Liver Preservation by Aortic Perfusion Alone Compared With Preservation by Aortic Perfusion and Additional Arterial Ex Situ Back-Table Perfusion With Histidine-Tryptophan-Ketoglutarate Solution: A Prospective, Randomized, Controlled, Multicenter Study.

BACKGROUND: Arterial ex situ back-table perfusion (BP) reportedly reduces ischemic-type biliary lesion after liver transplantation. We aimed to verify these findings in a prospective investigation. METHODS: Our prospective, randomized, controlled, multicenter study involved livers retrieved from patients in 2 German regions, and compared the outcomes of standard aortic perfusion to those of aortic perfusion combined with arterial ex situ BP. The primary endpoint was the incidence of ischemic-type biliary lesions over a follow-up of 2 years after liver transplantation, whereas secondary endpoints included 2-year graft survival, initial graft damage as reflected by transaminase levels, and functional biliary parameters at 6 months after transplantation. RESULTS: A total of 75 livers preserved via standard aortic perfusion and 75 preserved via standard aortic perfusion plus arterial BP were treated using a standardized protocol. The incidence of clinically apparent biliary lesions after liver transplantation (n = 9 for both groups; P = 0.947), the 2-year graft survival rate (standard aortic perfusion, 74%; standard aortic perfusion plus arterial BP, 68%; P = 0.34), and incidence of initial graft injury did not differ between the 2 perfusion modes. Although 33 of the 77 patients with cholangiography workups exhibited injured bile ducts, only 10 had clinical symptoms. CONCLUSIONS: Contrary to previous findings, the present study indicated that additional ex situ BP did not prevent ischemic-type biliary lesions or ischemia-reperfusion injury after liver transplantation. Moreover, there was considerable discrepancy between cholangiography findings regarding bile duct changes and clinically apparent cholangiopathy after transplantation, which should be considered when assessing ischemic-type biliary lesions.

Influence of Deceased Donor and Pretransplant Recipient Parameters on Early Overall Kidney Graft-Survival in Germany.

Background. Scarcity of grafts for kidney transplantation (KTX) caused an increased consideration of deceased donors with substantial risk factors. There is no agreement on which ones are detrimental for overall graft-survival. Therefore, we investigated in a nationwide multicentre study the impact of donor and recipient related risks known before KTX on graft-survival based on the original data used for allocation and graft acceptance. Methods. A nationwide deidentified multicenter study-database was created of data concerning kidneys donated and transplanted in Germany between 2006 and 2008 as provided by the national organ procurement organization (Deutsche Stiftung Organtransplantation) and BQS Institute. Multiple Cox regression (significance level 5%, hazard ratio [95% CI]) was conducted (n = 4411, isolated KTX). Results. Risk factors associated with graft-survival were donor age (1.020 [1.013-1.027] per year), donor size (0.985 [0.977-0.993] per cm), donor's creatinine at admission (1.002 [1.001-1.004] per µmol/L), donor treatment with catecholamine (0.757 [0.635-0.901]), and reduced graft-quality at procurement (1.549 [1.217-1.973]), as well as recipient age (1.012 [1.003-1.021] per year), actual panel reactive antibodies (1.007 [1.002-1.011] per percent), retransplantation (1.850 [1.484-2.306]), recipient's cardiovascular comorbidity (1.436 [1.212-1.701]), and use of IL2-receptor antibodies for induction (0.741 [0.619-0.887]). Conclusion. Some donor characteristics persist to impact graft-survival (e.g., age) while the effect of others could be mitigated by elaborate donor-recipient match and care.

The joint impact of donor and recipient parameters on the outcome of heart transplantation in Germany after graft allocation.

Organ shortage in heart transplantation (HTx) results in increased use of grafts from donors with substantial risk factors. It is discussed controversially which donor characteristics may be detrimental. Therefore, we evaluated the joint impact of donor- and patient-related risk factors in HTx on patient survival by multiple analysis in a nationwide multicentre study after donor selection was carried out. The research database consists of data concerning hearts donated and transplanted in Germany between 2006 and 2008 as provided by Deutsche Stiftung Organtransplantation and the BQS Institute. Multiple Cox regression (significance level 5%, hazard ratio [95% CI]) was conducted (n = 774, recipient age ≥ 18 years). Survival was significantly decreased by donor age (1.021 [1.008-1.035] per year), nontraumatic cause of death (1.481 [1.079-2.034]), troponin >0.1 ng/ml (2.075 [1.473-2.921]), ischaemia time (1.197 [1.041-1.373] per hour), recipient age (1.017 [1.002-1.031] per year) and in recipients with pulmonary vascular resistance ≥ 320 dyn*s*cm(-5) (1.761 [1.115-2.781]), with ventilator dependency (3.174 [2.211-6.340]) or complex previous heart surgery (1.763 [1.270-2.449]). After donor selection had been conducted, multiple Cox regression revealed donor age, nontraumatic cause of death, troponin and ischaemia time as well as recipient age, pulmonary hypertension, ventilator dependency and previous complex heart surgery as limiting risk factors concerning patient survival.

