Download Left Ventricular End-Diastolic Pressure-Volume

Comparative Medicine
Copyright 2003
by the American Association for Laboratory Animal Science
Vol 53, No 5
October 2003
Pages 493-497
Left Ventricular End-Diastolic Pressure-Volume
Relationship in Septic Rats with Open Thorax
Simona Cesar,* Nejka Potocnik, PhD, and Vito Starc, PhD, MD
To estimate changes in compliance, we evaluated the effects of sepsis on the end-diastolic pressure-volume relationship (EDPVR) in the left ventricle of rats that had undergone an open thorax procedure. Sepsis was induced in
male Wistar Hannover rats (n 7; 240 to 270 g) by intraperitoneal administration of a slurry of cecal contents;
control rats (n 7) were given 5% dextrose only. On the third day after induction of sepsis, left ventricular (LV)
pressure and LV dimensions were recorded simultaneously in animals of both groups. Using a micromanometer
and ultrasonic crystals, measurements were obtained at baseline and during the increase of afterload. Blood samples
were taken for determination of complete blood count, white blood cell differential count, and lactate concentration, and for bacteriologic examination. Septic rats lost weight, and developed changes in body temperature, ascites, and abscesses in the abdominal and thoracic cavities, gram-negative bacteremia, and increase in heart rate.
On the third day after induction of sepsis, LV EDPVR decreased, compared with that in the control rats (regression
coefficients: control group, 8.41 to 23.95; sepsis group, 3.94 to 7.92). Myocardial compliance in the left ventricle
increased on the third day of sepsis in the open-thorax rat model, as evidenced by the downward shift of LV EDPVR
in rats with sepsis, compared with controls.
Sepsis is a syndrome whereby, presumably as a consequence
of excess mediator release, multiple organs remote from the inciting inflammatory event sustain injury (8, 15). There is increasing evidence that sepsis alters myocardial function in
clinical and experimental settings (1, 5-7, 12-15, 17, 20, 21).
Most investigators have focussed on the characterization of systolic dysfunction (10, 16, 18), but little has been done to address
diastolic function during sepsis. Evaluation of any dysfunction
in diastole is important. Diastolic events set the conditions that
strongly determine subsequent systolic function (6). To the authors’ knowledge, characterization of diastolic dysfunction has
not yet been fully specified.
One important characteristic of diastolic function is the enddiastolic pressure-volume relationship (EDPVR). The slope of
this relationship could be interpreted as diastolic left ventricular (LV) compliance, which may alter during sepsis. Many authors came to different conclusions about EDPVR shifts during
sepsis. Using the canine model of sepsis in vivo, some authors
suggested that myocardial compliance decreases (implanted ultrasonic crystals) (21), similarly to what happens in the isolated
perfused heart of the Langendorff rat model of sepsis (6). A decrease in myocardial compliance is a fairly nonspecific response,
however, and occurs in a number of pathologic conditions, including ischemia, ventricular hypertrophy, experimental heart
transplantation, and cyanosis (21). On the contrary, in the canine model of sepsis in vivo, where live gram-negative bacteria
were introduced into the peritoneal cavity in a fibrin clot, the
authors suggested that myocardial compliance increases (12).
Finally, conflicting results regarding changes in ventricular
Received: 12/17/02. Revision requested: 2/12/03. Accepted: 5/05/03.
University of Ljubljana, School of Medicine, Institute of Physiology, Zaloska 4,
1105 Ljubljana, Slovenia.
*
Corresponding author.
compliance were documented by loss of myocardial collagen
that might contribute to the increase in myocardial compliance,
and by the presence of myocardial edema that might decrease
myocardial compliance (15).
In the study reported here, we evaluated changes in myocardial compliance, using a three-day experimental sepsis model in
rats that had undergone an open-thorax procedure, in comparison with healthy controls, as evidenced by the EDPVR shift obtained by measuring LV pressure (LVP) and LV volume (LVV),
using a micromanometer and ultrasonic crystals.
Materials and Methods
Rats. Wistar Hannover rats (n = 14, 240 to 270 g) were obtained from Harlan Italy (S. Pietro Al Natisone, Italy). The care
of laboratory animals followed accepted standards of the European Convention for the Protection of Vertebrate Animals used
for Experimental and other Scientific Purposes ETS No.: 123.
