Cardiac surgical patients exposed to heparin-bonded circuits develop less postoperative cerebral dysfunction than patients exposed to non-heparin-bonded circuits.

A prospective randomized trial was used to study the incidence of cerebral dysfunction in patients undergoing cardiopulmonary bypass (CPB) with heparin-bonded vs non-heparin-bonded circuits. Although the etiology of postoperative cerebral dysfunction is controversial, activation of the systemic inflammatory response may play a role. After institutional approval and informed written consent, 39 elective coronary artery bypass (CABG) patients were studied. A battery of neuropsychometric tests (NPMTs) was performed preoperatively, and 5 days and 6 weeks postoperatively. Significant change in NPMT performance was defined as a 25% or greater decrease in postoperative performance, compared to baseline. The number of abnormal tests per patient was calculated. Analysis using the Mann-Whitney rank test was performed for the first follow-up. Patients randomized to heparin-bonded circuits had fewer abnormal NPMTs (>1 abnormal test) on postoperative day 5 (58 vs 70%, n=19 and 20) than patients randomized to non-heparin-bonded circuits. Patients exposed to heparin-bonded circuits had fewer abnormal tests (>1 abnormal test) at 6 weeks (36 vs 63%, n=14 and 16). Results suggested that the attenuation of systemic inflammation by heparin-bonded CPB circuits may lower the incidence of cerebral injury in cardiac surgical patients.

A randomized trial of antithrombin concentrate for treatment of heparin resistance.

BACKGROUND: Heparin resistance is an important clinical problem traditionally treated with additional heparin or fresh frozen plasma. We undertook a randomized clinical trial to determine if treatment with antithrombin (AT) concentrate is effective for treating this condition. METHODS: Patients requiring cardiopulmonary bypass who were considered to be heparin resistant (activated clotting time < 480 seconds after > 450 IU/kg heparin) were randomized to receive either 1000 U AT or additional heparin. RESULTS: AT concentrate was effective in 42 of 44 patients (96%) for immediately obtaining a therapeutic activated clotting time. This compared favorably to 28 of 41 patients (68%) treated with additional heparin (p = 0.001). All patients who failed heparin therapy were successfully treated with AT. The patients receiving AT required less time to obtain an adequate ACT but there was no difference in clinical outcomes among the groups. Study patients had deficient AT activity at baseline (56%+/-25%), which improved in those given AT concentrate (75%+/-31% versus 50%+/-23%, p < 0.0005). CONCLUSIONS: Heparin resistance is frequently associated with AT deficiency. Treating this deficiency with AT concentrate is more effective and faster for obtaining adequate anticoagulation than using additional heparin.

Endotoxin in pooled pericardial blood contributes to the systemic inflammatory response during cardiac surgery.

Although endotoxin has been implicated as an important contributor to the systemic inflammatory response (SIR) during cardiopulmonary bypass (CPB), its source remains unclear. While gut translocation has traditionally been perceived as the primary source of endotoxemia, accumulation of endotoxin in pooled pericardial blood may represent an additional source of endotoxin that is continually reinfused into the CPB circuit. Eighteen patients undergoing primary coronary revascularization procedures were prospectively evaluated. Shed blood pooled in the pericardial space was returned to the CPB circuit through cardiotomy suction catheters at 45 min after placement of the aortic cross-clamp. Simultaneous samples of pooled pericardial and peripheral arterial blood were obtained and analyzed by a limulus amebocyte lysate assay for the determination of endotoxin concentration, and an enzyme-linked immunosorbent assay for tumor necrosis factor (TNF-alpha) levels. Significant elevations in endotoxin were demonstrated in pooled pericardial blood samples compared with arterial blood (3.5 +/- 0.5 vs 0.8 +/- 0.2 pg/ml; p < 0.05). TNF-alpha levels were below the limits of detection in both samples. These data implicate pooled pericardial blood as an important primary source of endotoxin that, when continually reinfused throughout CPB, may contribute to the overall SIR. Because endotoxemia has been identified as an important predictor of adverse outcomes following cardiac surgery, removal of endotoxin antigen in shed pericardial blood, prior to its reinfusion into the CPB circuit, may provide a directed means to improve perioperative outcome without compromising established blood conservation techniques.

