Sepsis remains a major challenge in the management of burn injury despite progress made in early wound debridement and fluid resuscitation. Although wound-related infections have gradually declined (1), the incidence of pulmonary complication remains at 50% (2, 3). Sepsis occurring during the late burn period (7-10 days) of burn is the major inciting event for multiorgan dysfunction syndrome (MODS) after burn injury. Clinical data from our institution showed that 81% patients who developed MODS after burn injury had one or more infectious complication. Clinically, burn complicated by subsequent sepsis is a classic “two-hit” injury model in which the first insult “primes” the subject to mount an altered response to the second insult. Depending on the type of the two sequential injuries, the priming process can be adaptive or maladaptive (4). When the initial burn is complicated by sepsis, it is typically a maladaptive response with more pronounced inflammatory response and higher mortality than either injury alone (5). In addition, it has been shown that time between burn and sepsis injuries also plays a role in determining the magnitude of immune dysfunction and mortality (6).
The mechanisms by which exaggerated tissue and organ injury occurs after burn complicated by sepsis are not clear. Many studies have characterized immunologic and organ function changes after this common sequential injury model. For instance, our laboratory analyzed cytokine changes in correlation with cardiac function in a rat model of 40% total body surface area (TBSA) burn followed by Streptococcuspneumoniae sepsis (7). It was found that circulating levels of TNF-α, IL-1β, IL-6, and nitric oxide were higher in burn complicated by sepsis than burn alone or sepsis alone, and these higher levels of proinflammatory cytokines correlated with decreases in the myocardial contractility measured in isolated perfused hearts (7). Despite the well-demonstrated myocardial depressive effects of experimental burn and sepsis, clinical sepsis is typically associated with a hyperdynamic response. Current methods of assessing cardiac function in vivo involve use of such global variables as cardiac output, blood pressure, and organ blood flow, all of which are sensitive to changes in systemic vascular resistance (SVR), which masks myocardial depression in intact subjects.
We previously analyzed the left ventricular pressure-volume relationship with a conductance catheter in mice after sepsis induced by cecal ligation and puncture (8, 9), and demonstrated a distinct pattern of myocardial depression associated with sepsis despite a seemingly normal cardiac output. Variables obtained with this model allowed us to separate myocardial depression from the global hemodynamic response. In the present study, we aim to characterize cardiac contractile dysfunction after burn injury complicated by S. pneumoniae pneumonia-related sepsis. We hypothesize that the previously documented profound inflammatory response after burn complicated by sepsis (7) is associated with significant myocardial depression regardless of changes in SVR in vivo; in addition, such myocardial depression is more profound after burn-complicated by sepsis than after burn alone or sepsis alone. Understanding the nature of myocardial depression may help find strategies for treatment of multiorgan dysfunction resulting from cardiac failure once infections occur after the initial burn injury.
MATERIALS AND METHODS
Male, 9- to 10-week-old C57/BL6 mice weighing 20 to 25 g (Harlan, Houston, TX) were used in the study. The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee. Hemodynamic and cardiac contractile function assessments were performed at the following time points after injury (n = 6-7 in each group): 24 h after sham burn or sham sepsis; 24 h after burn alone; 24 h after sepsis alone; 7 days after burn alone (late burn); and 24 h after sepsis, was induced at 7 days after the initial burn injury (burn complicated by late-occurring sepsis).
We used our murine model of 40% TBSA contact burn as previously described (10). Briefly, mice were placed in a plastic chamber insufflated with 2.5% isoflurane in oxygen. After loss of consciousness, the skin was clipped, shaved, and washed with alcohol. A contact burn was induced by brass probes preheated to 100°C in boiling water for 30 min. The size of the probes and number of contact areas were selected based on the animal's size to achieve a 40% TBSA burn area, which covered the entire dorsal area below the neck. Special precautions were taken to induce complete third-degree burn without any margins of first- or second-degree burn. This previously described model of injury was confirmed by histology (10). After the burn injury, animals were given 4 mL/kg/% TBSA intraperitoneal lactated Ringer's solution for resuscitation, and 0.05 mg/kg buprenorphine given every 12 h after burn for pain control. Mice were allowed to recover in cages placed on warming blankets with free access to food and water. Sham burn mice underwent the same anesthesia and skin preparation, but were only exposed to brass probes heated to 37°C.