Rats were maintained in cages in groups of three in a barrier
room at 22 ± 1°C and relative humidity of 50 ± 10%, and were
fed a commercial laboratory diet and water ad libitum. The light
cycle period was controlled at 12/12-h light/dark, with no twilight transition.
Health of the animals was monitored, and they were free of
antibodies to a number of routinely tested for murine viruses. In
addition, the rats were free of all endo- and ectoparasites and
mycoplasmal species. This protocol was approved by the official
authority—the Veterinary Administration of the Republic of
Slovenia.
Sepsis induction. A donor rat was anesthetized (intraperitoneal administration of pentobarbital sodium; 50 mg/kg of body
weight), and its cecal contents were mixed with 5% dextrose in
water at a concentration of 40 mg/ml to make a slurry. Before
they were euthanized, tissue specimens were obtained from do-
493
Vol 53, No 5
Comparative Medicine
October 2003
nor rats for use in other unrelated experiments. To induce sepsis,
rats were anesthetized with halothane and received an intraperitoneal injection of cecal slurry at a dosage of 200 mg/kg in a volume
of 5 ml/kg, using a 12-gauge needle and syringe. The time-matched
control rats were given 5% dextrose in water (6). The abdomen was
gently massaged to distribute the slurry, and the rats were allowed
to awake from anesthesia. Clinical signs were monitored, including
weight change, activity level, feces consistency, physical appearance, and rectal temperature (3, 21, 23).
Animal preparation. On the third day after sepsis induction, the animals were anesthetized with a mixture of ketamine
(100 mg/kg) and xylazine hydrochloride (0.65 ml/kg), and 2,500 U
of heparin was injected into the peritoneal cavity. Blood samples
had been taken from the bulbus oculi before induction of sepsis
and were again taken on the third day of postsepsis induction
for determination of a complete blood count (CBC) and a white
blood cell differential count (Bayer-Tehnicon H*1 automated laser hematology analyzer with species-specific software V30,
New York, N.Y.), and lactate concentration (Mini 8 Photometer,
Dr. Lange, Berlin, Germany), and for bacteriologic examinations
(blood samples were cultivated on agar medium for presence of
bacteria).
Body temperature was maintained by use of a heating lamp.
Electrocardiogram electrodes were tunneled subcutaneously to
the extremities for heart rate monitoring. The trachea was exposed through a midline incision, and an airway was established by puncturing the tracheal rings, using a 22-gauge
Y-catheter connected to a rodent ventilator (Model 683, Harvard
Apparatus, South Natick, Mass.), The trachea and catheter
were sutured and secured, and the rats were ventilated with
tidal volume of 2.5 ml at 95 cycles/min. A 20-gauge soft polyethylene catheter was inserted into the right carotid artery for infusion of substances, and through it, a micromanometer (Millar
2-F, Millar Instruments, Houston, Tex.) was inserted retrograde
to measure arterial blood pressure. The thorax was opened
through a midline incision, and the pericardium was dissected
to expose the heart. Another manometer (Millar 2F, Millar Instruments) was inserted into the left ventricle through the apex
to measure the LVP. Six ultrasonic crystals (1 mm; Sonometrics
Co., London, Ontario, Canada) were attached to the LV epicardium, using a cyanocrylate adhesive (Vetbond, 3M, Animal Care
Product), to measure changes in the LV short and long axes. The
short-axis diameter was measured, using a pair of crystals attached in the cranial-caudal orientation of the left ventricle,
whereas the long-axis diameter was measured by use of one
crystal attached to the apex and another attached to the base of
the heart at the level of the aortic valve (Fig. 1). The descending
aorta was occluded transiently to increase ventricular filling,
and thus, to shift the LV pressure-volume (P-V) loops toward
higher LVV.
Pressure-volume loops. After a brief period of stabilization,
the LVP and LV diameter were recorded simultaneously at
baseline and during increases in afterload generated by a sixsecond occlusion of the aorta. The P-V data were obtained at the
end of diastole for each cardiac cycle during transient aortic occlusion (Fig. 2). The ventilator was stopped during data acquisition to eliminate the effect of positive ventilation (11).
