/VF) during the acute stage of ST-segment elevation MI (STEMI). METHODS: A total of 58 patients experiencing VT/VF within 24 hours after the onset of MI were selected as the study group. A group of 58 patients with MI but without VT/VF was selected as the control group matched for sex (overall, 104 males), age (overall, 58 +/- 10 years), and the use of thrombolytic therapy (n = 82). The lipid profiles including total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), and triglyceride were measured during the first week and at the third month after the index MI. Other coronary risk factors, and clinical, hemodynamic and angiographic characteristics were also included in the assessment. RESULTS: During the acute stage, patients with VT/VF had higher levels of LDL-C and lower blood pressure on initial arrival at our hospital. At the 3-month follow-up, those patients with VT/VF showed higher levels of TC, LDL-C and triglyceride. Multivariate analysis revealed that LDL-C (p < 0.001) at the 3-month follow-up, mean blood pressure on arrival (p < 0.01), and the difference in triglyceride levels between the first week and the third month (p < 0.05) were independent predictors for the occurrence of VT/VF in the acute stage of MI. CONCLUSION: This study suggests that dyslipidemia imposes a higher risk of developing tachyarrhythmia in the acute phase of STEMI.
Mutations in cardiac potassium and sodium channel genes are responsible for several hereditary cardiac arrhythmia syndromes. We established a denaturing high-performance liquid chromatography (DHPLC) protocol for rapid mutation screening of these genes, and reported mutations and variations identified by this method. We included 28 patients with Brugada syndrome, 4 with congenital long QT syndrome (LQTS), 11 with drug-induced LQTS, 4 with idiopathic ventricular fibrillation, and 50 normal volunteers. Polymerase chain reactions were performed to amplify the entire coding region of these genes. DHPLC was used to screen for heteroduplexes then DNA sequencing was performed. With this method, we identified the mutation(s) in all four patients with congenital LQTS (KCNQ1 A341V, KCNH2 N633D, KCNH2 2768Cdel and KCNE1 K70 N Y81C double mutations). We also identified the SCN5A A551T mutation in 1 of the 28 patients with Brugada syndrome. All the above-mentioned mutations were novel except KCNQ1 A341V. No mutations were identified in patients with drug-induced LQTS or idiopathic ventricular fibrillation. In total, 25 single nucleotide polymorphisms were identified, 10 of which were novel. In conclusion, DHPLC is a sensitive and rapid method for detection of cardiac sodium and potassium channel gene mutations.
BACKGROUND: Despite the maintenance of atrioventricular (AV) synchrony, the detrimental effect of left ventricular asynchronization on mechanical performance and intraventricular flow by nonphysiologic right ventricular apical pacing in dual-chamber pacing, with and without rate adaptation, is not clear. METHOD: Twenty-seven consecutive patients receiving permanent pacemakers for symptomatic bradyarrhythmias (18 with DDD and 9 with AAI mode pacemakers) were evaluated with standard and tissue Doppler echocardiography before and 24 h after pacemaker implantation. The rate-response effect of pacing was studied by programmed rate with increments of 20, from 60 to 100/min. RESULTS: Color M-mode echocardiography demonstrated that much more DDD patients developed new biphasic intraventricular flow during isovolumic relaxation period than AAI patients (13/18 versus 0/9, P<0.001). In DDD patients, the ventricular relaxation represented by mitral annulus velocity in early diastole significantly attenuated (before vs. after DDDR, 8.5+/-2.8 vs. 5.2+/-1.2 cm/s, P<0.05), and also the mitral flow propagation velocity (33+/-11 vs. 25+/-5 cm/s, P<0.01). The myocardial performance index increased after DDD (0.70+/-0.15 vs. 0.79+/-0.24, P<0.05) but not after AAI (0.61+/-0.1 vs. 0.59+/-0.08, P=NS). For both pacing groups, the accelerated pacing rate prolonged the isovolumic relaxation time and shortened the diastole period (P<0.001). However, only DDD patients had a decreased mitral flow propagation velocity (P=0.026) and an attenuated force-frequency relation in programmed rate acceleration. CONCLUSION: Despite the AV synchrony, right ventricular apical pacing immediately attenuates the left ventricular contraction and relaxation performance, which deteriorated further and suppressed the physiologically positive force-frequency relation after accelerated pacing rate.