Extracorporeal tumor cell filtration during extended liver surgery: first clinical use of leukocyte depletion filters--case series.

BACKGROUND: During oncologic surgery, intraoperative manipulation of tumor tissue is almost unpreventable and causes a high risk of tumor cell dissemination into venous blood. A tumor cell-reducing effect of leukocyte adhesion filter systems has been shown under in vitro conditions. METHODS: In a first clinical attempt, leukocyte adhesion filters were integrated into veno-venous bypass systems in four patients undergoing extended liver surgery for secondary hepatic malignancies.Practicability, handling, and safety aspects as well as potency of cell removal and clinical side effects of the filter system were analyzed. RESULTS: All patients tolerated the application of the system without problems during operative and postoperative follow-up. Immunohistochemical staining of perioperative blood samples detected cytokeratin positive (CK+) cells in three cases during the hepatic mobilization. CONCLUSIONS: Effectiveness of CK+ cell depletion and safety of the procedure was shown. The presented surgical technique represents a safe and innovative tool; however, clinical significance has to be examined in a larger patient cohort.

The Donor-Risk-Index, ECD-Score and D-MELD-Score all fail to predict short-term outcome after liver transplantation with acceptable sensitivity and specificity.

BACKGROUND: Expansion of the donor pool by the use of grafts with extended donor criteria reduces waiting list mortality with an increased risk for graft and patient survival after liver transplantation. This study investigates the ability of the Donor-Risk-Index (DRI), the Extended-Criteria-Donor-Score (ECD-score) and the D-MELD-score to predict early outcome after liver transplantation. MATERIAL/METHODS: 291 consecutive adult liver transplants (01.01.2007-31.12.2010) were analysed in a single centre study with ongoing data collection. Primary study endpoints were 30-day mortality, 3-month mortality, 3-month patient and graft survival and the necessity of acute retransplantation within 30 days. For the primary study endpoints ROC-curve analysis was performed to calculate the sensitivity, specificity, and overall model correctness of the Donor-Risk-Index (DRI), Extended-Criteria-Donor-Score (ECD-score) and the D-MELD-Score as predictive models. Cut-off values were selected with the best Youden index. RESULTS: ROC-curve analysis showed areas under the curve (AUROCs) <0.7 for the DRI, the ECD-Score and the D-MELD-Score as models for the prediction of 30-day mortality, 3-month mortality, 3-month patient survival, 3-month graft survival as well as the necessity of acute retransplantation within 30 days after transplantation with unacceptable low levels of overall model correctness (<62%) and specificity (<56%). CONCLUSIONS: The DRI, the ECD-Score and the D-MELD-Score all fail to predict short-term outcome after liver transplantation with acceptable overall model correctness in a current European transplant setting.

The effect of education on the attitude of medical students towards organ donation.

BACKGROUND: The persisting shortage of organs for transplantation could be reduced by increasing the willingness to donate organs. An appropriate education and a positive attitude of medical students and future physicians towards organ donation may have a positive impact on the attitudes of the general public. MATERIAL/METHODS: During the summer semester 2010 we conducted a voluntary educational intervention study concerning organ donation among medical students in the course of the main surgery lecture at the University of Essen, Germany. RESULTS: The survey comprised 94 questionnaires. At the beginning of the lecture there were 67% of organ donor card carriers among the students. An additional 20% imagined they might carry an organ donor card in the future. 37% needed more information regarding organ donation. After the lecture 13% were still not willing to carry an organ donor card in the future. 18% of the students required further information. 42% of the students rated their attitude towards organ donation to be influenced positively by the lecture, 57% of the students stated there was no influence. One student documented a negative influence by the lecture towards organ donation. CONCLUSIONS: Well-directed interventions are needed to sensitize young adults to the topic of organ donation. Better understanding of medical students and future physicians in the field of organ donation will help them to become disseminators for this important topic in our society.