S. pneumoniae sepsis
The induction of sepsis was performed using aseptic techniques to avoid contamination from bacteria other than the S. pneumoniae intended for the study. Mice were anesthetized with isoflurane as described above, placed in a supine position, and the area over the trachea was prepared with povidone-iodine and alcohol. A midline incision was made over the trachea; the trachea was identified and isolated via blunt dissection. A 0.1-mL aliquot of bacterial suspension containing 1 × 105 CFU of S. pneumoniae was injected directly into the trachea using a 27-gauge needle. The concentration of the bacteria was determined by densitometry and was confirmed with culture. Sham mice received the 0.1 mL of phosphate-buffered solution as vehicle without bacteria. After bacterial or vehicle instillation, animals were placed in a 30° head-up position for 5 min to facilitate entry of bacteria into the lungs. The wound was closed with 6-0 polypropylene sutures. After receiving 2 mL of intraperitoneal lactated Ringer's solution for fluid resuscitation, mice were allowed to recover from anesthesia. This model of pneumonia-related sepsis was confirmed 24 h after bacterial challenge by positive blood cultures (>2 × 103 CFU) and bronchial lavage cultures (>1 × 106 CFU) of the original bacterial strain.
Hemodynamic and cardiodynamic assessments
Detailed methods of hemodynamic and cardiac contractility measurements using a miniature conductance catheter were described elsewhere (8). Briefly, mice were anesthetized with 1.5% to 2% inhaled isoflurane in oxygen. The animals' temperature was maintained between 36°C and 37.5°C with a heating blanket and heating lamp. Tracheostomy was performed to allow positive pressure mechanical ventilation. The left carotid artery was cannulated for mean arterial blood pressure measurements and fluid administration with lactated Ringer's solution. A clamshell thoracotomy was performed at 2 mm above the level of xyphoid process, and the heart was exposed. The pericardium was excised and a 6-0 silk tie was placed around the inferior vena cava (IVC). The left ventricular pressure-volume catheter (SPR 839; Millar Instruments, Houston, TX) was inserted via the apex of the heart and secured along the longitudinal axis of the left ventricle once all electrodes were within the left ventricular cavity. Steady-state hemodynamics and contractility were measured, followed by contractility parameters with preload reduction achieved by gently lifting the silk tie to cause transient (1-2 s) IVC occlusion while suspending mechanical ventilation. Data were digitally converted and displayed using Chart software (version 4.12; AD Instruments, Castle Hill, Australia). Hemodynamic and contractility measurements were made at a constant isoflurane concentration of 1.5% and a left ventricular end-diastolic volume of 15 to 18 real-time relative volume units corresponding to approximately 30 to 35 μL based on our previous calibrations (8). Parallel conductance of the myocardium was corrected using the hypertonic saline technique (11).
Hemodynamic and contractility measurements were made in two stages, steady state and IVC occlusion. Steady-state hemodynamic measurements included heart rate (HR), stroke volume (SV), cardiac output (CO), and pressure development during isovolumic contraction (dp/dtmax) and relaxation (dp/dtmin). Variables obtained during IVC occlusion included left ventricular end-systolic pressure volume relationship (ESPRV), preload-recruitable stroke work (PRSW), time-varying maximum elastance (Emax), and dp/dtmax corrected for varying left ventricular diastolic volume (dp/dtmax/LVEDV). The slopes of ESPVR and PRSW represented the end-systolic pressure and stroke work of the left ventricle generated in response to a changing left ventricular end-diastolic volume caused by transient IVC occlusion. Likewise, Emax was the maximum left ventricular pressure/volume response at various preload conditions caused by IVC occlusion. The slopes of these curves were numerically averaged for each experimental group. To overcome inconsistencies in pressure-volume relationship from curvilinearity at high left ventricular systolic pressures in mice (12), ESPVR was analyzed with pressure-volume loops with an end-systolic pressure of 80 mmHg or less. Data were analyzed with pressure-volume analysis software (PVAN version 3.1; Millar Instruments).
Data are expressed as mean ± SEM. Comparisons among groups were performed by one-way analysis of variance and Dunnett's multiple comparisons test. A value of P < 0.05 was considered statistically significant.