Data analysis. All data were recorded digitally at a sampling
rate of 2,000 Hz and were stored in the computer for off-line
analysis by use of commercially available software (Sonolab;
494
Figure 1. The open-thorax ventilated rat with a Millar pressure catheter (PCath) in the left ventricle and six ultrasound crystals (UCry)
placed on the left ventriclar (LV) epicardium to measure the changes in
LV short (cranial-caudal) and long (base-apex) axis of the left ventricle.
Figure 2. The LV pressure versus LV volume obtained, using ultrasound crystals and a micromanometer. The data were recorded from
one rat during aortic occlusion. The end-diastolic points shown as black
circles are defined at the pressure in the left ventricle just prior to
dP/dt increasing above 100 mmHg/s.
Sonometrics Co.). The following parameters were derived from
the measured LVP and LV diameter for healthy controls and for
animals in sepsis: LV systolic pressure (LVSP), LV end-diastolic
pressure (LVEDP), maximal rate of pressure increase (dP/dt > 0),
aortic pressure (AoP) and heart rate (HR). The LVV was evaluated as relative change of volume according to the formula:
LVV= (V– V0)/ V0
where V0 was defined as the volume in the left ventricle at a
transmural pressure of 0 mmHg. Pressure-volume data were
obtained at the end of diastole for each cardiac cycle during
transient aortic occlusion (EDPV pair). The LV end-diastolic
points were defined as the pressure in the left ventricle just
prior to the dP/dt increasing above the rate of 100 mmHg/s. End
diastolic LVP and LVV data for individual animals were collected from at least five aortic occlusions (occlusion EDPVR
data), to provide an individual P-V relationship (individual
EDPVR data), obtained by fitting LVP and LVV data points, us-
Left ventricular physiologic relationships in septic rats
Table 1. Physiologic, hemodynamic, biochemical, and hematologic variables
on the third day after sepsis induction
Control
(n = 7)
Percentage of body weight change
LV wet weight/body weight (%)
Heart rate (bpm)
Lactate (mmol/L)
Hemoglobin (g/L)
WBC count (× 109/L)
Platelet count (× 109/L)
3.07 ± 0.35
0.168 ± 0.01
245
± 14.29
1.11 ± 0.18
151
± 2.20
11.58 ± 0.59
779
± 10
A
Sepsis
(n= 7)
–10.79 ± 0.98*
0.192 ± 0.01
320
± 12.48*
1.67 ± 0.70*
131
± 4.36*
4.75 ± 0.71*
654
± 13*
Significant difference (P < 0.01) from values for nonseptic controls.
Data are expressed as mean ± SEM.
LV = left ventricular; bpm = beats per minute; WBC = white blood cells.
*
ing linear regression. The EDPVR was constructed, using the
two-axis ellipsoid heart model to calculate the LVV.
Statistical analysis. The Student’s t test was used to determine significant differences in the hemodynamic, biochemical,
and hematologic data between septic and healthy animals
(Table 1). All data were expressed as mean ± SEM. Statistical
significance was defined at P < 0.01.
For each animal, a linear regression model was considered
appropriate (Fig. 3a and 3b). Because of significant differences
between the slopes of straight lines for animals of the group
with sepsis and in the group of healthy animals, we decided to
compare slopes in the regression model for all animal pairs. For
this purpose, multiple linear regression, using dummy variables, was used to compare pairs of the slope regression coefficients of the individual P-V relationship. Bonferroni’s test was
used to detect significant differences in slopes (P < 0.00055). The
standard statistical description and multiple linear regression
analysis were made, using Statgraphics statistical software (22).
B
Results
Sepsis and its influence on myocardial compliance were
evaluated in three categories of data: clinical appearance, biochemical laboratory data, and the P-V relationship. Clinical appearance changed within one day of sepsis induction. Animals
of the sepsis group developed lethargy and anorexia, piloerection, and loose feces. They had increased or decreased rectal
temperature. On the third day after sepsis induction, ascites
and discrete abscesses in the abdominal cavity and the lungs
were seen with the naked eye. Gram-negative bacteremia and
left shift in the CBC were found in animals with sepsis. Hemodynamic values in the animals with sepsis were increased,
along with a significant increase in HR. Decrease in body
weight occurred subsequently in the septic rats and was 10% by
day 3. Of the 10% of animals that died, all died within the first
24 h of sepsis. The hemodynamic, biochemical, and hematologic
data are summarized in Table 1.
