|Home | About | Journals | Submit | Contact Us | Français|
Cardiovascular dysfunction frequently occurs after major vascular surgery or liver transplantation.
To evaluate the effects on myocardial activity of vasoactive agents released from ischemic-reperfused liver.
Isolated rat livers were perfused with Krebs-Henseleit solution (KH), propranolol 10−5 M, losartan 2×10−5 M and indomethacin 10−5 M, then made globally ischemic for 120 min (37°C) and reperfused. Isolated hearts from other rats were stabilized with KH and reperfused for 15 min with the perfusate exiting the livers. Livers were disconnected, and the hearts continued to be recirculated with the accumulated liver and heart effluent for an additional 50 min. Enzyme leakage, different vasoactive substances, left ventricular developed pressure (LVP) and coronary flow were measured during the experimental protocol.
Hepatic release of adrenaline, noradrenaline, angiotensin II, prostaglandin E2 and thromboxane B2 was significantly increased in the liver effluent following ischemia. When this effluent was directed to the heart, LVP was significantly raised in the first 10 min of reperfusion (137±5%) followed by marked decreased (46±6%) during the following 65 min of myocardial reperfusion. In the ischemic-reperfused drug-treated groups, the initial positive effect on LVP was milder than in controls (propranolol 112±12%, losartan 111±11%, indomethacin 113±9%) and the final LVP was lower (propranolol 29±6%, losartan 27±7% [P<0.05 versus ischemic control], indomethacin 46 ±12%).
During the initial phase of reperfusion, vasoactive substances released in the hepatic effluent potentiated LVP of the hearts exposed to this effluent. When the three inhibitory drugs were added to KH, this initial augmentation was not sustained. Propranolol and losartan, but not indomethacin, further depressed LVP. Vasoactive substances released from ischemic reperfused livers directly influenced heart function.
Liver transplantation is now accepted as the treatment of choice for end stage liver failure. Hypoxia of the donor liver is unavoidable during hepatic transplantation and results in hepatocellular injury (1,2). Similar tissue damage also follows vascular occlusion during hepatic lobe resection and when anastomoses are formed (3). Reperfusion of the hypoperfused or ischemic liver was found to magnify the injury, partly from the production of reactive oxygen species (4–7). Severe respiratory and cardiac dysfunction have been reported to follow major liver surgical procedures if the liver is subjected to a significant decrease in blood flow or ischemia followed by reperfusion (I/R) (8–10). Hemodynamic instability during liver transplantation in the reperfusion period has been attributed to hypovolemia, to acute left ventricular failure, as a result of the release of myocardial depressants from the postischemic donor liver and to concomitant decrease in left ventricular contractility (8–10). Postperfusion syndrome was characterized by hemodynamic changes such as bradyarrhythmias, decreased mean arterial pressure and systemic vascular resistance, and increased mean pulmonary artery and central venous pressure (10–12).
Several investigators have reported that liver ischemia (in rats and pigs) is associated with the release of adrenaline, noradrenaline, thromboxane A2 and angiotensin II (13,14). In a previous study, we found that liver I/R induced acute lung and myocardial dysfunction (15,16). We also found an immediate inotropic effect on the heart after reperfusion of the ischemic liver, followed by a rapid decline in myocardial function (15–17). The present study was designed to evaluate the presence and effects of vasoactive agents that may influence liver as well as myocardial function following liver I/R in an isolated perfused liver and heart model. We also performed experiments to selectively block the effects of specific vasoactive substances with propranolol, indomethacin and losartan, and to determine the resulting changes in hemodynamic parameters of both isolated organs.
Experiments were performed in accordance with the guidelines established by the Institutional Animal Care and Use Committee at the Rabin Medical Center, Petah Tikva, Israel.
Adult male Wistar rats (n=48) weighing 300 to 350 g were anesthetized by intraperitoneal injection of chloral hydrate (10 mg/100 g body weight). They underwent a laparotomy, and the portal vein and the supradiaphragmatic inferior vena cava were cannulated with 16 and 13 gauge cannulae, respectively. Both cannulae contained flow and pressure ports for continuous measurement. The intrahepatic inferior vena cava, the gastroepiploic vein and the hepatic artery were ligated, and the isolated liver was left intact in the rat, attached to the animal carcass within an environmental chamber. The liver was kept moist and warm (37°C). A thermistor was placed under the right lobe to control temperature (NJM-100 digital thermometer, Webster Laboratories, USA). The liver was perfused by a peristaltic pump (Watson Marlow 505U, United Kingdom) through the portal vein with oxygenated modified Krebs-Henseleit solution (KH) (in mM: 118 NaCl, 4.7 KCl, 25 NaHCO3, 2.5 CaCl2, 1.2 MgSO4, 1.2 KH2PO4, 11 d-glucose) at a rate of 3 mL/min/g liver weight. Liver outflow pressure was maintained at 0 mmHg. The perfusate was maintained at a constant temperature (37°C) and equilibrated with 95% O2 and 5% CO2 to achieve an influent PO2 of 450 to 550 mmHg, PCO2 of 30 to 40 mmHg and pH 7.34 to 7.5. This model has previously been described (15,18).
