Author + information
- Received January 23, 2002
- Revision received May 29, 2002
- Accepted June 7, 2002
- Published online October 16, 2002.
- Ryosuke Nishio, MD, PhD*,
- Shigetake Sasayama, MD, PhD, FACC* and
- Akira Matsumori, MD, PhD, FACC*,* ()
- ↵*Reprint requests and correspondence:
Dr. Akira Matsumori, Department of Cardiovascular Medicine, Kyoto University Graduate School of Medicine, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8397, Japan.
Objectives This study, performed in a murine model of encephalomyocarditis virus myocarditis, used a new Millar 1.4F conductance-micromanometer system for the in vivo determination of the left ventricular (LV) pressure-volume relationship (PVR).
Background Viral myocarditis is an important cause of congestive heart failure and may lead to dilated cardiomyopathy. However, the hemodynamic changes associated with its acute phase have not been analyzed in detail.
Methods Four-week-old DBA/2 mice were inoculated with EMCV (day 0). Serial hemodynamic measurements, compared with uninfected control mice were made on days 0, 1, 3, 4, 5, 7, 9, 12, and 14.
Results On day 1, the hearts of infected mice manifested enhanced contractile function, decreased LV compliance, and abnormal diastolic function with increased LV end-diastolic pressure (EDP). Mean stroke index, ejection fraction (EF), and cardiac index (CI) were significantly higher than in uninfected control mice (p < 0.05). Contractile function decreased from days 4 to 14. On day 7, when hemodynamic abnormalities consistent with heart failure culminated, end-diastolic volume (EDV), EDP, and EDPVR were significantly higher, and CI, EF, end-systolic pressure (ESP), and ESPVR significantly lower in the infected than in control mice. Heart rate remained comparable in both groups. Although heart failure receded between day 9 and day 14, ESPVR, ESP, and EF remained significantly depressed up to day 14, and EDV and EDP remained significantly higher.
Conclusions These hemodynamic data provide new insights into the pathophysiology of acute viral myocarditis and may be useful in the development of therapeutic interventions.
Viral myocarditis is an important cause of congestive heart failure (1). In its acute phase, it is associated with systolic dysfunction, myocardial necrosis, and inflammatory cellular infiltration. A clearer understanding of the disease’s pathophysiology would help in the choice of appropriate treatment. We have described a murine model of viral myocarditis caused by the encephalomyocarditis virus (EMCV) infection (1,2). Pathologic studies performed during the acute phase in this model have shown the onset of visible myocardial changes three days after virus inoculation. Focal myocytolysis appeared on day 3, and necrotic myocardial foci with infiltration of small mononuclear cells and interstitial edema were present on day 5. Expansion of myocardial necrosis was visible after day 7, becoming extensive between day 10 and day 14. Dilation of the left ventricle (LV) was prominent by day 8, increasing further between 10 and day 14, a stage when pleural effusion, pulmonary congestion, and ascites developed (1). Pathological studies were consistent with death from congestive heart failure, although the hemodynamic characteristics of this and other murine models of viral myocarditis have not been analyzed in detail. This study applied the new conductance-micromanometer system described by Georgakopoulos et al. (3–6) to measure the LV pressure-volume relationships in the acute phase of myocarditis in our in vivo model.
Conductance catheter system design
We used the Millar 1.4F catheter (SPR-719, Millar Instruments, Houston, Texas) composed of four conductance electrodes and a micromanometer. The distance between the conductance sensor electrodes is 4.5 mm (3). The conductance system and the pressure transducer controller (Integral 3 [VPR-1002], Unique Medical Co., Tokyo, Japan) were set at a frequency of 20 kHz, the full-scale current selected at 20 μA, and the pressure transducer at 5 μV/V/100 mm Hg (7). The pressure-volume loops and intracardiac electrocardiogram were monitored online, and the conductance, pressure, and intracardiac electrocardiographic signals were digitized at 2 kHz, stored on disk, and analyzed with Integral 3 software (Unique Medical Co., Tokyo, Japan) (7).