Joint impact of donor and recipient parameters on the outcome of liver transplantation in Germany.

BACKGROUND: The shortage of donor organs in Germany has led to the use of organs from donors with extended donor criteria (EDC). EDC have been defined on the basis of expert opinions, but their clinical relevance is controversial. This may cause loss of organs otherwise available for transplantation. We evaluated the impact of donor and recipient factors in liver transplants on patient and graft survival in a nationwide multicenter analysis, with special focus on EDC and donor risk index. METHODS: A database was created from data on livers donated and transplanted in Germany between 2006 and 2008 as provided by Deutsche Stiftung Organ transplantation and BQS Institute. Cox regression (significance level 5%, risk ratio [95% confidence interval]) was used for calculating the impact on patient survival (n=2095) and on graft survival (n=2175). RESULTS: Patient and graft survival were significantly affected only by donor age (1.012 and 1.011/year), recipient age (1.019 and 1.014/year), creatinine (1.248 and 1.205/mg/dL), bilirubin (1.022 and 1.023/mg/dL), and high urgency status (1.783 and 1.809). Inferior organ quality resulted in lower graft survival (1.243) and donor history of smoking in lower patient survival (1.249). CONCLUSION: Multiple Cox regression revealed no significant impact of EDC or donor risk index on patient and graft survival except for donor age after donor selection at recovery. Among recipient variables, only age, creatinine and bilirubin, and high urgency status were associated with poorer outcome.

Porcine endogenous retrovirus released by a bioartificial liver infects primary human cells.

BACKGROUND: Porcine endogenous retrovirus (PERV) remains a safety risk in pig-to-human xenotransplantation. There is no evidence of in vivo productive infection in humans because PERV is inactivated by human serum. However, PERV can infect human cell lines and human primary cells in vitro and inhibit human immune functions. AIMS: We investigated the potential of primary porcine liver cells to transmit PERV to primary human cells in a bioreactor-based bioartificial liver (BAL). METHODS: Primary human hepatocytes, endothelial cells and the human cell line HEK 293 were exposed to supernatants from BAL or from the porcine cell line PK-15. PERV polymerase-specific reverse-transcriptase polymerase chain reaction (RT-PCR) and PCR were used to investigate PERV transmission to human cells. An assay of RT activity was used to detect the presence of retrovirus in the supernatants of BAL, primary human hepatocytes and endothelial cells. RESULTS: Primary human hepatocytes (hHep), endothelial cells and HEK 293 cells were reproducibly infected by PERV, originating from primary porcine liver cells within the BAL and from PK-15 cells. Infected cells were positive for PERV-specific DNA and RNA after 8-10 days on an average, and RT activity was detectable in the supernatants of infected hHep and HEK 293 cells. CONCLUSION: A risk of PERV infection in human cells is documented in this study, indicating that short-term contact of primary porcine liver cell supernatants with primary human cells could result in PERV transmission.

Survival of patients with synchronous and metachronous colorectal liver metastases--is there a difference?

BACKGROUND: The aim of this study was to compare outcomes in patients with synchronous and metachronous colorectal liver metastases, with special emphasis on prognostic determinants. STUDY DESIGN: We analyzed prospectively collected data on 101 patients with synchronous metastases (group A) who were treated surgically during the time period from April 1998 to December 2006 in regard to overall and disease-free survival, impact of chemotherapy, as well as several serum parameters. A group of patients with metachronous colorectal liver metastases (group B) was considered for baseline comparison. RESULTS: Twenty-three patients in group A received only an explorative laparotomy. Surgical treatment included right hepatectomy (n = 7), left hepatectomy (n = 5), right trisectionectomy (n = 10), left trisectionectomy (n = 1), left lateral resection (n = 11), and sectionectomy (n = 44). Thirty-day mortality was 3%. Morbidity was observed in 10% of the patients. One-, 3-, and 5-year overall survival rates for synchronous metastases were 86%, 68%, and 47%, respectively. The corresponding rates for metachronous metastases were 94%, 68%, and 39% (p > 0.05). Disease free survival was 74%, 42%, and 33% in group A versus 84%, 62%, and 13% in group B (p = 0.28). There was no difference in survival between patients receiving neoadjuvant chemotherapy and no chemotherapy (p > 0.05). Out of all serum parameters, carcinoembryonic antigen levels were a negative predictor for overall and disease-free survival only. CONCLUSIONS: Patients with synchronous colorectal liver metastases had a similar 5-year overall and disease-free survival, which corresponds to patients with metachronous metastases. The impact of neoadjuvant chemotherapy in patients with synchronous metastases needs to be further clarified.