Adult male Wistar rats (n=48) weighing 280 to 320 g were injected with 500 U of heparin intraperitoneally and anesthetized. The heart was mounted on a stainless steel cannula (nonworking Langendorff preparation) and perfused with fresh KH (at 37°C). The solution was equilibrated with 95% O2 and 5% CO2 using a membrane oxygenator and passed through a 5 μm filter (Schliecher & Scholl FP-050 Dassel, Germany) into a premyocardial reservoir, at a constant pressure of 80 mmHg. Epicardial pacing wires were connected to the right ventricle and the surface of the aortic cannula. Hearts were paced at 300 beats/min (5 V, 10 ms duration) by an external Harvard stimulator (Edenbridge, United Kingdom). A latex balloon filled with water was inserted into the left ventricular cavity through a small incision in the left atrium and connected to a Statham Medical P132284 pressure transducer (Mennen Medical, Inc, USA). The balloon was tied and inflated to a volume that produced a diastolic pressure of 0 to 5 mmHg and was continuously monitored throughout the experiment. A thermistor was implanted in the right ventricle to monitor cardiac temperature. The pulmonary artery was cannulated to obtain perfusate samples for pH, PO2 and PCO2. Coronary flow was measured constantly. Perfusate afferent and efferent gases were measured at 5, 15 and 50 min of reperfusion, after stabilization. Samples were withdrawn from the aortic inflow port and the right ventricle catheter at the same time points. Figure 1 details the dual organ (liver and heart) experimental time protocol.
Pairs of one liver and one heart were allocated to eight main experimental groups (n=6 in each group). All organs were allowed to stabilize separately for 30 min. The livers were then either perfused for 2 h or made globally ischemic. Perfusion in each case was followed by a 15 min in-series reperfusion of the liver and the heart where the hepatic effluent was pumped directly into the heart. At the end of this step, the liver was removed from the circuit and the heart was left to recirculate the accumulated effluent for an additional 50 min. The hearts were not subjected to ischemia in any of these experiments (Figure 1).
In group 1a (control), the liver was stabilized and then perfused for 120 min with KH. Meanwhile, a heart from a second animal was stabilized, so that this latter phase coincided with the termination of the 120 min of liver perfusion. The hepatic effluent was then directed into the Langendorff preparation for 15 min and the heart was perfused with the postliver KH. Liver perfusion was stopped and the heart was recirculated with the accumulated perfusate for an additional 50 min.
Pairs of livers and hearts were perfused as described above with propranolol 10−5 M (group 2a), losartan 2×10−5 M (group 3a) and indomethacin 10−5 M (group 4a) to determine the spontaneous decline of the measured parameters of the heart perfused with liver effluent.
Group 1b (ischemia) liver was stabilized for 30 min and then subjected to global ischemia for 120 min. The liver was reperfused with KH and the hepatic effluent was directed into the heart for 15 min. Liver perfusion was stopped. The accumulated liver effluent was used to perfuse the heart for an additional score.
In Group 2b, the I/R protocol as in group 2 was used, with the addition of propranolol 10−5 M (Zeneca, United Kingdom) to the KH. The heart was stabilized with KH.
In Group 3b, the I/R protocol as in group 2 was used. The liver was perfused with KH containing losartan 2×10−5 M (Merck, USA). The heart was stabilized with KH.
In group 4b the I/R protocol as in group 2 was used. The liver was perfused with KH containing indomethacin 10−5 M (Sigma Chemicals, USA). The heart was stabilized with KH.
The time points for assessment of liver and myocardial functions, although recorded continuously, were as follows:
Effluents from liver and heart were collected for analysis of aspartate aminotransferase (AST) (19). Concentrations of noradrenaline, adrenaline and dopamine in the liver effluent were evaluated by high performance liquid chromatography with electrochemical detection, as described by Halbrugge et al (20). Thromboxane B2, 6-keto prostaglandin E2 (PGE2) and angiotensin II were tested by specific assay kits (San Juan, Cabiotrano, USA)
Liver and cardiac function values during stabilization were defined as 100%. Results are expressed as mean ± SEM. Statistical difference between the groups was assessed by multivariate ANOVA with repeated measurements using Duncan’s multiple comparison option. P<0.05 was considered significant.