Sham burn and sham sepsis groups exhibited similar hemodynamics and cardiac contractility results; these groups were combined and used as the control group for statistical analysis.
Global hemodynamic and cardiac contractile function assessments were made at the early (24 h) and late (7 days) period after the 40% TBSA burn. Global hemodynamic variables were measured during steady state without IVC occlusion. There were no significant changes in any of the variables during early or late burn (Table 1).
Cardiac contractility was measured during steady-state and IVC occlusion. dp/dtmax and dp/dtmin was obtained during steady state, and showed no statistically significant changes (Fig. 1A). When dp/dtmax was corrected for dp/dtmax/LVEDV, a variable obtained with IVC occlusion and considered more sensitive than dp/dtmax alone (13), there was a significant decrease in dp/dtmax/LVEDV during early burn, but the value returned to baseline during late burn (Fig. 1B). Other cardiac contractility variables obtained with IVC occlusion included ESPVR, PRSW, and Emax, which all indicated marked myocardial depression associated with early burn (Fig. 2). During late burn, these cardiac contractility indices returned toward baseline, but ESPVR remained significantly depressed, indicating a certain degree of residual myocardial depression (Fig. 2).
Sepsis was confirmed by the presence of S. pneumoniae in blood cultures (>2 × 103 CFU) 24 h after intratracheal bacterial challenge. There were no statistically significant changes in HR and mean arterial pressure (MAP) 24 h after the bacterial challenge. There was a significant decrease in CO paralleled by an increase in SVR in septic mice, which was responsible for the maintained MAP (Table 1).
Changes in dp/dtmax and dp/dtmin measured during steady state with sepsis alone were not statistically significant (Fig. 1A). However, changes in dp/dtmax corrected for varying left ventricular end-diastolic volume accomplished by IVC occlusion (dp/dtmax/LVEDV) were statistically significant (Fig. 1B). All other indices of load-insensitive contractility variables, including ESPVR, PRSW, and Emax, showed significant decreases with sepsis alone (Fig. 2).
Burn complicated by sepsis
We induced S. pneumoniae sepsis 7 days after the initial burn injury. The timing of this late-occurring septic complication was based on the evidence of maximum immune dysfunction (6), as well as on our previous finding of worsened myocardial function with the infectious complication (14) during this late period of burn injury. It is also consistent with the timing of the development of infection and multiorgan dysfunction in adult burn patients in our institution (15).
When burn injury was complicated by pneumonia sepsis, there was a significant decrease in HR. CO was also reduced by 50% compared with sham group. However, because of a 100% increase in SVR, MAP was not significantly changed despite a dramatic decrease in CO (Table 1). All cardiac contractility variables consistently demonstrated severe myocardial depression after burn complicated by sepsis. For example, dp/dtmax and dp/dtmin, measured during steady state, as well as dp/dtmax/LVEDV during IVC occlusion, showed significant reductions (Fig. 1). Other variables measured with IVC occlusion, including ESPRV, PRSW, and Emax, showed even more significant myocardial depression with burn complicated by sepsis compared with either burn alone or sepsis alone (Fig. 2).
Sepsis is a major contributor to morbidity and mortality after burn injury. Clinical data from the adult burn population in our own institution clearly demonstrated that most cases of multiorgan dysfunction after the initial burn injury resulted from an infection complication (15). Burn complicated by sepsis is one “two-hit” injury model whose pathophysiology and consequences are still a focus of intense investigation. We used a mouse model of cutaneous burn followed by intratracheal instillation of S. pneumoniae 7 days after the initial burn to closely mimic the classic picture of burn injury complicated by late-occurring pneumonia-related sepsis. Previously, our group characterized sepsis induced at various periods postburn, and determined that sepsis induced 48 h postburn caused a greater degree of myocardial depression than that induced 72 h postburn (7), and the greatest inflammation and cardiac defects occurred when sepsis was induced 7 to 10 days postburn (14). In addition, cardiac contractility slowly returned toward baseline 7 days after burn in the absence of sepsis (14). The concept of inducing pneumonia sepsis during the recovering phase (7 days) after the initial burn injury was consistent with the development of ventilator-related pulmonary infections in the clinical setting of burn injury (2, 3). Thus, we used this model to evaluate the impact of late-occurring pneumonia sepsis on the course of burn injury in terms of cardiac contractile dysfunction and its interaction with changes in systemic vascular resistance.