The effects of hypercholesterolemia on the myocardium per se include electrophysiological and mechanical alterations. Since gap junctions are essential in electromechanical coupling throughout the heart, we examined the correlation between the temporal expression of cardiac connexin 43 (Cx43), contractile function, and conduction velocity in cholesterol-fed rabbits. After a 12-week feeding period, serum cholesterol levels gradually increased (P<0.001). In contrast, expression of cardiomyocyte Cx43 protein progressively decreased (60% reduction at 12 weeks, P<0.001). Such a reduction was also demonstrated by immunoconfocal microscopy, which further showed redistribution of Cx43 gap junctions at the lateral cell membrane. The downregulation of Cx43 protein was associated with increased levels of Cx43 mRNA (3.5 -fold at 12 weeks, P<0.001) and phosphorylated c-Jun N-terminal kinase (three-fold at 12 weeks, P=0.001). Functionally, although fractional shortening of the left ventricle remained unchanged throughout the feeding protocol, the cholesterol-fed rabbits had a reduced cardiac cycle-dependent variation of integrated backscatters, a decreased mitral ring systolic velocity, and an increased modified Tei index (all P<0.001), all of which indicated impaired intrinsic myocardial contractility and attenuated ventricular systolic performance. In Langendorff-perfused hearts of cholesterol-fed rabbits, decreased conduction velocity was observed (P<0.005). Withdrawal of the cholesterol-enriched diet for 18 weeks restored the contractile parameters and Cx43 protein expression. These findings suggest that Cx43 is highly involved in the molecular mechanism of hypercholesterolemia-induced cardiac contractile dysfunction and dysrhythmias.
OBJECTIVES: The purpose of this study was to test the hypothesis that radiofrequency (RF) catheter ablation results in cardiac nerve sprouting. BACKGROUND: Nerve sprouting plays a role in cardiac arrhythmogenesis. Whether or not nerve sprouting occurs after RF catheter ablation is unclear. METHODS: We performed RF catheter ablation in the right atrium (RA) and right ventricle (RV) in 10 dogs, which then were sacrificed in 2 hours (acute group, n = 5) or 1 month (chronic group, n = 5). Seven normal dogs were used as control. Immunohistochemical staining for growth-associated protein 43 (GAP-43) was performed to measure growing (sprouting) nerves. RESULTS: A significant increase of GAP-43 immunoreactive nerve fiber density was observed at the RA ablation sites in 2 hours (4,410 +/- 1,379 microm(2)/mm(2)) and in 1 month (2,948 +/- 666 microm(2)/mm(2)) after ablation compared to controls (1,377 +/- 471 microm(2)/mm(2), P = .0001). At remote sites (>2 cm away from ablation sites) of RA, RF ablation also resulted in robust nerve sprouting in both the acute group (5,846 +/- 3241 microm(2)/mm(2)) and the chronic group (6,030 +/- 2226 microm(2)/mm(2)). RF ablation in the RV did not increase nerve density at the ablation sites, but nerve density was increased at remote sites in 2 hours (1,345 +/- 451 microm(2)/mm(2), P = .0136) that was reduced down to the normal control level (722 +/- 337 microm(2)/mm(2)) in 1 month. CONCLUSIONS: Nerve sprouting occurred within 2 hours after RF ablation in both the RA and RV and persisted for at least 1 month in the RA but not the RV. The increased GAP-43(+) nerve densities developed at both the ablation and the remote sites.