Vasopressin deficiency and pressor hypersensitivity in hemodynamically unstable organ donors.

BACKGROUND: Solid organ donors often develop hypotension due to vasodilation, and recently we observed that a variety of vasodilatory states are characterized by vasopressin deficiency and hypersensitivity. Thus, we investigated the prevalence of vasopressin deficiency in hypotensive solid organ donors without clinical evidence of diabetes insipidus; we also investigated the vasopressor effect of vasopressin replacement in hypotensive donors. METHODS AND RESULTS: Fifty organ donors were evaluated for hemodynamic instability, (mean arterial pressure [MAP]

Cardiac transplantation in a patient with protein S deficiency.

Cardiac transplantation was successfully performed in a patient with end-stage ischemic cardiomyopathy and hereditary protein S deficiency who had undergone two previous coronary artery bypass graft procedures. Routine intraoperative heparinization and reversal with protamine was undertaken, and the antifibrinolytic agent aprotinin was infused throughout the procedure without perioperative hemorrhage or thrombosis. Systemic anticoagulation with intravenous heparin was resumed on postoperative day 2, and the patient was then converted to Lovenex as outpatient anticoagulation to facilitate routine surveillance endomyocardial biopsies.

Time-dependent cellular population of textured-surface left ventricular assist devices contributes to the development of a biphasic systemic procoagulant response.

OBJECTIVE: Textured-surface left ventricular assist devices (LVAD) have been shown to enhance ventricular function and survival in patients with end-stage heart failure. Furthermore, we have described a procoagulant physiology in our LVAD population with sustained thrombin generation (elevated thrombin-antithrombin III complex and prothrombin fragment 1+2) and fibrinolysis (D-dimers), even up to 335 days after LVAD placement. To explain such sustained activation of coagulation, we speculated that the LVAD surface selectively adsorbed and promoted activation of circulating blood cells. METHODS: In a prospective study of 20 patients with LVADs, we examined samples of peripheral blood as well as cells harvested from the surface of the LVADs at the time of their explantation for procoagulant proinflammatory markers. RESULTS: Analysis of the cells populating the LVAD surface revealed the presence of pluripotent hematopoietic CD34(+) cells, as well as cells bearing monocyte (CD14)/macrophage (CD68) markers, which also expressed procoagulant tissue factor. Reverse transcriptase-polymerase chain reaction confirmed cellular activation on the LVAD surface, revealing transcripts for interleukin 1alpha, interleukin 2, and tumor necrosis factor alpha, in addition to vascular cell adhesion molecule-1 consistent with their capacity to continually recruit and activate circulating cells, thereby propagating their response. In the periphery, elevated levels of tissue factor were found in the plasma of patients with LVADs, along with enhanced procoagulant activity. CONCLUSION: These observations suggest that the LVAD surface selectively absorbs and activates circulating hematopoietic precursor and monocytic cells, thereby creating a sustained prothrombotic and potentially proinflammatory systemic environment.

Activation-induced T-cell death and immune dysfunction after implantation of left-ventricular assist device.