Although numerous studies have shown myocardial contractile dysfunction after burn and sepsis, clinical experience seldom points to cardiac failure as the fatal event. In a porcine model of flame burn injury, Tadros et al. (16) showed that CO decreased only during the immediate hour after burn, and by the second hour, CO had increased to almost 40% above baseline, likely reflecting aggressive fluid resuscitation (16). When burn is complicated by sepsis, the typical cardiovascular response is even more hyperdynamic, despite the fact that sepsis is well documented to cause myocardial depression (17). Unfortunately, clinically available in vivo measurements such as CO and ejection fraction do not provide sensitive and adequate determination of cardiac contractility in the face of changes in SVR.
SVR may be the key component in determining the clinical manifestations of sepsis. In a murine model of polymicrobial sepsis caused by cecal ligation and puncture (CLP), we showed “normalized” CO despite decreased cardiac contractility, with a reduction of SVR and MAP (8). In the present study using the pneumonia sepsis model, mice maintained HR and MAP after burn, sepsis, or even burn complicated by sepsis. CO, however, decreased as SVR increased. The differences in these variables between CLP and pneumonia sepsis models represent the clinical septic shock and septic syndrome, respectively, reflecting different manifestations determined by variations in the SVR. It appears that the response of the systemic vasculature determines the overall hemodynamic picture-whether the SVR is increased to maintain the blood pressure or decreased to maintain the CO. Severe myocardial depression may limit the ability to maintain both, which would result in decreased organ perfusion and increased mortality.
The fact that dp/dtmax and dp/dtmin did not show significant changes with burn alone or sepsis alone was somewhat interesting. Our previous studies using isolated heart preparations repeatedly showed significant decreases in dp/dtmax and dp/dtmin with burn alone or sepsis alone (7, 10). We contribute this discrepancy to the different models used for assessing cardiac contractility. In isolated perfused hearts, dp/dt is typically generated with a relatively constant preload, afterload, perfusate composition, and absence of sympathetic influence, thus it serves as reliable indicator of myocardial contractility. In the intact animal, dp/dtmax and dp/dtmin are sensitive to many factors, including afterload and heart rate, and further modulated by changes in sympathetic outflow, all could mask the underlying changes in dp/dtmax and dp/dtmin as indicators of myocardial contractile function. The fact that in vitro heart preparations typically generate much lower dp/dtmax and dp/dtmin values than those measured in the intact animal (7, 10) supports the theory of neurohumoral modulation. The use of IVC occlusion allowed us to obtain dp/dtmax corrected for LVEDV, which examined dp/dt in wide range of preload conditions and offered a more accurate assessment of the intrinsic adaptability of the left ventricle to generate force in response to a varying load challenge. This is probably the reason why dp/dtmax/LVEDV showed a significant decrease with burn alone or sepsis alone, whereas steady-state dp/dtmax did not. Our finding confirmed that dp/dtmax/LVEDV may be a more sensitive and useful measurement of cardiac performance in intact animals (13).
Cardiac contractility variables obtained with varying preload by IVC occlusion, such as ESPRV, PRSW, and Emax, all showed differences with burn, sepsis, and especially burn complicated by sepsis. Cardiac contractile defects were confirmed despite a wide range of SVR encountered in the present and other sepsis models (8), unmasking the underlying myocardial depression from the confounding changes in afterload. Findings in the present study further confirm that left ventricular pressure-volume analysis is insensitive to load conditions (18), making this approach desirable in burn and sepsis research.
Our previous studies in isolated mouse and rat hearts showed that myocardial depression occurred as early as 8 h and peaked at 24 h; if sepsis did not occur, myocardial contractility gradually returned to baseline 7 days after burn (7, 10, 14). In the present study using intact mice, we saw the same trend of early myocardial depression and gradual recovery with burn alone. When burn was complicated by sepsis, the severity of myocardial depression was greater than after burn alone or sepsis alone; moreover, it occurred at a time when myocardial contractility was expected to return toward normal after burn injury alone. This observation supports the fact that sepsis after an initial burn injury is a destructive or maladaptive priming process (4) in which the inflammatory response is more exaggerated than burn alone or sepsis alone.