On the third day after induction of sepsis, HR increased significantly (P < 0.01), compared with that for the control group.
Blood lactate concentration was increased significantly (P < 0.01)
in rats of the sepsis group, but was still within the normal range.
Total white blood cell and platelet counts decreased significantly
(P < 0.01) in the sepsis group; hemoglobin concentration decreased significantly (P < 0.01) in rats of the sepsis group, but
was still within the normal range. There were no significant differences in LV wet weight per body weight between groups.
The P-V relationship in diastole was significantly affected in
the animals with sepsis since the curve was displaced down-
Figure 3. The LV pressure versus LV volume in end-diastolic points
for individual healthy controls (a, n = 7) and for individual septic
animals (b, n = 7). End-diastolic pressure volume points for individual
animal are collected from at least five aortic occlusion episodes for
each heart.
ward, and there was a reduction in the slope (Fig. 3b), compared
with the value for the control group (Fig. 3a). When considering
ventricular compliance obtained by normalizing the slope of
EDPVR for each heart by its V0 , according to formula 1, the
shift of EDPVR downward occurred for the septic, compared
with the control heart, suggesting an increase of myocardial
compliance in sepsis on the third day.
Considering a linear EDPVR for each healthy control and
each septic animal, the multiple regression analysis, using
dummy variables, indicated a significant difference in the slope
of P-V lines between the all animal pairs (control/sepsis), except
in healthy control rat 4 and septic rats 12 and 13. In all instances, except in healthy control rat 4, coefficients for animals
in sepsis were significantly lower than coefficients for healthy
controls (for healthy controls, the regression coefficients ranged
495
Vol 53, No 5
Comparative Medicine
October 2003
Figure 4. Mean and SEM regression coefficients (1/mmHg) for EDPVR
data from individual healthy controls (n = 7; white bars 1–7) and for
individual septic animals (n = 7; pattern bars 8–14). In all instances,
except in healthy control rat 4, coefficients for septic animals were significantly (P < 0.00055) lower than coefficients for healthy controls.
between 8.41 to 23.95; for animals in sepsis, values were 3.94 to
7.92) (Fig. 4). Using the Bonferroni test, P < 0.00055 denotes significant difference.
Discussion
Because the nature of the changes in myocardial stiffness (or
compliance properties) in sepsis remains controversial among
various authors, our goal was to determine changes in LV compliance by determining EDPVR in a clinically relevant animal
model of chronic sepsis. With the method of obtaining LV P-V
loops, using ultrasound crystals and a micromanometer, our results confirmed the modified diastolic function in experimentally induced sepsis manifested in the apparent downward shift
of the EDPVR of the septic, compared with the control heart.
This downward shift of EDPVR suggests an increase of myocardial compliance on the third day of sepsis. This technique, using
ultrasound crystals, is precise and helpful in assessing the
changes in diastolic cardiac function.
Our findings regarding the increased compliance of the left
ventricle in the rat model of sepsis are supported by results of
several studies published previously. Price and co-workers (19)
documented response of the left ventricle to volume loading in
patients with sepsis. They observed an abnormal increase in LV
end-diastolic diameter in those who survived sepsis, implying
increased ventricular compliance. These changes in LV function
were of rapid onset and reversible in survivors within seven to
10 days.
By analogy, in the canine model of sepsis in vivo, where live
gram-negative bacteria were introduced into the peritoneal cavity in a fibrin clot, Natanson and co-workers (12) found diastolic
ventricular abnormalities that were characterized by significant increase in volume in relationship to pressure in animals
of the sepsis group, compared with values for healthy controls
at baseline. Those authors suggested that myocardial compliance was increased during sepsis (12).
On the other hand, Stahl and co-workers (21) suggested that
myocardial compliance was decreased since transmural pressure-volume-strain curves were shifted to the left in a canine
496
model of sepsis. We speculate that the reason for differences in
the measurements of LV compliance could be due to the continuous high-volume fluid application those investigators used
(rehydration) and because of the implantation of ultrasonic cardiac crystals 10 days before they induced sepsis. It is well
known that hydration of the myocardium has a strong impact
on compliance by reducing it. However, the implantation of crystals into a myocardium similarly causes tissue injury at the site
of the implantation. This injured tissue is replaced with connective tissue, which can, in turn, decrease compliance (21).