BACKGROUND: Cardiac transplantation is a limited option for end-stage heart failure because of the shortage of donor organs. Left-ventricular assist devices (LVADs) are currently under investigation as permanent therapy for end-stage heart failure, but long-term successful device implantation is limited because of a high rate of serious infections. To examine the relation between LVAD-related infection and host immunity, we investigated immune responses in LVAD recipients. METHODS: We compared the rate of candidal infection in 78 patients with New York Heart Association class IV heart failure who received either an LVAD (n=40) or medical management (controls, n=38). Fluorochrome-labelled monoclonal antibodies were used in analyses of T-cell phenotype. Analysis of T-cell function included intradermal responses to recall antigens and proliferative responses after stimulation by phytohaemagglutinin, monoclonal antibodies to CD3, and mixed lymphocyte culture. We measured T-cell apoptosis in vivo by annexin V binding, and confirmed the result by assessment of DNA fragmentation. Activation-induced T-cell death was measured after T-cell stimulation with antibodies to CD3. All immunological tests were done at least 1 month after LVAD implantation. Between-group comparisons were by Kaplan-Meier actuarial analysis and Student's t test. FINDINGS: By 3 months after implantation of LVAD, the risk of developing candidal infection was 28% in LVAD recipients, compared with 3% in controls (p=0.003). LVAD recipients had cutaneous anergy to recall antigens and lower (<70%) T-cell proliferative responses than controls after activation via the T-cell receptor complex (p<0.001). T cells from LVAD recipients had higher surface expression of CD95 (Fas) (p<0.001) and a higher rate of spontaneous apoptosis (p<0.001) than controls. Moreover, after stimulation with antibodies to CD3, CD4 T-cell death increased by 3.2-fold in LVAD recipients compared with only 1.2-fold in controls (p<0.05). INTERPRETATION: LVAD implantation results in an aberrant state of T-cell activation, heightened susceptibility of CD4 T cells to activation-induced cell death, progressive defects in cellular immunity, and increased risk of opportunistic infection.

100 long-term implantable left ventricular assist devices: the Columbia Presbyterian interim experience.

BACKGROUND: The use of left ventricular assist devices (LVADs) as bridge to transplantation is now accepted as a standard of care for a subset of end-stage heart failure patients. Our interim experience with both pneumatically and electrically powered ThermoCardiosystems LVADs is presented to outline the benefits and limitations of device support as well as discuss its potential role as bridge to recovery and as destination therapy. METHODS AND RESULTS: Detailed records were kept prospectively for all patients undergoing LVAD insertion. One hundred LVADs were inserted over 7 years into 95 patients, with an overall survival rate of 75% and a transplantation rate of 70%. Four patients underwent device explant for recovered myocardial function. Three patients received LVADs as destination therapy in the ongoing REMATCH (Randomized Evaluation of Mechanical Assist Treatment for Congestive Heart failure) trial. Overall mean patient age was 51 years, and mean duration of support was 108 days. There were 25 device-related infections including the drive line, device pocket, and blood-contacting surfaces. Cerebral vascular accidents and other embolic events occurred in 7 patients with six deaths. There were four device malfunctions and nine graft-related hemorrhages, resulting in six reoperations and three deaths. CONCLUSIONS: The use of long-term implantable LVADs will likely not be limited to bridge to transplantation. The REMATCH trial has commenced to study the role LVADs may have as an alternative to medical management. Furthermore, as the issues of myocardial recovery are examined, the "bridge to recovery" may be an important additional role for these assist devices.

Aprotinin in deep hypothermic circulatory arrest.

Early experience with aprotinin in deep hypothermic circulatory arrest (DHCA) raised alarm about hazards associated with its use. Based on what little is known about possible mechanistic interactions between hypothermia, stasis, and aprotinin, there is no evidence that aprotinin becomes unusually hazardous in DHCA. Excessive mortality and complication rates have only been reported in clinical series in which the adequacy of heparinization is questionable. Benefits associated with use of aprotinin in DHCA have been inconsistently demonstrated. The only prospective, randomized series showed significant reduction in blood loss and transfusion requirements. Use of aprotinin in DHCA should be based on the same considerations applied in other cardiothoracic procedures.

Improved survival rates support left ventricular assist device implantation early after myocardial infarction.