Using left ventricular pressure-volume analysis, we were able to demonstrate a significant pattern of myocardial depression during sepsis that occurred in the recovery phase of the initial burn injury despite certain seemingly normal hemodynamic parameters. The high morbidity and mortality from burn injury complicated by sepsis may be related to the drastic impairment in myocardial contractility. Our data are consistent with our previously identified time course of burn-related myocardial depression measured with isolated heart preparations. In addition, we have shown that left ventricular pressure-volume analysis with IVC occlusion is a desirable method to characterize cardiac dysfunction during burn and sepsis within a wide range of load conditions.
1. Mayhall CG: The epidemiology of burn wound infections. Then and now. Clin Infect Dis
2. Rue LW III, Cioffi WG, Mason AD Jr, McManus WF, Pruitt BA Jr: The risk of pneumonia in thermally injured patients requiring ventilatory support. J Burn Care Rehabil
3. de La Cal MA, Cerda E, Garcia-Hierro P, Lorente L, Sanchez-Concheiro M, Diaz C, van Saene HK: Pneumonia in patients with severe burns: a classification according to the concept of the carrier state. Chest
4. Meldrum DR, Cleveland JC Jr, Moore EE, Partrick DA, Banerjee A, Harken AH: Adaptive and maladaptive mechanisms of cellular priming. Ann Surg
5. Sasaki J, Fujishima S, Iwamura H, Wakitani K, Aiso S, Aikawa N: Prior burn insult induces lethal acute lung injury in endotoxemic mice: effects of cytokine inhibition. Am J Physiol Lung Cell Mol Physiol
6. Moss NM, Gough DB, Jordan AL, Grbic JT, Wood JJ, Rodrick ML, Mannick JA: Temporal correlation of impaired immune response after thermal injury with susceptibility to infection in a murine model. Surgery
7. White J, Thomas J, Maass DL, Horton JW: Cardiac effects of burn injury complicated by aspiration pneumonia-induced sepsis. Am J Physiol Heart Circ Physiol
8. Tao W, Deyo DJ, Traber DL, Johnston WE, Sherwood ER: Hemodynamic and cardiac contractile function during sepsis caused by cecal ligation and puncture in mice. Shock
9. Tao W, Sherwood ER: β-2 microglobulin knockout mice treated with anti-ASAIOGM-1 exhibit improved hemodynamics and cardiac contractile function during acute intraabdominal sepsis. Am J Physiol Regul Integr Comp Physiol
10. White J, Maass DL, Giroir B, Horton JW: Development of an acute burn model in adult mice for studies of cardiac function and cardiomyocyte cellular function. Shock
11. Yang B, Larson DF, Beischel J, Kelly R, Shi J, Watson RR: Validation of conductance
catheter system for quantification of murine pressure
loops. J Invest Surg
12. Georgakopoulos D, Kass DA: Pressure
relations. In Hoit BD, Walsh RA (ed): Cardiovascular Physiology in the Genetically Engineered Mouse
. Ed 2. Boston: Kluwer Academic Publishers, 2002, pp 207-222.
13. Little WC, Cheng CP, Mumma M, Igarashi Y, Vinten-Johansen J, Johnston WE: Comparison of measures of left ventricular contractile performance derived from pressure
loops in conscious dogs. Circulation
14. Horton JW, Maass DL, White J, Sanders B: Myocardial inflammatory responses to sepsis complicated by previous burn injury. Surg Infect
15. Fitzwater J, Purdue GF, Hunt JL, O'Keefe GE: The risk factors and time course of sepsis and organ dysfunction after burn trauma. J Trauma
16. Tadros T, Traber DL, Herndon DN: Opposite effects of prostacyclin on hepatic blood flow and oxygen consumption after burn and sepsis. Ann Surg
17. Court O, Kumar A, Parrillo JE, Kumar A: Clinical review: myocardial depression in sepsis and septic shock. Crit Care
18. Sagawa K, Maughan L, Suga H, Sunagawa K: Physiologic determinants of the ventricular pressure
relationship. In Sagawa K, Maughan L, Suga H, Sunagawa K (ed): Cardiac Contraction and the Pressure-Volume Relationship
. New York: Oxford University Press, 1988, pp 110-170.