Mice were anesthetized with a mixture of ketamine, 100 mg/kg, and xylazine, 5 mg/kg, intraperitoneally; additional smaller doses were given occasionally, as needed. The animals were placed in the supine position under a dissecting microscope (model MZ75: Leica Microsystems Wetzlar GmbH, Wetzlar, Germany), and a vertical midline cervical incision was made to expose the trachea by microsurgical techniques. After endotracheal intubation, the cannula was connected to a volume cycled rodent ventilator (Shinano Co., Tokyo, Japan) with supplemental oxygen, a tidal volume of 7 μl/g, and a respiratory rate of 140/min. The right carotid artery and external jugular vein were exposed via the same midline incision. To allow the use of a physiologic closed-chest preparation (8), a 1.4F SPR-719 Millar catheter was advanced via the right carotid artery into the ascending aorta for measurements of aortic pressure, then inserted into the LV. Left ventricular pressure-volume relations were measured by transiently compressing the inferior vena cava. The data were recorded as a series of pressure-volume loops (about 20 loops). Parallel volume (Vp) of each mouse was calibrated by the injection of 10 μl of hypertonic saline into the external jugular vein (9).
Measurements of right atrial (RA) pressure
A modified 0.014-in. Pressure Wire Sensor (Radi Medical System AB, Uppsala, Sweden) was used for right atrial (RA) pressure measurements (10). The pressure wire was inserted through the external jugular vein into the RA and connected to the Pressure Wire Interface (Radi Medical System AB, Uppsala, Sweden). The pressure tracings and surface electrocardiogram were monitored online (Biomedical Research System [LEG-1000], Nihon Koden Co., Tokyo, Japan). The RA pressure was measured, and the pressure waveforms and electrocardiogram were digitized at 2 kHz, stored to disk, and analyzed with commercially available software (Nihon Koden Co., Tokyo, Japan).
Volume calibration of the conductance catheter
The volume calibration of this conductance system was performed as described by Yang et al. (9). Briefly, seven cylindrical holes in a block 1-cm deep and with known diameters ranging from 1.4 to 5 mm were filled with fresh heparinized whole murine blood. An interelectrode distance of 4.5 mm was used to calculate the absolute volume in each cylinder. In this calibration the linear volume-conductance regression of the absolute volume in each cylinder versus the raw signals acquired by the conductance catheter was used as the volume calibration formula (Fig. 1) (5).
Analysis of the signals
All pressure-volume loop data were analyzed with the Integral 3 software (Unique Medical Co., Tokyo, Japan). Indexes of contractility and LV stiffness were calculated, including end-systolic pressure (ESP) volume relationship (ESPVR), end-systolic volume (ESV) elastance (Ees), stroke work (SW)-end-diastolic volume relation (preload recruitable stroke work [PRSW]), slope of maximum derivative of change in systolic pressure over time (dP/dtmax)-EDV relation [(dP/dtmax)/EDV], end-diastolic pressure volume relationship (EDPVR), and end-diastolic volume elastance (Eed). The Ees and Eed were normalized by 100 mg of heart weight. The ventricular-to-vascular coupling ratio was assessed by the arterial elastance (Ea)-to-Ees ratio (Ea/Ees). The time constant of isovolumic relaxation (τ) was also calculated by linear regression of dP/dtmax versus pressure from data between the minimum derivative of change in diastolic pressure over time (dP/dtmin) and 5 mm Hg above EDP. Stroke volume (SV), cardiac output (CO), and SW were normalized by body weight (stroke volume index [SVI], cardiac index [CI], and SW index [SWI], respectively). The systemic vascular resistance index (SVRI) was calculated by the following equation: SVRI = [mean aortic pressure (AOmean) − mean RA pressure (RAmean)]/CI.
A dose of 0.1 ml of the M variant of EMCV diluted in Eagle’s modified essential medium (Nissui Pharmaceutical Co., Tokyo, Japan) to a concentration of 1,000 plaque-forming U/ml was inoculated intraperitoneally in 32-day-old, 17 g, inbred male DBA/2 mice (7). A dose of 0.1 ml of phosphate-buffered saline was inoculated intraperitoneally to uninfected control mice. The day of virus inoculation was defined as day 0.