OBJECTIVES: Implantation of left ventricular assist devices (LVADs) early after acute myocardial infarction (MI) has traditionally been thought to be associated with high mortality rates due to technical limitations and severe end-organ dysfunction. At some experienced centers, doctors have refrained from earlier operation after MI to allow for a period of hemodynamic and end-organ stabilization. METHODS: We retrospectively investigated the effect of preoperative MI on the survival rates of 25 patients who received a Thermocardiosystems Incorporated LVAD either <2 weeks (Early) (n = 15) or >2 weeks (Late) (n = 10) after MI. Outcome variables included perioperative right ventricular assistance (and right-sided circulatory failure), hemodynamic indexes, percent transplanted or explanted, and mortality. RESULTS: No statistically significant differences were demonstrated between demographic, perioperative or hemodynamic variables between the Early and Late groups. Patients in the Early group demonstrated a lower rate of perioperative mechanical right ventricular assistance, but had a higher rate of perioperative inhaled nitric oxide use. In addition, 67% of patients in the Early group survived to transplantation and 7% to explantation, findings comparable to those in the Late group (60% and 0% respectively). CONCLUSIONS: This clinical experience suggests that patients may have comparable outcomes whether implanted early or late after acute MI. These data therefore support the early identification and timely application of this modality in post-MI LVAD candidates, as this strategy may also reveal a subgroup of patients for whom post-MI temporary LVAD insertion may allow for full ventricular recovery.

Improved survival in patients with acute myocarditis using external pulsatile mechanical ventricular assistance.

BACKGROUND: Acute myocarditis remains a disease with a variable clinical course, from full ventricular recovery to complete heart failure; to date, few cases have been reported that describe the efficacy of temporary mechanical ventricular assistance for its treatment. METHODS: We evaluated the voluntary world registry with the use of an external pulsatile ventricular assist device (the ABIOMED BVS 5000 [BVS]) for acute myocarditis to determine the impact of mechanical ventricular assistance on outcome. Variables analyzed included patient demographics, serum chemistries, and overall hemodynamics prior to BVS, while on BVS support, and after BVS explanation. Postoperative parameters included re-operation, bleeding, respiratory failure, renal failure, and infections, neurologic, or embolic events. RESULTS: Eighteen patients in the ABIOMED world registry underwent BVS implantation for myocarditis; 11 (61.1%) had complete pre-operative and hemodynamic data for analysis. Patients were supported for 13.2 +/- 17.0 days, after which time 7 (63.6%) patients survived to explanation of the device and 2 (18.2%) underwent transplantation. Elevated admission serum chemistries (blood ureanitrogen [BUN], creatinine, transaminases) and hemodynamics (central venous pressure [CVP], mean pulmonary arterial pressure [PAP], pulmonary capillary wedge pressure [PCW], cardiac index [CI], all normalized during the period of device support. Estimated ejection fractions in the 7 explanted patients ranged between 50 to 60% at routine evaluation 3 years after device removal. CONCLUSIONS: Temporary mechanical ventricular assistance represents an efficacious therapy for acute myocarditis in patients with hemodynamic decompensation despite maximal medical therapy. Failure to achieve full ventricular recovery while on device support still allows for other surgical alternatives, including implantation of a long-term implantable ventricular assist device, or cardiac transplantation.

Selective anticoagulation with active site-blocked factor IXA suggests separate roles for intrinsic and extrinsic coagulation pathways in cardiopulmonary bypass.

BACKGROUND: Multiple stimuli converge in cardiopulmonary bypass to create a tremendous prothrombotic stimulus. The ideal anticoagulant for cardiopulmonary bypass should selectively target only the intravascular stimuli, thereby eliminating pathologic clotting in the bypass circuit while preserving hemostasis in the thoracic cavity. We propose the inhibition of factor IX as such a targeted anticoagulant strategy. METHODS: We prepared an inhibitor of activated factor IX and applied it to a primate model of cardiopulmonary bypass to confirm the anticoagulant efficacy of activated factor IX in this setting and to assess more subtle markers of thrombin generation, macrophage procoagulant activity, and cellular tissue factor expression. Seven baboons that received activated factor IX (460 microg/kg) and 7 that received heparin (300 IU/kg) and protamine underwent cardiopulmonary bypass for 90 minutes and were followed after the operation for 3 hours. RESULTS: Analysis of plasma factor IX activity demonstrated adequate inhibition (<20%) of factor IX throughout cardiopulmonary bypass. Activated factor IX-treated baboons demonstrated similar circuit patency to heparin-treated baboons but had significantly diminished intraoperative blood loss. Preservation of extravascular hemostasis was further demonstrated in activated factor IX-treated animals by (1) significantly increased levels of thrombin-antithrombin III complex and prothrombin activation peptide (F1+2) without intravascular thrombosis, (2) significantly greater macrophage procoagulant activity in pericardial-derived monocytes, and (3) immunohistochemical evidence of tissue factor expression in pericardial mesothelial cells and macrophages. CONCLUSIONS: Anticoagulation with activated factor IX allows for intravascular anticoagulation with maintenance of extravascular hemostasis. These findings suggest activated factor IX as an agent that not only exemplifies a targeted approach to selective anticoagulation in cardiac surgery but also further characterizes the procoagulant milieu during cardiopulmonary bypass.