INTRODUCTION: We previously demonstrated that wavefront synchronization by spatiotemporal excitable gap pacing (Sync P) is effective at facilitating spontaneous termination of ventricular fibrillation (VF). Therefore, we hypothesized that a spatiotemporally controlled defibrillation (STCD) strategy using defibrillation shocks preceded by Sync P can improve defibrillation efficacy. METHOD AND RESULTS: We explored the STCD effects in 13 isolated rabbit hearts. During VF, a low-voltage gradient (LVG) area was synchronized by Sync P for 0.92 second. For Sync P, optical action potentials (OAPs) adjacent to four pacing electrodes (10 mm apart) were monitored. When one of the electrodes was in the excitable gap, a 5-mA current was administered from all electrodes. A shock was delivered 23 ms after the excitable gap when the LVG area was unexcitable. The effects of STCD was compared to random shocks (C) by evaluating the defibrillation threshold 50% (DFT(50); n = 35 for each) and preshock coupling intervals (n = 208 for STCD, n = 172 for C). Results were as follows. (1) Sync P caused wavefront synchronization as indicated by a decreased number of phase singularity points (P < 0.0001) and reduced spatial dispersion of VF cycle length (P < 0.01). (2) STCD decreased DFT(50) by 10.3% (P < 0.05). (3) The successful shocks showed shorter preshock coupling intervals (CI; P < 0.05) and a higher proportion of unexcitable shock at the LVG area (P < 0.001) than failed shocks. STCD showed shorter CIs (P < 0.05) and a higher unexcitable shock rate at LVG area (P < 0.05) than C. CONCLUSION: STCD improves defibrillation efficacy by synchronizing VF activations and increasing probability of shock delivery to the unexcitable LVG area.
INTRODUCTION: We previously reported that a normal ventricle can demonstrate two types of ventricular fibrillation (VF), depending on the underlying electrophysiologic characteristics at the time of VF induction. We hypothesize that the two types of VF can coexist in acutely ischemic ventricles. METHODS AND RESULTS: Optical mapping studies were performed with di-4ANEPPS in 15 Langendorff-perfused rabbit hearts. Coronary artery branches were ligated to create regional ischemia in 10 hearts. Action potential duration measured to 50% repolarization (APD50) during ischemia showed an area with uniformly shortened APD50 (zone 1), an area with normal or lengthened APD50 (zone 3), and an area in between with an APD50 gradient (zone 2). Ischemia flattened APD restitution (APDR) slope and reduced conduction velocity in zone 1, creating a condition for type II VF. APDR steepened and the conduction velocity changed little in the nonischemic zone (zone 3), creating a condition for type I VF. During induced VF, the dominant frequency in zones 2 and 3 progressively increased after ischemia onset. The dominant frequency in zone 1 (ischemic zone) first decreased and then slightly increased but typically remained less than the dominant frequency in zone 3. The number of wavebreaks increased with time in all three zones (baseline: 4.3 +/- 1.5; 30 min: 11.7 +/- 5.6; 60 min: 15.6 +/- 11 per frame; P < 0.01). CONCLUSION: Two types of VF can coexist during acute regional ischemia. Both ischemic and nonischemic regions develop proarrhythmic changes during regional ischemia, thus contributing to increased ventricular vulnerability to VF and sudden death during acute coronary occlusion.