Time course of hemodynamics in uninfected or infected mice
On day 0, the uninfected group included 45 mice, and the infected group included 200 mice. Serial hemodynamic measurements were made in both groups on days 0, 1, 3, 4, 5, 7, 9, 12, and 14. The animals were randomly selected to be catheterized at each time point. In the infected group, when the measurements could not be made for technical reasons, for instance when the animal died of cardiogenic shock during the operation, or because of arrhythmias, the data were excluded from analysis. None of the uninfected mice died during an experiment. After the successful collection of hemodynamic measurements, the mice were sacrificed, the hearts were removed, and body and heart were weighed.
Statistical comparisons were made by analysis of variance with unpaired Student t test. The standard volume lines were analyzed by simple linear regression. Values are expressed as mean ± SE. A p value < 0.05 was considered statistically significant.
Time course of hemodynamic function in uninfected versus infected mice
Table 1compares multiple indexes of hemodynamic function in control versus infected mice on day 1, day 7, and day 14. No difference was found between the two groups at baseline. Heart rate remained stable in both groups over the 14 days of observation.
On day 1, at a stage when the hearts of infected mice show no gross pathologic changes (1), contractility was enhanced, diastolic function was abnormal, and EDP was increased in the group of infected animals (Figs. 2A and 2B). The ESPVR in the infected mice had a markedly steeper slope in comparison with controls (Fig. 2A). The normalized Ees (NL Ees) of the infected mice was significantly greater than that of the uninfected mice (p < 0.05) (Table 1, Fig. 3A). Consistent with the changes in ESPVR, PRSW and (dP/dtmax)/EDV in infected mice were significantly greater than in controls (Table 1). The dP/dtmax and dP/dtmin were significantly increased in the infected group (Table 1, Fig. 2B), although these two parameters vary with chamber volume, contractility, and heart rate. To normalize for these factors, dP/dtmax/dP/dtmin (Fig. 3C), which is independent of preload and afterload (4,5), was calculated and was increased by approximately 72% in the infected compared with the uninfected mice. In addition, τ, an index of diastolic relaxation, was significantly longer in infected than in uninfected mice (p < 0.01, Table 1). Prolongation of τ and an increase in dP/dtmax/dP/dtmin are both indicative of a delayed diastolic relaxation in the infected animals.
In addition to a delayed relaxation, chamber compliance was decreased in the infected group. Figure 2A also shows a markedly steeper slope of EDPVR in the infected group in comparison with controls. Normalized Eed (NL Eed) in the infected mice was significantly greater than in the uninfected mice (p < 0.05; Table 1, Fig. 3B). The EDP (Table 1, Fig. 4A) was significantly increased in the infected mice compared with controls (p < 0.05), and EDV (Table 1, Fig. 4B) was decreased, though this difference was not statistically significant. End-systolic pressure (Table 1, Fig. 4A) (p < 0.01), CI (Fig. 5A), ejection fraction (EF) (Fig. 5B), and SVI (Table 1) were significantly higher in the infected group than in the uninfected group (p < 0.05). Finally SVRI, an index of afterload, was significantly increased in the infected group (p < 0.05, Table 1).
On day 3, a stage at which small foci of myocardial necrosis are found without evidence of cellular infiltration (1), abnormal diastolic function persisted in the infected mice, however, without findings of enhanced contractility (Fig. 3A). The NL Ees, PRSW, and (dP/dtmax)/EDV were comparable in both groups (Fig. 3A). τ, dP/dtmax/dP/dtmin, dP/dtmin, and NL Eed were significantly increased in the infected mice as well as EDP (p < 0.05, Figs. 3B and 3C, and 4A). End-systolic pressure, ESV, EDV, SVI, EF, CI, dP/dtmax, RAmean, and SVRI were comparable in both groups (Figs. 4A and 4B, and 5A and 5B).
On day 4, contractility began to fall in the infected mice (Fig. 3A). The NL Ees and (dP/dtmax)/EDV of the infected mice were significantly lower than in the uninfected mice (p < 0.05). Delayed diastolic relaxation persisted, and decreased chamber compliance further decreased. τ, dP/dtmax/dP/dtmin, and dP/dtmin were significantly increased in the infected mice (p < 0.01, Fig. 3C), along with NL Eed, EDP, and RAmean, compared with controls (p < 0.01, Figs. 2B and 3A). Ejection fraction was significantly lower (p < 0.05, Fig. 5B), though SVI and CI were compensated by body weight loss (Fig. 5A). End-systolic pressure, EDV, and dP/dtmax were comparable in both groups (Figs. 4A and 4B).