Heparinless cardiopulmonary bypass with active-site blocked factor IXa: a preliminary study on the dog.

OBJECTIVE: Cardiopulmonary bypass is a potent stimulus for activation of procoagulant pathways. Heparin, the traditional antithrombotic agent, however, is often associated with increased perioperative blood loss because of its multiple sites of action in the coagulation cascade and its antiplatelet and profibrinolytic effects. Furthermore, heparin-mediated immunologic reactions (that is, heparin-induced thrombocytopenia) may contraindicate its use. Cardiopulmonary bypass with a selective factor IXa inhibitor was tested to see whether it could effectively limit bypass circuit/intravascular space thrombosis while decreasing extravascular bleeding, thereby providing an alternative anticoagulant strategy when heparin may not be safely administered. METHODS: Active site-blocked factor IXa, a competitive inhibitor of the assembly of factor IXa into the factor X activation complex, was prepared by modification of the enzyme's active site by the use of dansyl glutamic acid-glycine-arginine-chlormethylketone. Twenty mongrel dogs (five were given standard heparin/protamine; 15 were given activated site-blocked factor IXa doses ranging from 300 to 600 microg/kg) underwent 1 hour of hypothermic cardiopulmonary bypass, and blood loss was monitored for 3 hours after the procedure. RESULTS: Use of activated site-blocked factor IXa as an anticoagulant in cardiopulmonary bypass limited fibrin deposition within the extracorporeal circuit as assessed by scanning electron microscopy, comparable with the antithrombotic effect seen with heparin. In contrast to heparin, effective antithrombotic doses of activated site-blocked factor IXa significantly diminished blood loss in the thoracic cavity and in an abdominal incisional bleeding model. CONCLUSION: These initial studies on the dog suggest that administration of activated site-blocked factor IXa may be an effective alternative anticoagulant strategy in cardiopulmonary bypass when heparin is contraindicated, affording inhibition of intravascular/extracorporeal circuit thrombosis with enhanced hemostasis in the surgical wound.

Alternative technique of right-sided outflow cannula insertion for right ventricular support.

Right ventricular assist devices are an important part of the armamentarium of cardiac surgeons for the treatment of right-sided circulatory failure after cardiac transplantation or insertion of a left ventricular assist device. However, right ventricular assist device insertion can be technically challenging in the setting of pulmonary hypertension because of a number of concomitant anatomic and physiologic phenomena. We present a technique for the insertion of the right ventricular assist device outflow cannula that is easier and faster to insert, and safer to explant, especially if cardiopulmonary bypass is to be avoided.

Selective anticoagulation with active site blocked factor IXa in synthetic patch vascular repair results in decreased blood loss and operative time.