Perfusion pressure at the portal vein was 6.9±1.6 in livers of all control groups and remained stable throughout the protocol. In the I/R livers, perfusion pressure increased from 6.9±1.3 mmHg under baseline conditions to 125±10% at 1 min reperfusion (P<0.05). Perfusion pressure to baseline values (98±12%) gradually decreased during the 15 min of reperfusion. In the drug ischemic group the same trend of perfusion pressure was observed – at first it increased and then declined: propranolol 120±13% to 106±6%; losartan 122±8% to 98±13% and indomethacin 127±15% to 112±15%.
AST concentrations of the nonischemic liver effluent during all perfusion periods were very low (0 to 15 U/L) – lower than the I/R livers (P<0.0001). AST concentration in the ischemic liver effluent was maximal at 1 min of reperfusion and was similar in all ischemic groups. Later, it declined (Figure 2). AST concentrations in the coronary effluent of hearts reperfused in the ischemic group were significantly higher (P<0.0001) than in the controls but were lower than those in the ischemic liver effluent (Figure 2).
There was a significant increase in adrenaline, noradrenaline, dopamine, thromboxane B2, PGE2 and angiotensin II (Table 1) released from the livers within 5 min following ischemia. These substances were undetected in the controls.
During the stabilization period, left ventricular developed pressure (LVP) and coronary flow were similar in all four groups (Table 2). There was a 20% spontaneous decline from the baseline values after 50 min of recirculation. This is the normal fate of myocardial contractility in this type of model (15). LVP increased to 137±5% within 15 min of reperfusion in ischemic group 1b. The LVP of hearts perfused with the ischemic liver effluent (group 1b) gradually decreased to 46±6% of baseline at the end of the reperfusion period. The addition of propranolol, losartan and indomethacin to KH before and during I/R was associated with a reduction in the increase of LVP to 113±8%, 103±8% and 113±14%, respectively, falling to a minimum of 29±6%, 27±7% (P<0.05 versus both I/R and control groups) and 46±12%, respectively, of baseline for the indomethacin-treated group (P<0.05 versus control group but similar to the ischemia only group) (Figure 3). In hearts perfused with nonischemic liver effluent containing all of the specific blockers of the vasoactive substances, depression of LVP was markedly reduced by all these drugs. In addition, it was higher than in the groups that were perfused with ischemic effluent (P<0.005, Figure 3). After 65 min of reperfusion, LVP was similar in all hearts compared with their respective controls.
The measured coronary flow was stable in the control groups and ranged from 10 to 16 mL/min. Liver ischemic effluent caused an increase in coronary flow at 10 to 15 min reperfusion (P<0.005 compared with the nonischemic groups): 152±7% (ischemia) and 141±6% (ischemia plus propranolol), 157±12% (ischemia plus losartan) and 135±6% (ischemia plus indomethacin). It then decreased significantly (P<0.05). No statistical difference was observed between any ischemic groups (P<0.05, Figure 4). Coronary flow in the control group was stable throughout the experiment. Addition of drugs to groups perfused with nonischemic effluent, 1a to 4d, did not cause any change in coronary flow 15 min after drug perfusion, which nevertheless later declined gradually (Figure 4).
The vulnerability of the liver to ischemia is one of the major nonimmunological problems of liver transplantation (11). I/R leads to a massive liberation of various vasoactive substances and to the generation of reactive oxygen species (21–24). In our previous double isolated-perfused organ model study, where the liver was subjected to similar normothermic global ischemia (ie, no flow) and then reperfused, we observed remote heart and lung injury (15,16).
In the present study, we focused on the release of vasoactive substances after liver ischemia. These substances may have a role in the pathophysiology of the postreperfusion syndrome, including cardiovascular hemodynamic changes. In all of the experimental ischemic groups, damage to liver, as expressed by AST concentrations, was similar. In our isolated liver model, substances liberated from the ischemic liver were directed to the isolated heart. We found a significant rise in the concentration of catecholamines, prostaglandins and angiotensin. The increase in these substances was accompanied by a marked rise in LVP and coronary flow. Blocking the activity of these compounds by propranolol, indomethacin and losartan partially lowered these inotropic effects. Later, however, LVP was further depressed following 50 min heart reperfusion with liver effluent. This was higher in propranolol and losartan than in ischemia and indomethacin groups (groups 2b and 3b compared with groups 1b and 4b). The increase in coronary flow (about 40%) that was observed after reperfusion with ischemic effluent was similar in all of the hearts. The addition of the various drugs did not prevent the ischemia-induced rise in coronary flow. However, we observed a decline of coronary flow in all hearts reperfused for 50 min with ischemic effluent.