Similarly, Farias and co-workers (6) found a progressive leftward shift in the EDPVR in the Langendorff model of isolated,
perfused heart that was more pronounced on day 7 of sepsis.
This could be the result of examining the heart ex vivo (absence
of nerve regulation, absence of pleural pressure, absence of
right-to-left ventricular “cross-talk”).
Conflicting results regarding changes in ventricular compliance are also confirmed by results of the investigation where Yu
and co-workers (15) documented loss of myocardial collagen
that might contribute to an increase in myocardial compliance
and myocardial edema that decreases myocardial compliance.
Since myocardial stiffness may be affected by myocardial edema
and collagen loss in opposite directions, the net effect can only
be determined by measuring myocardial compliance.
Our study is limited by the fact that, because of the invasive
procedure, where the animal is euthanatized at the end of the
experiment, the pressure and volume on the same animal before induction of sepsis and after the third day of sepsis cannot
be measured. We overcame this problem by increasing the number of measurements. We considered that the dependence of
end-diastolic pressure on the relative volume for each animal
during sepsis is linear. For some healthy animals, exponential
dependence could also be considered. In the small interval of
volumes that described the end-diastolic points during measurement, we described the course of end-diastolic points by use
of linear regression, where some information could have been
lost. The EDPVR is neither linear nor clearly exponential, and
most investigators have not used a purely mathematical model
of the V-P relationship, but rather have demonstrated shifts in
the V-P curve.
The measurements we made define the movement of the LV
epicardial surface, not the endocardial surface. For this reason,
the changes measured in intercrystal distances may not reflect
actual volumetric changes within the ventricle. In addition, as
our measurements were made in two dimensions (base-apex
and cranial-caudal side of the left ventricle) and the volume of
the left ventricle was estimated, using the two-axis ellipsoid
heart model, the estimation of the volume is only approximate,
because the ellipsoid model is not accurate. The relationship
between the LV pressure and dimension has the typical appearance of an LV P-V loop, even though changes in ventricular size
were measured in only two dimensions by use of our technique.
Future work in embedding the crystals within the endocardium
and at multiple sites and modeling the dimensional changes in a
volumetric manner may overcome this limitation in the future.
In conclusion, our results confirmed the modified diastolic
function during experimentally induced sepsis manifested in
the downward shift of EDPVR in the early stage, suggesting an
increase of myocardial compliance the third day after sepsis induction. The precise mechanism of increasing myocardial com-
Left ventricular physiologic relationships in septic rats
pliance is unknown. Whether this diastolic abnormality is a direct consequence of bacteremia, or whether it represents a compensatory mechanism of the ventricle to allow a larger diastolic
volume without any pressure increase that would results in pulmonary edema, is not known. Elucidation of the cause of this
ventricular diastolic abnormality will require further investigation. Contributing factors to this shift may include myofibrillar
disruption, changes in myocardial collagen content, collagen remodeling, or myocardial edema (2, 9, 21). Further histologic
studies may offer some insight into discovering the predominating factor.
References
1. Bone, R. C. 1991. The pathogenesis of sepsis. Ann. Intern. Med.
115:457-469.
2. Borg, T. K., W. F. Ranson, F. A. Moslhy, and J. B. Caulfield.
1981. Structural basis of ventricular stiffness. Lab. Invest.
44(1):49-54.
3. Deitch, E. A. 1998. Animal models of sepsis and shock: a review
and lessons learned. Shock 9:1-11.
4. Esposito, G., L. F. Santana, K. Dilly, J. D. S. Cruz, L. Mao, W.
J. Lederer, and H. A. Rockman. 2000. Cellular and functional
defects in a mouse model of heart failure. Am. J. Physiol. Heart
Circ. Physiol. 279:H3101-H3112.
5. Fang, K., R. L. Krahmer, E. B. Rypins, and W. R. Law. 1996.
Starling resistor effect on pulmonary artery occlusion pressure
in endotoxin shock provides inaccuracies in left ventricular compliance assessment. Crit. Care Med. 24:1618-25.