Heparin has been the mainstay of anti thrombic therapy in arterial repair procedures. With increasing use of synthetic patch angioplasty (polytetrafluoroethylene [PTFE] or Dacron, Medical Products, Flagstaff, AZ) to improve long-term patency and limit aneurysmal dilation, however, the use of heparin has been associated with excessive needle hole bleeding, resulting in time delay in the operating room to achieve hemostasis, as well as clinically significant blood loss. Because of the multiple sites of action of heparin in the coagulation cascade, both intravascular (desired effect) and extravascular (untoward side effect) hemostasis are impaired. The authors therefore tested the hypothesis that selective inhibition of intravascular coagulation, without significant impairment of extravascular hemostasis, would prevent clotting intraluminally while preserving hemostasis at the suture line of the patch graft. The unique position of factor IX/IXa in the coagulation cascade renders its inhibition an ideal target in this setting. The authors prepared active site blocked factor IXa (IXai) using dansyl-Glu-Gly-Arg chloromethylketone, and tested this hypothesis in a New Zealand rabbit aortotomy model with PTFE patch closure using either heparin (25 i.u./kg; n = 16) or IXai (300 micrograms/kg; n = 21). The infrarenal aorta was identified and isolated, the anti coagulant infused, aortic cross clamp placed, and aortotomy repaired with a 2 x 6 mm PTFE patch. After cross-clamp removal, blood loss was measured and time to hemostasis was recorded. Compared with heparin, IXai resulted in significantly reduce blood loss (6.97 +/- 4.4 g vs 2.72 +/- 2.51 g, respectively, p < 0.008), and time to hemostasis (2.94 +/- 0.77 min vs 2.0 +/- 0.63 min, respectively, p < 0.003). To assess long-term patency and thrombosis, 12 rabbits (given heparin; n = 6 and IXai; n = 6) were observed for up to 2 months post-operatively. No differences were observed between rabbits treated with heparin or IXai; 100% of the grafts were patent with no differences in degree of intimal hyperplasia by histologic analysis. Together, these data suggest that use of IXai in PTFE vascular repair will safely allow realization of the benefits of long-term patency and decreased aneurysmal dilatation, while eliminating the intraoperative morbidity of needle hole bleeding.

Angiogenesis: a possible mechanism underlying the clinical benefits of transmyocardial laser revascularization.

While clinical reports indicate that significant relief of angina is achieved with transmyocardial laser revascularization (TMLR), the mechanisms of benefit are still a matter of considerable controversy. Studies in our laboratory, as well as in the laboratories of other investigators, have challenged the classic hypothesis that benefits are derived from blood flow through chronically patent channels. While several alternatives have been proposed, our work has focused on investigating the possibility that TMLR stimulates vascular growth in the region around the TMLR channels. We have performed studies looking at histologic markers of vascular growth (including vessel counting and cellular proliferation assays) in order to test this hypothesis, the results of which are reviewed. In brief, we find that TMLR markedly enhances myocardial vascular growth above what is seen normally in ischemic myocardium. We hypothesize that the underlying mechanism relates to liberation of growth factors by inflammatory cells, which are recruited in response to the laser induced myocardial injury. Clarification of whether this mechanism contributes to observed clinical benefits is of fundamental importance, since such understanding may suggest means of enhancing the process.

Activation of coagulation and fibrinolytic pathways in patients with left ventricular assist devices.

Left ventricular assist devices have provided successful supportive therapy for patients awaiting cardiac transplantation for extended periods of time. Although thromboembolic events have complicated support with these devices, the HeartMate left ventricular assist device developed by Thermo Cardiosystems, Inc., Woburn, Massachusetts, was specifically designed with a textured blood-contacting surface to minimize this risk. Clinical experience with this device has been encouraging, inasmuch as minimal thromboembolic complications have occurred despite the absence of anticoagulation. The coagulation and fibrinolytic pathways in these individuals were investigated to better understand the hematologic status of patients treated with the Thermo Cardiosystems device. Despite apparently normal prothrombin and activated partial thromboplastin times, as well as platelet counts, evidence of significant thrombin generation and fibrinolysis was present. To eliminate underlying cardiac failure as the responsible factor for these abnormalities, we made similar measurements in patients with end-stage heart failure who were not supported by an assist device or anticoagulation. These measurements revealed no evidence of thrombin generation or fibrinolysis. These data demonstrate that patients supported with a left ventricular assist device, while successfully sustained without systemic anticoagulation, nevertheless have evidence of activation of coagulation. These phenomena appear to be related to the presence of the device rather than to the underlying cardiac abnormalities. Although procoagulant and fibrinolytic pathways are apparently balanced in these patients, these data underscore the potential for the development of bleeding or thrombosis in clinically relevant settings.