Impact of intraoperative radiotherapy (IORT) on survival of patients with unresectable hilar cholangiocarcinoma.

BACKGROUND/AIMS: The cure or long-term survival of hilar cholangiocarcinoma patients can only be achieved after complete tumor resection. Many patients though suffer from unresectable hilar cholangiocarcinoma, and palliative treatment is therefore the only therapeutic option. In cases of unresectable bile duct cancer, intraoperative radiotherapy (IORT) is an additional option during surgery. The aim of this study was to compare the efficacy of IORT to surgery alone in patients with unresectable hilar cholangiocarcinoma. METHODOLOGY: Palliative IORT (group 1) was performed on 9 patients (4 female/5 male); surgery alone (group 2) was performed in a case-matched group of 9 patients (4 female, 5 male) with unresectable hilar cholangiocarcinoma. The mean ages were 52.9 years (group 1) and 57.2 years (group 2). The two groups had comparable local tumor extension and stages of tumor disease according to UICC 6th edition. Group 1 was also compared to all 36 patients (n=36) suffering from unresectable cholangiocarcinoma treated without IORT (Group 3). RESULTS: The survival of patients after IORT was significantly improved compared to surgery alone in this study. The median survival time was 23.3 months (group 1) versus 9.4 months (group 2) and 5.7 month (group 3). The one year and two year actuarial survival rates are: 56% and 42% (group 1), 33% and 0% (group 2), 25% and 8% (group 3). CONCLUSIONS: Palliative surgery including IORT is safe for patients with unresectable hilar cholangiocarcinoma. The benefit shown by this investigation is a significant improvement of survival time after application of IORT in the palliative situation. A prospective study with randomization is needed to confirm these first results in a larger cohort.

Value of staging laparoscopy for multimodal therapy planning in esophago-gastric cancer.

The importance of multimodal treatment for advanced esophago-gastric cancer has contributed to the development of more accurate preoperative staging strategies. The impact of staging laparoscopy and cytology after conventional staging is evaluated in this study. Staging laparoscopy was performed in 125 patients with potentially resectable cancer of the distal esophagus or gastric cancer. Results were registered separately on a database according to the TNM classification of the International Union Against Cancer (UICC). Laparoscopy changed TNM classification in 46 cases. Explorative laparoscopy resulted in up-staging concerning the N-factor (n = 15) and M-factor (n = 28). Downstaging of the T-factor was recorded in three cases. Cytologic examination gave no additional information in our series. Our experience suggests a clear benefit of laparoscopy in staging of patients with distal esophagus or gastric malignancy. Laparoscopy is a safe and effective staging modality, avoiding unnecessary laparotomies and providing new means of directing appropriate treatment strategy.

Anesthesia for cardiovascular interventions and magnetic resonance imaging in pigs.

Large animal models are still required for many experimental purposes. The aim of the current study was to define a viable narcotic procedure for experimental cardiovascular interventions and imaging in pigs. A total of 32 domestic pigs were used. Animals received propofol, midazolam, and fentanyl as continuous intravenous infusion anesthesia for complex vascular interventions, angiographic X-ray imaging, and magnetic resonance imaging (MRI). Anesthesia was maintained for 6 to 10 h. The initial hourly doses were 2.29 mg/kg of propofol, 1.14 mg/kg of midazolam, and 0.009 mg/kg of fentanyl, with controlled ventilation. Anesthesia, interventions, imaging, periods of apnea of as long as 2 min, and transportation were well-tolerated. Stress-induced arrhythmias were not noted, and artifact-free imaging was achieved. The combination of propofol, midazolam, and fentanyl is well-suited for experimental angiographic interventional studies, experimental cardiovascular MRI, and MR-guided interventions in pigs.

Method of intracranial pressure monitoring and cerebrospinal fluid sampling in swine.

