Tumor markers in colorectal cancer, gastric cancer and gastrointestinal stromal cancers: European group on tumor markers 2014 guidelines update.

Biomarkers currently play an important role in the detection and management of patients with several different types of gastrointestinal cancer, especially colorectal, gastric, gastro-oesophageal junction (GOJ) adenocarcinomas and gastrointestinal stromal tumors (GISTs). The aim of this article is to provide updated and evidence-based guidelines for the use of biomarkers in the different gastrointestinal malignancies. Recommended biomarkers for colorectal cancer include an immunochemical-based fecal occult blood test in screening asymptomatic subjects ≥50 years of age for neoplasia, serial CEA levels in postoperative surveillance of stage II and III patients who may be candidates for surgical resection or systemic therapy in the event of distant metastasis occurring, K-RAS mutation status for identifying patients with advanced disease likely to benefit from anti-EGFR therapeutic antibodies and microsatellite instability testing as a first-line screen for subjects with Lynch syndrome. In advanced gastric or GOJ cancers, measurement of HER2 is recommended in selecting patients for treatment with trastuzumab. For patients with suspected GIST, determination of KIT protein should be used as a diagnostic aid, while KIT mutational analysis may be used for treatment planning in patients with diagnosed GISTs.

Tumor markers in pancreatic cancer: a European Group on Tumor Markers (EGTM) status report.

Pancreatic ductal adenocarcinoma is one of the most difficult malignancies to diagnose and treat. The aim of this article is to review how tumor markers can aid the diagnosis and management of patients with this malignancy. The most widely used and best validated marker for pancreatic cancer is CA 19-9. Inadequate sensitivity and specificity limit the use of CA 19-9 in the early diagnosis of pancreatic cancer. In non-jaundiced patients, however, CA 19-9 may complement other diagnostic procedures. In patients with resectable pancreatic cancer, presurgical and postresection CA 19-9 levels correlate with overall survival. In advanced disease, elevated pretreatment levels of CA 19-9 are associated with adverse patient outcome and thus may be combined with other factors for risk stratification. Most, but not all, reports indicate that serial levels of CA 19-9 correlate with response to systemic therapy. Use of CA 19-9 kinetics in conjunction with imaging is therefore recommended in monitoring therapy. Although several potential serum and tissue markers for pancreatic cancer are currently undergoing evaluation, none are sufficiently validated for routine clinical use. CA 19-9 thus remains the serum pancreatic cancer marker against which new markers for this malignancy should be judged.

[Diagnosis and multimodal therapy for hepatocellular carcinoma]. / Diagnose und multimodale Therapie des hepatozellulären Karzinoms.

Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related death in the world. The majority of HCCs develops on the basis of a chronic liver disease. This often complicates diagnosis and therapy. Non-invasive diagnostic criteria are based on dynamic imaging techniques and the serum level of AFP (alpha-fetoprotein). When evaluating HCC patients for therapy, besides tumor burden and localisation, the therapeutic evaluation must also consider the general condition of the patient and his/her liver function. For this purpose, the BCLC algorithm of the Barcelona Clinic for Liver Disease has proven helpful. Only one-third of the patients can be cured by resection, transplantation or local tumour ablation. In locally advanced cases transarterial procedures including transarterial chemoembolisation and radioembolisation are applied. HCC is a chemo-resistant tumour and chemotherapy is not accepted as standard of care in HCC. Sorafenib is the first systemic treatment with proven efficacy approved for the treatment of advanced and metastatic HCC. Interdisciplinary management of HCC patients is essential in order to provide every patient with the optimal therapy at his specific stage of disease.

Tumour markers in colorectal cancer: European Group on Tumour Markers (EGTM) guidelines for clinical use.

The aim of this article is to present updated guidelines for the use of serum, tissue and faecal markers in colorectal cancer (CRC). Lack of specificity and sensitivity preclude the use of all existing serum markers for the early detection of CRC. For patients with stage II or stage III CRC who may be candidates for either liver resection or systemic treatment should recurrence develop, CEA should be measured every 2-3 months for at least 3 years after diagnosis. Insufficient evidence exists to recommend routine use of tissue factors such as thymidylate synthase, microsatellite instability (MSI), p53, K-ras and deleted in colon cancer (DCC) for either determining prognosis or predicting response to therapy in patients with CRC. Microsatellite instability, however, may be used as a pre-screen for patients with suspected hereditary non-polyposis colorectal cancer. Faecal occult blood testing but not faecal DNA markers may be used to screen asymptomatic subjects 50 years or older for early CRC.

Changes of thymidine kinase (TK) during adjuvant and palliative chemotherapy.