Multiple organ failure after liver transplantation.

OBJECTIVE: To examine the effect of multiple organ failure after liver transplantation on mortality and resource utilization. DESIGN: Retrospective cohort study. SETTING: Surgical intensive care unit in a tertiary care university hospital. PATIENTS: Consecutive series of 113 adults undergoing liver transplantation between 1984 and 1992. Patients were excluded if they died intraoperatively (n = 2), required retransplantation (n = 8), or had incomplete records (n = 7). INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We prospectively developed definitions for organ failure, and quantitated the frequency and related outcomes for mortality and resource utilization. Multiple organ failure was defined as the presence of two or more organ failures. Patients were grouped according to the presence (n = 31) or absence (n = 65) of multiple organ failure. Preoperative severity of illness was assessed by the Acute Physiology and Chronic Health Evaluation (APACHE II) and United Network for Organ Sharing (UNOS) scoring systems. Postoperative outcome data, including hospital survival rate, hospital length of stay, and charges were recorded. The frequency of multiple organ failure after liver transplantation was 32%. The mortality rate in the patients who developed multiple organ failure was 42% vs. only 2% in those patients without multiple organ failure (p < .0001). Patients with four or more organ failures had a 100% mortality rate. Postoperative multiple organ failure was associated with increased hospital length of stay (46 +/- 7 days vs. 29 +/- 2 days; p = .026) and increased hospital charges ($271,497 +/- 29,994 vs. $136,372 +/- 8,310; p < .0001). Higher preoperative APACHE II and UNOS scores predicted postoperative multiple organ failure, but were less accurate tools for predicting risk of death. CONCLUSIONS: Multiple organ failure is associated with death and increased resource utilization in liver transplantation. Pretransplantation severity of illness, as measured by APACHE II and UNOS scoring systems, is an important determinant of postoperative multiple organ failure and outcome.

Hemodynamic correlates of outcome in patients undergoing orthotopic liver transplantation. Evidence for early postoperative myocardial depression.

OBJECTIVE: To describe the hemodynamic and oxygen transport patterns in survivors and nonsurvivors following liver transplantation (LT) and to assess their relationship to organ failure and mortality. DESIGN: Retrospective cohort. SETTING: Surgical ICU in a tertiary care university teaching hospital. PATIENTS: Consecutive series of 113 adults undergoing LT between 1984 and 1992. Patients were excluded if they died intraoperatively (n = 2), required retransplantation (n = 8), or their records were incomplete (n = 7). MEASUREMENTS AND MAIN RESULTS: Preoperative severity of illness was assessed by the acute physiology and chronic health evaluation (APACHE) II scoring system. Hemodynamic and oxygen transport variables were recorded immediately preoperatively and sequentially every 12 h during the first 2 postoperative days. Organ failures (pulmonary, renal, cardiovascular, hepatic, and central nervous system) were assessed for patients in the postoperative period. Patients were grouped as survivors (n = 82) or nonsurvivors (n = 14) with a mortality rate of 15%. Preoperative APACHE II scores were significantly lower in survivors compared with nonsurvivors (7 +/- 0 vs 11 +/- 2; p = 0.029). Both preoperatively and postoperatively, survivors sustained a relatively higher mean arterial pressure, stroke volume index, left ventricular stroke work index, cardiac index, and oxygen delivery as compared with nonsurvivors (p < 0.01). The postoperative decline in systemic blood flow that was seen in both groups was particularly prominent in nonsurvivors during the first 12 h following LT (p < 0.03). Nonsurvivors sustained an approximately fivefold increase in the rate of organ failure (p < 0.0001); all patients (n = 6) with 4 or more organ failures died. CONCLUSION: Nonsurvivors of LT have less cardiac reserve pretransplant; postoperatively, they demonstrate early myocardial depression and subsequently lower levels of cardiac index and oxygen delivery. Patients who develop these hemodynamic patterns are more prone to organ failure and death.