On day 5, decreased contractility and abnormal diastolic function had progressed, the chambers began to dilate, and cardiac output fell in the infected animals (Figs. 3A to 3C, 4B, and 5A). End-diastolic volume and ESV were significantly greater in the infected than uninfected mice (p < 0.05 and p < 0.01, respectively, Fig. 4B). The NL Ees, PRSW, (dP/dtmax)/EDV, and dP/dtmax had fallen significantly in the infected mice compared with the controls (p < 0.05, Fig. 3A). In contrast, τ, dP/dtmax/dP/dtmin, dP/dtmin, NL Eed, EDP, and RAmean had increased significantly (Figs. 3B and 3C, and 4A). Finally, EF and CI were decreased in the infected group (Figs. 5A and 5B), while SVI and ESP were comparable in both groups (Fig. 4A).
Changes consistent with heart failure culminated on day 7. Decreased contractility, abnormal diastolic function, chamber dilation, and low output had each progressed in the infected mice (Figs. 2C and 2D, 4A and 4B, and 5A and 5B). The NL Ees, PRSW, (dP/dtmax)/EDV, dP/dtmax, EF, ESP, CI, and SVI were significantly depressed in the infected compared with the uninfected mice (Table 1). Accordingly, τ, dP/dtmax/dP/dtmin, and dP/dtmin were significantly, and EDV, ESV, NL Eed, EDP, and RAmean markedly increased in the infected mice compared with the controls (Table 1).
Days 9 to 14
Abnormal systolic and diastolic dysfunction receded, though chamber dilation progressed between day 9 and day 14 (Figs. 2E and 2F, 4A and 4B, and 5A and 5B). Though ESP, EDP, ESV, EF, dP/dtmax, dP/dtmin, dP/dtmax/dP/dtmin, τ, NL Ees, PRSW, (dP/dtmax)/EDV, AOmean, and RAmean each returned toward baseline between days 9 and 14, the differences between infected and uninfected groups on day 14 remained statistically significant (Figs. 2E and 2F, 4A and 4B, and 5B, Table 1). In contrast, SVI, CI, and NL Eed returned toward baseline between days 9 and 14 to an extent such that the differences between infected animals and controls were no longer significant on day 14 (Fig. 5A, Table 1). End-diastolic volume continued to increase past day 9 and, on day 14, remained significantly greater in the infected than in the uninfected mice (Fig. 4B).
The SVRI remained comparable in both groups from day 3 to day 14, while Ea rose significantly between day 4 and day 12 in the infected group (Table 1).
Time course of efficiency of LV work
Efficiency of LV work (SW/PVA) was depressed in the infected mice on day 4, when the contractility began to decrease (Fig. 5C). The fall in efficiency reached a nadir of 13.4 ± 2.9% on day 7, in contrast with 78.6 ± 1.4% in the uninfected group. Past day 9, the efficiency in the infected mice recovered, though remained significantly decreased on day 14 (p < 0.05). Concordant with these measurements of efficiency, Ea/Ees increased past day 4 in the infected mice, reached its peak on day 7, and recovered between day 9 and day 14 (Table 1). The SWI was decreased on day 5 when CI began to decrease in the infected group, reached its peak on day 7, and recovered between day 9 and day 14 (Table 1).
This study revealed that acute myocarditis induced by EMCV was characterized by three phases of hemodynamic evolution (Fig. 6):
1) A hyperdynamic phase was observed from day 1 to day 3. In this phase, increases in contractility, cardiac output, ESP, and vascular resistance were observed, probably from activation of sympathetic activity, as has been described in another type of viral infection (11). A hyperdynamic state is often clinically observed in the hyperacute phase of myocarditis, though has not been specifically reported. It is noteworthy that diastolic dysfunction was found concomitantly. Abnormal relaxation and increased chamber stiffness were present, despite the known improvement in diastolic function expected from catecholamines (12). This diastolic dysfunction may be caused by direct viral activity. Murine cardiac troponin T is increased, and plaque assay shows the presence of EMCV in the myocardial homogenate as early as day 1 in this same model (13) (unpublished data). Trivial injury to the plasma membrane and myocardial structural proteins may be caused by the initial viral attack on the myocytes, despite the absence of gross pathologic abnormalities on day 1. Other immune mediators, including cytokines and proteins induced by EMCV, may cause this abnormal diastolic function.