6. Farias, S., F. M. Powers, and W. R. Law. 1999. EDPVR in sepsis: relative contributions of compliance and equilibrium chamber volume differ. J. Surg. Res. 82:172-179.
7. Grocott-Mason, R. M. and A. M. Shah. 1998. Cardiac dysfunction in sepsis: new theories and clinical implications. Intensive
Care Med. 24:286-295.
8. Hersch, M., A. A. Gnidec, A. D. Bersten, M. Troster, F. S.
Rutledge, and W. J. Sibbald. 1990. Histological and ultrastructural changes in nonpulmonary organs during early hyperdynamic sepsis. Surgery 107:397-410.
9. Hunter, P. and B. Smaill. Structure and function of the diastolic heart: material properties of passive myocardium, p. 1-26.
In P. Glass, P. Hunter, and A. McCulloch (ed.), Theory of the heart.
Biomechanics, biophysics and nonlinear dynamics of cardiac function. Springer-Verlag, New York.
10. McDonaught, K. H., C. H. Lang, and J. J. Spitzer. 1984. Depressed function of isolated hearts from hyperdynamic septic rats.
Circ. Shock 12:241-251.
11. Miyamoto, M. I., C. S. Kim, J. L. Guerrero, A. Rosenzweig,
J. K. Gwathmey and R. J. Hajjar. 1999. Ventricular pressure
and dimension measurements in mice. Lab. Anim. Sci. 49:305307.
12. Natanson, C., M. P. Fink, H. K. Ballantyne, T. J. MacVittie,
J. J. Coklin, and J. E. Parrillo. 1986. Gram-negative bacteremia produces both severe systolic and diastolic cardiac dysfunction in a canine model that simulates human septic shock. J. Clin.
Invest. 78:259-270.
13. Pedersen, P., B. Biber, S. Martinell, T. Seeman, and P-O.
Hasselgren. 1984. Hemodynamic and hematologic changes in a
standardised trauma-sepsis model in rats. Circ. Shock 14:13-23.
14. Pedersen, P. V., B. W. Warner, H. S. Bjornson, D. T. Hyiama,
S. Li, D. F. Rigel, P-O. Hasselgren, and J. E. Fischer. 1989.
Hemodynamic and metabolic alterations during experimental
sepsis in young and adult rats. Surg. Gynecol. Obstet. 168:148156.
15. Pei, Y., D. R. Boughner, W. J. Sibbald, J. Keys, J. Dunmore,
and C. M. Martin. 1997. Myocardial collagen changes and edema
in rats with hyperdynamyic sepsis. Crit. Care Med. 25: 657-62.
16. Piper, R. D., F. Y. Li, M. L. Myers, and W. J. Sibbald. 1997.
Structure-function relationship in the septic rat heart. Am. J.
Respir. Crit. Care Med. 156:1473-1482.
17. Postel, J. and P. R. Schloerb. 1977. Metabolic effect of experimental bacteremia, p. 475-480. Ann. Surg.
18. Powers, F. M., S. Farias, H. Minami, A. F. Martin, R. J. Solaro,
and W. R. Law. 1998. Cardiac myofilament protein function is
altered during sepsis. Mol. Cell. Cardiol. 30:967-978.
19. Price, S., P. B. Anning, J. A. Mitchell, and T. W. Evans. 1999.
Myocardial dysfunction in sepsis: mechanisms and therapeutic
implications. Eur. Heart J. 20: 715-724.
20. Raper, R. F. and W. J. Sibbald. 1988. The effect of coronary
artery disease on cardiac function in nonhypotensive sepsis. Chest
94:507-511.
21. Stahl, T. J., P. B. Alden, W. S. Ring, R. C. Madoff, and F. B.
Cerra. 1990. Sepsis-induced diastolic dysfunction in cronic canine peritonotis. Am. J. Physiol. 258:H625-H633.
22. Statgraphics Plus for Windows. 1998. Manugistic, Inc. 32-Bit
Operating System. Clagenfurt O’stria, Md.
23. Wichterman, K. A., A. E. Baue, and I. H. Chaudry. 1980. Sepsis and sepsis shock—a review of laboratory models and a proposal. J. Surg. Res. 29:189-201.
497