UNLABELLED: Thymidine kinase is involved in nucleic acid synthesis and is, therefore, considered to be an important proliferation tumor marker. For this reason, we monitored this marker in the course of colorectal cancer chemotherapy. MATERIALS AND METHODS: We examined thymidine kinase (TK) levels in 30 patients with colorectal cancer who underwent adjuvant or palliative chemotherapy (CHT schemes). The condition for being included in the study was a minimum of 3 cycles of chemotherapy. TK was always assessed with radio-receptor analysis, before and after every chemotherapy cycle, together with other tumor markers. RESULTS: From the monitored tumor markers, only TK changed typically in the course of chemotherapy. In adjuvant chemotherapy, it was mostly low at the beginning of the cycle and its values usually increased considerably at the end. On the other hand, in palliative chemotherapy the dynamics of TK varied mainly depending on the effect of the therapy. Other tumor markers showed nonstandard behavior and minimal correlation with TK changes. CONCLUSION: Thymidine kinase seems to be a suitable parameter for monitoring the effect of adjuvant and palliative chemotherapy in colorectal cancer.

Clinical relevance of tumor markers for the control of chemotherapy.

This presentation is based on our experience with tumor marker monitoring of surgery therapy and chemotherapy effects. The control of chemotherapy is one of the most important problems in oncological practice. The correlation between the clinical status of the patient and tumor size changes, based on the results of different imaging methods, has been the most important and most frequently used method. However, the therapy effect has been recently assessed by markers of the biological activity of the tumor. Using tumor markers for the assessment of the effect of surgery therapy is already part of routine practice in many different types of cancer. Pre-operative and post-operative values of tumor markers are essential for a proper evaluation. However, tumor marker monitoring of the effect of radiotherapy and chemotherapy has been used very rarely, mostly only for research purposes. Besides monitoring by classical tumor markers, monitoring by markers of angiogenesis and apoptosis seem to be promising for the assessment of chemotherapy effect. Measurement of circulating cancer cells by means of mRNA also seems to be intriguing for therapy effect control and monitoring of the course of disease. Unfortunately, the routine use of these methods has been applied in only a few institutes worldwide. A completely different situation has been observed in palliative treatment. In most cases, changes of serum levels of tumor markers correlate with therapy effect. Thus, the effect of treatment on tumor proliferation can be successfully estimated by decreasing tumor marker levels.

Does serum tumor marker half-life complement pretreatment risk stratification in metastatic nonseminomatous germ cell tumors?

The goal of this study was to determine whether the serum tumor marker half-life (MHL) of human chorionic gonadotropin (HCG) and alpha-fetoprotein (AFP) during initial chemotherapy can complement pretreatment risk stratification in metastatic nonseminomatous germ cell tumors. One hundred forty-seven patients were assessable for MHL during the first two cycles of platinum-based chemotherapy. MHL calculation was based on two consecutive values using Kohn's apparent half-life formula (MHL =ln 1/2/G, where G was the gradient of the marker slope) or on three (or more) values using simple linear regression. MHL was regarded as prolonged if it was more than 3.5 days for HCG or more than 7 days for AFP. The median MHL for HCG was 2.8 days (range, 0.7-16.7) and for AFP was 6.2 days (range, 2. 6-65.4). Thirty-five of 108 patients (32%) had a prolonged MHL for HCG, 41 of 114 (36%) had a prolonged MHL for AFP, and in 59 of 147 patients (40%), either or both MHLs were prolonged. If patients with both MHLs normal were compared against patients with either or both MHLs prolonged, highly significant differences in progression-free survival (P < 0.0001) and overall survival (P = 0.0005) were demonstrated. The test accuracy was 70% for both progression-free and overall survival, and it was slightly greater than the overall predictive value of the Medical Research Council prognostic classification. A combination of Medical Research Council criteria and MHL analysis allowed us to refine prognostic assessment. Because MHL analysis is able to complement pretreatment risk stratification and can support selection of patients for early-dose intensified chemotherapy, it should be included in prospective clinical trials for patients with poor-prognosis disease.

Prognostic implications of tumour marker analysis in non-seminomatous germ cell tumours with poor prognosis metastatic disease.

86 unselected patients with poor risk metastatic non-seminomatous germ cell tumours (NSGCT) treated from 1979 to 1990 at a single institution were reviewed with regard to the prognostic relevance of tumour marker analysis. The number of elevated tumour markers was not able to distinguish patients into prognostic subgroups. Pretreatment levels of human chorionic gonadotropin (HCG), alpha-fetoprotein (AFP) and lactate dehydrogenase (LDH) did not have a significant influence on clinical outcome. HCG and AFP half-life analysis during the first chemotherapy cycles also failed to define prognostic subgroups. If early deaths within 90 days after the onset of chemotherapy were excluded, patients with a half-life of HCG decline greater than 3.5 days tended to have a poorer prognosis which did not reach significance.