The amount of adrenaline released was approximately fourfold that of noradrenalin and 11-fold that of dopamine. The various catecholamines found in the liver effluent amounted to less than 10−6 M. In preliminary experiments, a lower concentration (10−6 M) of propranolol did not block the inotropic effect occurring after ischemia. We therefore used 10−5 M of this beta-blocker. This amount should have blocked all the catecholamine-induced changes because this concentration is much higher than the maximal concentrations of the all of the catecholamines found in the liver effluent. Propranolol partially blocked the inotropic effect of the first 10 to 15 min of reperfusion, as seen by the rise in left ventricular pressure (137±5%) in the hearts perfused with ischemic liver effluent (compared with 113±9% in those perfused with ischemic liver effluent plus propranolol). These values were higher than the LVP values of the hearts perfused with nonischemic liver effluent (103±8% in group 1a compared with 89±8% in group 2a with propranolol). These results suggest that other vasoactive compounds are involved in this process. The increase in coronary flow immediately following ischemia was similar in all groups. The increased catecholamine release from organs such as liver and heart has already been described (13,14,25). The release of catecholamines may explain the increase in mean pulmonary arterial pressure that follows the reperfusion of transplanted liver (13,14,25). Myocardial damage is accompanied by adenoreceptor activation and is related to the extent of ischemic damage (25). Because the inotropic effect was not blocked completely with propranolol, we studied the effects of other vasoactive substances.
Cardiac angiotensin II type 1 (AT1) receptors have been found in several animal species (26–28). In vitro studies of cardiac preparations have shown that angiotensin II exerts a positive inotropic effect. However, in vivo results have not led to any definitive conclusions (27,28). These controversial results may originate from different experimental conditions and techniques used to assess myocardial contractility. The level of angiotensin II found in the liver effluent in our study was very low. Nevertheless, we studied the influence of its blocker, losartan, to evaluate whether the initial inotropic effect after liver ischemia induction could be blocked. Losartan mildly decreased this effect. Later, however, it had severe myocardial depression effects. In a rabbit ventricular myocardium, the positive inotropic effects and the increase in the amplitude of Ca2+ transient induced by angiotensin II were negated by losartan (10−5 M), but were not affected by a selective angiotensin II type 2 (AT2) receptor antagonist, PD123319 (29). Angiotensin II elicits positive inotropic effects by a dual mechanism: activation of AT1 receptors and increased Ca2+ transient by myofilament Ca2+ sensitivity. Angiotensin-converting enzyme inhibitors were found to lower liver and myocardial damage after ischemia (29,30). In the ischemic liver, captopril lowered reactive oxygen metabolites, endothelins and thromboxane (29).
Because concentrations of both thromboxane B2 and PGE2 were raised after liver ischemia, we added indomethacin before reperfusion to block prostaglandin metabolism. Thromboxane B2 is a stable hydrolysis product of thromboxane A2, which is a potent vasoactive, proaggregatory and chemotactic prostanoid produced from arachidonic acid by the cyclooxygenase pathway (14). It is known that in experimental liver transplantation in large animals, thromboxane B2 is increased after reperfusion of the ischemic liver (14). Indomethacin completely blocked prostaglandin synthesis and there were no traces of either thromboxane B2 or PGE2 in the liver effluent. Concurrently, indomethacin blocked the positive inotropic effect on the heart of the ischemic liver effluent (137±7% in the controls versus 113±14% in the indomethacin group, P>0.05). Prostaglandins may have a role in the inotropic effect observed after liver ischemia. Unlike indomethacin, propranolol and losartan are known to lower cardiac contractile properties (31–35). In the former, LVP continued to decline, and after 50 min of reperfusion in the indomethacin group, a similar decline of LVP and coronary flow was observed compared with the ischemic control group (46±6% compared with 46±12% for LVP and 73±5% compared with 65±5% for coronary flow).
Within the limitations of an ex vivo isolated rat double organ model, we showed that the I/R liver perfusate induced changes in myocardial function as a result of the release of different vasoactive substances. Blocking their activity diminished some of their effects. This study may explain part of the physiological mechanism involved in reperfusion injury of the graft in liver transplantation. These effects must be taken into account when treating patients after liver transplantation with vasoactive substances. These findings may have important practical applications in the clinical management of liver transplantation, as well as procedures involving no flow-reflow conditions.