2) A depressive phase was observed between day 4 and day 7, characterized by a progressive fall in contractility, lower cardiac output, impaired myocardial relaxation, and decreased chamber compliance. Left ventricular dilation was observed past day 5, while, on day 7, cardiogenic shock and severe congestion were documented in these experiments. Impaired contractility may be caused directly by myocyte injury and indirectly by nitric oxide (NO) and cytokines, including tumor necrosis factor-α and interleukin-1β. We have reported, in this model, the expression of messenger RNA of inducible NO synthase and of these cytokines, and the importance of these mediators in its pathophysiology (13–17). The expression of these cytokines was significantly increased on day 3 and peaked on day 7 (13). Recent reports have emphasized the importance of NO and cytokines in the pathophysiology of congestive heart failure (13–21). The well-described direct and indirect negative inotropic effects of these immune-mediators are suspected to explain the decrease in myocardial contractility observed in this model (22–25). In addition, the progressive diastolic dysfunction occurring during this phase may be caused by myocytic injury, interstitial edema, and cellular infiltrations. It has been reported that NO, in contrast with its negative effects on systolic function, probably improves diastolic function, and is one of the mediators involved in compensatory mechanisms (26). However, the tissue injury due to viral infection apparently prevailed over the beneficial effects on diastolic function conferred by NO.
3) A recovery phase was observed between day 9 and day 14, during which contractility recovered slightly, and chamber compliance improved markedly. The decrease in contractility may have been caused by a fall in the production of NO and cytokines, while the improvement in chamber compliance may be explained by regressions in interstitial edema and inflammatory cellular infiltrations (1,2,13). Vascular resistance was not different between the two groups, perhaps because of production of NO and cytokines.
To the best of our knowledge, this is the first report of hemodynamic measurements, particularly of pressure-volume relationship and RA pressure, in an animal model of viral myocarditis. Our immunologic and molecular biologic studies of viral myocarditis in an EMCV animal model, which began in 1982, have provided important insights into the mechanisms of the disease. In this new hemodynamic study, distinct pathophysiologic abnormalities were identified, including a rapid and persistent fall, after an early increase, in systolic function, although, on day 14 the animals remained apparently vigorous. In contrast, diastolic function decreased from day 1 to day 7, then improved markedly to day 14. This improvement in diastolic function compensated for systolic dysfunction, with remodeling and dilation of the LV chamber. Therefore, the vitality of the mice may have reflected the recovery from congestion by an improvement of chamber compliance. These functional anomalies and remodeling process are the results of direct viral attack, immunological responses, including cytokine production, and of endocrine activity probably vastly different from the pathophysiology involved in ischemic or valvular heart disease. A detailed description of the time course of hemodynamic alterations and remodeling occurring in the acute phase of viral myocarditis, as is presented for the first time in this report, should be clinically helpful to diagnose the disease, as well as plan its management.