BACKGROUND: Arrhythmia and sudden death represent striking features in patients with end-stage renal disease (ESRD). Increased QT dispersion has been shown to be associated with arrhythmias. Abnormal excitability and heterogeneous cardiac iron deposition may cause the arrhythmogenesis of human siderotic heart disease. Iron overload and precipitation with its toxicity in cardiac muscles may, therefore, cause QT prolongation in dialysis patients. METHODS: A total of 102 (65 women, 37 men; mean age, 47.7 +/- 13.4 years) nondiabetic patients undergoing peritoneal dialysis (PD) were enrolled in this study. Another 102 subjects with a serum creatinine level less than 1.5 mg/dL (133 micromol/L) were used as matched control subjects. The PD patients were divided into 2 groups according to whether their computerized measurements of corrected QT (QTc) dispersion were longer than 74 ms. A value of 74 ms has been associated with risk for serious arrhythmias related to sudden death in dialysis patients. RESULTS: The QTc dispersion of PD patients was significantly longer than that of the control subjects (69.8 +/- 40.0 versus 55.2 +/- 33.6 ms; P < 0.01). Thirty-eight PD patients with QTc dispersion longer than 74 ms had lower blood pressure ( P = 0.01), fewer left ventricle masses ( P = 0.036), and lower serum albumin levels (P = 0.046) but higher levels of serum calcium (P = 0.038) and transferrin saturation (TSAT; P = 0.022) than the other patients. Multivariate analysis identified TSAT as an independent factor for QTc dispersion (r = 0.432, P < 0.001). A linear relationship showed that at 74 ms of QTc dispersion, TSAT was 35.2%. CONCLUSION: Long-term PD patients have longer QTc dispersion than subjects with normal renal function. The high body iron stores in these patients increase the risk of increased QT dispersion. The concern over iron overload in dialysis patients is not only because of its oxidative toxicity, but also its precipitation of arrhythmias, which may be measured by the surrogate marker of QTc dispersion.
BACKGROUND: A therapeutic implication of the focal-source hypothesis of ventricular fibrillation (VF) is that VF can be terminated by focal ablation. We hypothesize that beta-adrenergic receptor blockade converts multiple-wavelet VF to focal-source VF and that this focal source is located near the papillary muscle (PM). METHODS AND RESULTS: We used optical mapping techniques to study the effects of propranolol (0.3 mg/L) on VF dynamics in Langendorff-perfused rabbit hearts. The left ventricular (LV) anterior wall was mapped and optical action potential duration restitution (APDR) was determined at 25 epicardial sites. We performed ablation during VF of the left anterior PM in hearts with (N=6) or without (N=6) cytochalasin infusion, the LV lateral epicardium (Epi group, N=3), and the LV endocardium (Endo group, N=3). The PM was also ablated in 3 hearts without propranolol (control group). Propranolol converted multiple-wavelet VF to slow VF with reentry localized to the PM. Propranolol decreased the maximal slope of the APDR curve (P<0.001) as well as its spatial heterogeneity (P<0.01) and conduction velocity (P=0.01) while increasing the VF cycle length (P<0.001). PM ablation terminated VF during propropranolol infusion with (6 of 6, 100%) or without (4 of 6, 67%) cytochalasin D and significantly reduced inducibility. VF did not terminate in the Epi, Endo, and control groups (P<0.001). CONCLUSIONS: Propranolol flattens the APDR curve and reduces conduction velocity, converting multiple-wavelet VF into VF with a focal source anchored to the PM. Ablation of this focal source may terminate VF.
Wavefront synchronization is an important aspect preceding the termination of ventricular fibrillation (VF). We evaluated the defibrillation efficacy of a novel multisite pacing algorithm using optical recording-guided synchronized pacing (SyncP) in the excitable gaps. We compared the effects of SyncP with traditional overdrive pacing (ODP) at 90% of the VF cycle length (VFCL) and high-frequency pacing (HFP; 43-215 Hz) on spontaneous VF termination in isolated rabbit hearts. For SyncP, the pacing current was triggered by the activation of a reference site and was delivered when the optical potential of the pacing site was in an excitable gap. We measured VFCL and the spatial dispersion of VFCL (SDCL) from five points (3 points in the paced area and 2 points in the nonpaced area) and the distribution of phase singularities during the prepacing, pacing, and postpacing periods. The results showed that 1) the VF termination rate of SyncP (16.0%, n = 106) was higher than that of ODP (2.1%, n = 48, P < 0.01) or HFP (1.6%, n = 129, P < 0.0001); 2) energy consumption for SyncP (7.6 +/- 9.3 mJ) was significantly lower than that of ODP (14.0 +/- 14.8 mJ, P < 0.0001); and 3) SyncP, but not ODP or HFP, decreased SDCL in the paced area during the pacing (P < 0.01) and postpacing (P < 0.05) periods compared with the prepacing period. We conclude that SyncP is effective in inducing wavefront synchronization and is more effective at facilitating spontaneous VF termination than non-SyncP.