Cerebral oedema and encephalopathy have been noted to occur frequently in patients severely ill or dying after trauma, ischaemia, infections or even metabolic disorders. The objective of the present study was to establish continuous monitoring of the intracranial pressure (ICP) and sampling of cerebrospinal fluid (CSF) for further investigations in swine. ICP monitoring was established in eight pigs by using a ventricular drainage system, implemented after paramedian trepanation of the os frontale. CSF and serum samples were taken for measurement of the levels of glucose and protein. Operating time was 21+/-8 min for the trepanation until ICP monitoring was performed. No complications occurred during surgery. Continuous monitoring of ICP and CSF sampling was easy to perform, and without any side-effects in any animal. At autopsy, no iatrogenic lesions were found and monitoring catheters were still in place. For several types of research requiring ICP monitoring and sampling of CSF, this method can be used successfully.

Split liver transplantation for hepatocellular carcinoma.

BACKGROUND/AIMS: Liver transplantation (LTx) is recognized as the treatment of choice for small hepatocellular carcinomas (HCC) in patients with end-stage liver failure. However, because of limited organ availability, not all those who qualify can benefit from full-size LTx. METHODOLOGY: Over a 28-month period, we transplanted 6 deceased donor split liver allografts in stable recipients with HCC and cirrhosis. RESULTS: There were 5 men and one woman with a median age of 60 years. Median waiting time to LTx was 82 days. Extended right split grafts (segments I, IV-VIII) were transplanted in 5 patients. One patient received a right split graft. In situ harvesting was performed in 4 instances and ex situ in the remaining 2. Median cold ischemia time was 10 hours. Primary non-function was not observed. Median intensive care unit stay was 4 days. There were neither vascular nor biliary complications. The postoperative course was uneventful in half of the patients. Two patients were re-operated because of an abscess caused by a liver necrosis in the marginal zone of segment IV. One patient died for reasons unrelated to liver function or to the surgical intervention. The remaining 5 patients are alive after a median follow-up of 20 months. CONCLUSIONS: Deceased donor split LTx constitutes an additional option for patients with HCC and cirrhosis. The potential risks of using "split livers" as well as the potential benefits of transplanting patients unlikely to survive the waiting list period must be evaluated on an individual basis.

Portopulmonary hypertension in the early phase following liver transplantation.

BACKGROUND: Portopulmonary hypertension (PPH) is a severe complication of liver cirrhosis, which poses a high risk for postliver transplantation (LT) mortality. In most liver transplant centers, severe PPH is viewed as an absolute contraindication for LT, but recent reports challenge this. The purpose of our study was to determine the incidence of PPH, its influence on the 30-day mortality rate following LT and to determine the sensitivity and specificity of Doppler echocardiography and electrocardiography as noninvasive tools to determine PPH. METHODS: We studied 74 consecutive patients that underwent LT between February 2004 and November 2005. Pulmonary arterial pressure and cardiac index were repeatedly determined during surgery and postoperatively. PPH was defined as mild (mean pulmonary arterial pressure (MPAP) 25-35 mm Hg), moderate (MPAP of 35-45 mm Hg) and as severe (MPAP >45 mm Hg). RESULTS: The total incidence of PPH was 31% (16 mild, 5 moderate, and 2 severe). There was a tendency towards increased 30-day mortality rate in patients with PPH compared to controls (22% vs. 12%, P=0.1). However, the two patients with the most severe PPH survived. The duration of ventilation and total stay at the intensive care unit did not differ significantly between groups. The positive predictive value of Doppler echocardiography for PPH was 39% and the negative predictive value 90%. CONCLUSIONS: Mild pulmonary hypertension is common in patients with liver failure, whereas moderate and severe hypertension is not. Severe PPH should not be considered as absolute contraindication for LT.

Does PEEP impair the hepatic outflow in patients following liver transplantation?

OBJECTIVE: Evaluation of the impact of end-expiratory pressure (PEEP) ventilation on venous liver outflow, portal vein, and hepatic artery flows as well as systemic hemodynamics in patients following liver transplantation (LT). DESIGN: Prospective, interventional patient study. SETTING: University hospital intensive care unit. PATIENTS: 65 consecutive patients after LT. INTERVENTIONS: All patients were intubated and mechanically ventilated with biphasic positive airway pressure (BIPAP). The effects of three levels of PEEP (0, 5, and 10 mbar) applied at random order on hepatic inflow and outflow were studied in the immediate postoperative period. MEASUREMENT AND RESULTS: Central venous-, arterial pressure, and cardiac index was recorded from every patient at three different PEEP levels (0, 5, and 10 mbar). Simultaneously, flow velocities in the hepatic-, portal vein, and hepatic artery were determined by Doppler ultrasound. PEEP of 10 mbar significantly increased central venous pressure in comparison with zero PEEP. Mean arterial pressure and cardiac index was not influenced. Hepatic inflow and outflow of the transplanted livers were not impaired by any of the used PEEP levels. CONCLUSIONS: BIPAP ventilation with PEEP levels up to 10 mbar does not affect systemic hemodynamics. Furthermore, neither venous outflow nor portal venous or hepatic artery inflow of the liver are impaired at PEEP levels up to 10 mbar immediately following liver transplantation. Although these results suggest that PEEP ventilation up to 10 mbar does not affect liver hemodynamics, further studies are needed to determine whether these findings could be confirmed for a longer ventilation period with PEEP.