Clinical utility of biochemical markers in colorectal cancer: European Group on Tumour Markers (EGTM) guidelines.

In recent years, numerous serum and cell/tissue-based markers have been described for colorectal cancer (CRC). The aim of this article was to provide guidelines for the routine clinical use of some of these markers. Lack of sensitivity and specificity preclude the use of any available serum markers such as carcinoembryonic antigen (CEA), CA 19-9, CA 242, CA 72-4, tissue polypeptide antigen (TPA) or tissue polypeptide-specific antigen (TPS) for the early detection of CRC. However, preoperative measurement of CEA is desirable as this may give independent prognostic information, help with surgical management and provide a baseline level for subsequent determinations. For patients with stage 2 (Dukes' B) and 3 (Dukes' C) disease who may be candidates for liver resection, CEA levels should be measured every 2-3 months for at least 3 years after diagnosis. For monitoring treatment of advanced disease, CEA should also be tested every 2-3 months. Insufficient evidence is presently available to recommend the routine use of other serum markers for monitoring purposes. Similarly, the new cell and tissue-based markers (e.g, ras, P53) cannot yet be recommended for routine clinical use.

Immunological comparison of ovarian and colonic CEA.

Ovarian and colonic CEA were compared immunologically by means of antisera prepared against each of them. CEAs of both origins were found identical by immunodiffusion methods. In radioimmunological experiments, slight differences were observed between some but not all ovarian CEAs and colonic CEAs and also between different preparations of colonic CEA: no organ specificity of ovarian CEA could be demonstrated. Finally, CEA level was measured in 41 sera of patients with ovarian carcinoma by two radioimmunoassays, one using colonic CEA as tracer and standard and anti-colonic CEA serum, the other using ovarian CEA and anti-ovarian CEA serum: the values given by the two assays were highly correlated (rs = 0.8107), meaning that an organ specific assay for ovarian CEA is not needed.

Identification of membrane antigens in granulocytes and colonic carcinoma cells by a monoclonal antibody specific for biliary glycoprotein, a member of the carcinoembryonic antigen family.

The gene coding for 'biliary glycoprotein (BGP)' is a member of the carcinoembryonic antigen (CEA) gene family. A monoclonal antibody (MAb) was induced against a BGP-preparation isolated from human bile. The antibody did not crossreact with the carcinoembryonic antigen (CEA) and different non-specific crossreacting antigens. The anti-BGP MAb was used to identify BGP-related antigens in membrane extracts from granulocytes and the colonic carcinoma cell line HT-29. In granulocyte membranes, a single antigen of Mr 160,000 was bound. In membranes from HT-29 cells, a main antigen of Mr 85,000 was present. At high antigen concentration, an additional antigen of Mr 115,000 was identified. Since several transcripts of the BGP gene have been identified, the different BGP related antigens are probably products of alternatively spliced mRNAs.

Diagnostic significance of carcinoembryonic antigen in the differential diagnosis of malignant mesothelioma.

The histologic and cytologic distinction of malignant mesothelioma from carcinomas metastatic to the pleura or peritoneum is often problematic. For this reason immunologic methods are being increasingly used as diagnostic adjuncts. This review summarizes 40 studies on the expression of carcinoembryonic antigen in mesotheliomas and in lung and other carcinomas involving the pleura or peritoneum. Carcinoembryonic antigen was identified immunohistochemically in 11% of mesotheliomas and in 84% of carcinomas examined and immunocytochemically (in serous effusions) in 4% and 58%, respectively. In serum and in pleural or ascitic fluid, significantly elevated levels of carcinoembryonic antigen are commonly associated with (lung) carcinomas but rarely with mesotheliomas. Thus, together with identification of the antigen in serum, pleural fluid, or ascitic fluid, immunohistochemical and immunocytochemical techniques for detecting carcinoembryonic antigen provide a valuable aid for distinguishing malignant mesothelioma from metastatic carcinomas.

Characterization of myeloma cells by means of labeling index, bone marrow histology, and serum beta 2-microglobulin.