Whether hypercholesterolemia (HC) can induce proarrhythmic neural and electrophysiological remodeling is unclear. We fed rabbits with either high cholesterol (HC, n=10) or standard (S, n=10) chows for 12 weeks (protocol 1), and with HC (n=12) or S (n=10) chows for 8 weeks (protocol 2). In protocol 3, 10 rabbits were fed with various protocols to observe the effects of different serum cholesterol levels. Results showed that the serum cholesterol levels were 2097+/-288 mg/dL in HC group and 59+/-9 mg/dL in S group for protocol 1 and were 1889+/-577 mg/dL in HC group and 50+/-21 mg/dL in S group for protocol 2. Density of growth-associated protein 43- (GAP43) and tyrosine hydroxylase- (TH) positive nerves in the heart was significantly higher in HC than S in protocol 1. Compared with S, HC rabbits had longer QTc intervals, more QTc dispersion, longer action potential duration, increased heterogeneity of repolarization and higher peak calcium current (ICa) density (14.0+/-3.1 versus 9.1+/-3.4 pA/pF; P<0.01) in protocol 1 and 2. Ventricular fibrillation was either induced or occurred spontaneously in 9/12 of hearts of HC group and 2/10 of hearts in S group in protocol 2. Protocol 3 showed a strong correlation between serum cholesterol level and nerve density for GAP43 (R2=0.94; P<0.001) and TH (R2=0.91; P<0.001). We conclude that HC resulted in nerve sprouting, sympathetic hyperinnervation, and increased ICa. The neural and electrophysiological remodeling was associated with prolonged action potential duration, longer QTc intervals, increased repolarization dispersion, and increased ventricular vulnerability to fibrillation.
UNLABELLED: Phase Singularity and Wavebreak. INTRODUCTION: Phase maps and the detection of phase singularities (PSs) have become a well-developed method for characterizing the organization of ventricular fibrillation (VF). How precisely PS colocalizes with wavebreak (WB) during VF, however, is unknown. METHODS AND RESULTS: We performed optical mapping of 27 episodes of VF in nine Langendorff-perfused rabbit hearts. A WB is a point where the activation wavefront and the repolarization waveback meet. A PS is a site where its phase is ambiguous and its neighboring pixels exhibit a continuous phase progression from -pi to +pi. The correlation coefficient between the number of WBs and PSs was 0.78 +/- 0.09 for each heart and 0.81 for all VF episodes (P < 0.001), indicating a significant temporal correlation. We then superimposed the WBs and PSs for every 100 frames of each episode. These maps showed a high degree of spatial colocalization. To quantify spatial colocalization, the spatial shifts between the cumulative maps of WBs and PSs in corresponding frames were calculated by automatic alignment to obtain maximum overlap between these two maps. The spatial shifts were 0.04 +/- 0.31 mm on the x-axis and 0.06 +/- 0.27 mm on the y-axis over a 20 x 20 mm2 mapped field, indicating highly significant spatial correlation. CONCLUSION: Phase mapping provides a convenient and robust approach to quantitatively describe wave propagation and organization during VF. The close spatiotemporal correlation between PSs and WBs establishes that PSs are a valid alternate representation of WB during VF and further validated the use of phase mapping in the study of VF dynamics.