Neurological complications after cadaveric and living donor liver transplantation.

Problems related to the central nervous system have a major impact on survival and quality of life. The aim of this retrospective study was to evaluate the incidence of neurological complications after liver transplantation (LT), including both cadaveric and living donor liver transplantation. Between April 2001 and March 2004 174 patients (120 cadaveric liver transplantations, 54 living donor transplantations) were admitted to our intensive care after liver transplantation. Of the transplanted patients 24.7% developed neurological complications. These patients' stay in the intensive care (14.2 +/- 17.2 days) was much longer than that of all admitted patients (8.4 +/- 10.5 days, p < 0.05). The most common neurological complications were encephalopathy (72.1%) and seizures (11.6 %). The incidence of neurological complications in living donor liver transplanted patients was significantly lower than in cadaveric transplantation patients (20.4% vs 26.7 %). The cold ischemia time in living donor transplanted patients was significantly shorter in comparison with cadaveric transplanted patients (215 +/- 119.3 vs. 383.7 +/- 214.7). The survival rate after liver transplantation of patients with neurological complications was lower than that of patients without, but not significantly different (79.1 % vs. 82.4%, p > 0.05). The incidence of neurological symptoms was found to be similar between the patients treated with cyclosporine (25%) and tacrolimus (23.8 %) in this study. In conclusion, there was a high incidence of neurological complications after LT, prolonging the patients' stay in intensive care significantly. The major neurological manifestation in our patients was encephalopathy followed by seizures. Living donor liver transplantation was associated with a significantly lower incidence of neurological complications compared with patients who had received a cadaveric graft. This might be due to the good quality of the organ and the much shorter cold ischemia time of the graft when the donor was alive.

Tumor cell dissemination in colon cancer does not predict extrahepatic recurrence in patients undergoing surgery for hepatic metastases.

Patients undergoing resection of hepatic metastases of colorectal cancer have a high risk of extrahepatic recurrence, most likely caused by early tumor cell dissemination or the manipulation of liver tumors during surgical resection. Using immunocytochemistry, we studied 47 patients for cytokeratin (CK)-positive (+) cells in: a) bone marrow (BM) samples to determine whether tumor cell dissemination had already occurred before surgery; and b) blood samples directly taken from the hepatic vein before and during surgery of liver metastases. In addition, normal and malignant liver tissues were evaluated for markers known to be involved in tumor progression and metastasis [urokinase plasminogen activator (uPA), Her-2/neu, epidermal growth factor receptor (EGF-R)] using sandwich enzyme immunoassays. CK+ cells were detected in the BM of 26/47 patients (55%), in blood samples of 14/47 patients (30%) before surgery and 11/47 patients (23%) during surgery with a median detection rate of 1 (range, 1-14) CK+ cell per 4x10(6) MNC. No CK+ cells were found in 15/47 patients (32%) in any sample studied. Tumor tissue was obtained from 32/47 patients and normal liver tissue from 24/32 patients. While no differences were found for EGF-R and Her-2/neu, a 9-fold higher expression of uPA could be demonstrated in tumor tissue of 20/32 patients (63%) compared to normal liver tissue. When all obtained results were correlated with clinical outcome, neither the detection of CK+ cells nor the expression pattern in the tumor tissue, or the combination of both, was predictive for extrahepatic recurrence or overall survival after a mean observation time of 43 months (range, 26-54 months). Although uPa is overexpressed in liver metastases of colorectal cancer, and dissemination of CK+ cells during surgery of these metastases is a frequent event in colon cancer, these findings do not predict extrahepatic recurrence. Further characterization of single cells, especially those spread during surgery, will help to identify those patients with an increased risk of later relapse.