We recognize some limitations in our investigations. First, Vp was calibrated by the hypertonic saline method. This method assumes that hemodynamics are unchanged by the bolus of hypertonic saline (27). However, volume loading and the hypertonicity of the solution may both modify hemodynamics in this model. There were no significant differences in EDP, ESP, dP/dtmax, and dP/dtmin between measurements made immediately before the saline bolus and those made for the determination of Vp in either the uninfected or the infected group over the two weeks of study (Table 2). However, a volumetric analysis was not performed. Ultrasonic crystal measurements in mice have been reported, a method used in open-chest conditions (28). While this method accurately estimates LV volume, it could not be applied in this study, because after day 5, the infected mice were intolerant of open-chest surgery and any invasive cardiac intervention, developing arrhythmias or sudden death. Likewise, echocardiography could not be used because of the heavy pericardial calcifications present in this model (1) and the pronounced artifacts present in the small LV chambers of uninfected, five-week-old mice. Magnetic resonance imaging was not applicable because of its magnetic effects on the experimental instrumentation. Finally, though unlikely, if LV volume was modified by the saline bolus, EDV and ESV would have been underestimated (27). Second, normalization of the elastance by heart weight may or may not be optimal (29). Elastance can be influenced by chamber size. A decrease in Ees may have been caused by a significant increase in heart weight in the infected group on days 5 and 7. Maximum systolic stiffness, which is derived from the stress-strain relationship, is an index of LV contractility independent of chamber size (28). Esposito et al. (28) have described the in vivo stress-strain relationship in mice by the ultrasonic crystal method (28). However, this elegant and accurate method could not be applied in this model for technical reasons described earlier. Despite the lesser accuracy of Ees normalized by heart weight, the difference in heart weight between the two groups was small compared with that of Ees on days 5 and 7. Furthermore, there was no significant difference in heart weight between the two groups after day 9. Therefore, we believe that the significant decrease in NL Ees in the infected group after day 4 truly reflected a depression of myocardial contractility in this study.
In conclusion, these hemodynamic observations made during the first 14 days of murine EMCV-induced myocarditis provide new insights into the pathophysiology of the acute phases of the disease, and may be useful in the development of therapeutic interventions.
☆ Supported by a Research Grant from the Japanese Ministry of Health and Welfare and a Grant-in-Aid for General Scientific Research from the Japanese Ministry of Education, Science, and Culture.
- mean aortic pressure
- maximum derivative of change in systolic pressure over time
- minimum derivative of change in diastolic pressure over time
- elastance of artery
- end-diastolic pressure volume relationship
- encephalomyocarditis virus
- end-systolic pressure volume relationship
- NL Eed
- normalized end-diastolic volume elastance
- NL Ees
- normalized end-systolic volume elastance
- preload recruitable stroke work
- pressure-volume area = stroke work + potential energy
- mean right atrial pressure
- stroke volume index
- systemic vascular resistance index
- stroke work index
- parallel volume
- Received January 23, 2002.
- Revision received May 29, 2002.
- Accepted June 7, 2002.
- American College of Cardiology Foundation
- Matsumori A.,
- Kawai C.
- Matsumori A.,
- Kawai C.
- Georgakopoulos D.,
- Mitzner W.A.,
- Chen C.H.,
- et al.
- Murphy A.M.,
- Koegler H.,
- Georgakopoulos D.,
- et al.
- Georgakopoulos D.,
- Kass D.A.
- Sato T.,
- Shishido T.,
- Kawada T.,
- et al.
- Hoit B.D.,
- Ball N.,
- Walsh R.A.
- Yang B.,
- Larson D.F.,
- Watson R.
- Takagi A.,
- Tsurumi Y.,
- Ishii Y.,
- Suzuki K.,
- Kawana M.,
- Kasanuki H.
- Shioi T.,
- Matsumori A.,
- Sasayama S.
- Wang W.Z.,
- Matsumori A.,
- Yamada T.,
- et al.
- Nishio R.,
- Matsumori A.,
- Shioi T.,
- Ishida S.,
- Sasayama S.
- Nishio R.,
- Matsumori A.,
- Shioi T.,
- et al.
- Yamada T.,
- Matsumori A.,
- Sasayama S.
- Packer M.
- Bristow M.R.
- Bozkurt B.,
- Kribbs S.B.,
- Clubb F.J.,
- et al.
- Finkel M.S.,
- Oddis C.V.,
- Jacob T.D.,
- Watkins S.C.,
- Hattler B.G.,
- Simmons R.L.
- Gulick T.,
- Chung M.K.,
- Pieper S.J.,
- Lange L.G.,
- Schreiner G.F.
- Feldman M.D.,
- Erikson J.M.,
- Mao Y.I.,
- Korcarz C.E.,
- Lang R.M.,
- Freeman G.L.
- Esposito G.,
- Rapacciuolo A.,
- Prasad S.V.,
- et al.
- Hsia H.H.,
- Starling M.R.