Although hypercholesterolemia is a well-known risk factor for atherosclerosis, little is known about the effect of hypercholesterolemia on cardiac contractile function. The objective of this study was to examine the effect of hypercholesterolemia on myocardial contractility. Fifteen New Zealand white rabbits were fed standard chow (control group) and another 15 were fed a cholesterol-enriched diet (HC group) for 12 weeks. The contractile response of ventricular muscle strips was measured in various extracellular calcium concentrations and at different pacing rates. The whole-cell calcium current recording, and mRNA and protein levels of cellular calcium-handling proteins were also analyzed. With 2 mM Ca2+ and stimulation at 3 Hz, the contractile force of HC strips was less than that of the controls (3.63 +/- 0.20 vs. 4.61 +/- 0.50 mN, p < 0.05). The time to peak tension was longer for HC strips (93.3 +/- 2.16 vs. 82.2 +/- 2.81 ms, p < 0.05). The peak L-type calcium inward current density was slightly higher in HC myocytes but did not reach statistical significance (-14.90 +/- 0.94 vs. -12.44 +/- 0.84 pA/pF, p = 0.15). The mRNA level of sarcoplasmic reticulum Ca2+-ATPase (SERCA), normalized to GAPDH, was significantly lower in the HC than that in the control group (2.85 +/- 0.14 vs. 7.67 +/- 0.67, p < 0.05), as was the ryanodine receptor (RyR; 0.42 +/- 0.06 vs. 0.71 +/- 0.13, p < 0.05). The mRNA of the Na+/Ca2+ exchanger (NCX) was statistically higher in the HC group (0.90 +/- 0.12 vs. 0.48 +/- 0.05, p < 0.05). Western blot experiments revealed that protein expression of SERCA in the HC strips decreased, but that of the NCX increased. The protein expression of the dihydropyridine receptor was similar between these two groups. We concluded that hypercholesterolemia results in suppression of the maximal contractile function and in a longer systolic contractile time course. These changes may partially be mediated through a decrease in SERCA and RyR but an increase in NCX expression.
In order to assess the effect of hypercholesterolemia on cardiac sympathetic nerve function, New Zealand white rabbits were fed a normal diet (the control group) or one enriched with 0.5% cholesterol [the hypercholesterol (HC) group] for 3 months. Before and after the 3-month diet treatment, we performed serial imaging examinations and analyzed the uptake and washout ratio of (123)I-meta-iodobenzylguanidine ((123)I-MIBG) from the myocardium by administration of (123)I-MIBG through an ear vein. At the end of the experiments, the rabbits were sacrificed, and right ventricular strips were taken from their hearts. The inotropic response of the right ventricular strips to isoproterenol (ISO) and norepinephrine (NE) were evaluated. The cardiac MIBG uptake of the HC group, which was evaluated using the heart to mediastinum ratio, was higher than that of the age-matched control group. However, the washout ratios of (123)I-MIBG did not differ statistically between the two groups. On pretreatment with cocaine, NE-enhanced contractility was greater in papillary muscles isolated from the HC group. The concentration-response curve to ISO was shifted to the right in the HC group, compared with that in the control group. In conclusion, hypercholesterolemia in rabbits resulted in an increase in sympathetic nerve density in the myocardium, a decrease in the inotropic response to ISO and an increase in the inotropic response to NE in cocaine-treated myocardium. Both the in vivo and in vitro studies demonstrated the functional significance of neural remodeling induced by hypercholesterolemia.
Reactive oxygen species (ROS) are short-lived, highly reactive chemical entities that play significant roles in all levels of biology. However, their measurement requires destructive preparation, thereby limiting the continuous measurement of ROS in a living tissue. We develop an optical mapping system to visualize ROS production in an isolated and perfused rat heart. By staining the heart with dihydroethidium (DHE), a 532-nm laser beam is directed to the epicardial surface, where we collect the red fluorescence (>600 nm) for semiquantitative analysis. With this system, ROS production as well as ventricular pressure and ECG in isolated perfused rat hearts are monitored throughout the reperfusion of global ischemia. Ischemia would decrease myocardial ROS production, while reperfusion would immediately result in sustained ROS overproduction. Optical mapping would provide information regarding the spatial distribution and temporal evolution of myocardial ROS production, which would enhance knowledge of the role of free radicals in cardiovascular biology.