The relationship between two cellular prognostic parameters of multiple myeloma, the plasma cell labeling index (LI%) and bone marrow histology was studied. The LI% as the percentage of monoclonal plasma cells in the S-phase was measured by bromodeoxyuridine incorporation using the anti-bromodeoxyuridine antibody BU-1. The histologic classification was based on six plasma cell types that allow prognostic grading as the Marschalko, small cell, cleaved, polymorphous, asynchronous, and blastic types. The biopsies also were used for estimating the degree of infiltration. Beta 2-microglobulin (IMx assay, Abbott, North Chicago, IL), the most significant serum parameter for myeloma also was measured for comparison. Bone marrow specimens and sera were obtained from 50 myeloma patients. Fourteen patients with smoldering myeloma were characterized by well-differentiated Marschalko or small cell type cells, a low LI% and a low beta 2-microglobulin concentration. Considering all myeloma patients, plasma cell type and degree of infiltration showed a significant correlation with LI% and beta 2-microglobulin concentration. Patients with plasma cells that were not mature cells of the Marschalko or small cell type revealed a high LI% or high beta 2-microglobulin level in 16 of 19 cases. However, a high LI% or high beta 2-microglobulin level could be detected in only 8 of 31 patients with plasma cells of the Marschalko or small cell type. Three of 21 stage I patients did not show the typical finding of a low LI%, low beta 2-microglobulin level, and a favorable grade. Stage II patients were not uniformly characterized by LI%, beta 2-microglobulin and plasma cell morphology. The percentage of nucleolated plasma cells was not associated with the LI%. Bone marrow histology, LI%, and beta 2-microglobulin concentration appear to be supplementary prognostic factors. If LI% and beta 2-microglobulin levels are not measured, a higher risk could be overlooked in cases with mature plasma cells.

Tumour associated antigens in diagnosis of serous effusions.

The use of tumour associated antigens in the diagnosis of serous effusions was studied in 76 patients with benign and 200 patients with malignant disease. Tissue polypeptide antigen (TPA), alpha fetoprotein, and CA 125 were found to be of little value. At cut off points of 3 ng/ml, 10 U/ml, and 30 U/ml, respectively, carcinoembryonic antigen (CEA), biliary glycoprotein I (BGP I), and CA 19-9 discriminated between benign and malignant serous effusions with a sensitivity of between 24% and 67%. The immunocytochemical staining for these markers resulted in malignant cells being detected in 18% to 33% of cases. Various combinations of conventional cytological examination, effusion fluid tumour marker determination, and immunocytochemical analysis identified malignant cells in serous effusions in up to 72% of cases; conventional cytology alone detected tumour cells in only 30%.

Plasma cell proliferation in monoclonal gammopathies: measurement using BU-1 antibody in flow cytometry and microscopy: comparison with serum thymidine kinase.

AIMS: The labelling index as defined by the percentage of bone marrow plasma cells doubling their DNA in the S phase is a useful prognostic factor in multiple myeloma. The aim of this study was to examine the specificity and sensitivity of a new flow cytometric method for measuring the labelling index. METHODS: Bone marrow specimens from five patients with monoclonal gammopathy of undetermined significance and 61 patients with multiple myeloma were investigated. The labelling index (LI%) was determined by means of a microscopic and flow cytometric method using the antibromodeoxyuridine antibody BU-1. Serum thymidine kinase, another index of proliferation, was measured by radioimmunoassay. RESULTS: Good comparability (r = 0.83) and nearly equal imprecision (CV < 20%) were found with microscopic and flow cytometric methods of LI% measurement. However, 1000 or more cells had to be counted by microscopy around the cutoff value to avoid an unacceptable imprecision. Plasma cells with increased S phase (LI% > 1%) were characterised by their reduced light chain fluorescence intensity ratio between plasma cells and nonspecifically stained cells (7.9 v 14.8, p < 0.002), that is, by their generally lowered cytoplasmic immunoglobulin content. There was a moderate correlation between thymidine kinase and labelling index (r = 0.56, p < 0.001). At 100% specificity, myelomas with proliferating plasma cells were more sensitively detected by the labelling index than by serum thymidine kinase (55% v 32% sensitivity). CONCLUSIONS: The labelling index represents a more specific and sensitive proliferation marker than serum thymidine kinase. Flow cytometry does not result in greater precision.

AFP isoforms and their clinical significance (overview).

AFP has maintained great clinical relevance as target tumor marker of germ cell tumors (YST = yolk sac tumors) and hepatocellular cancer (HCC) besides its occasional elevation in other malignancies. Its organ and tumor specificity is further limited by elevation in benign liver diseases (BLD) and in pregnancy. The binding of various lectins to its carbohydrate chain (5%) in addition to mere isoelectric focussing has been successfully used in the differentiation of AFP source and disease. The primarily used lectins in affinity electrophoresis and chromatography are Concanavalin A and lens culinaris agglutinin (LCA). The inclusion of other lectins (e.g. erythroagglutinating E-PHA, allomyrina dich. aggl. A lectin) and a more practical nomenclature have been used for characterization of 4 reactive patterns of cord serum/liver/BLD, HCC, g.i. tumor and YST type. In particular, data are reviewed concerning the clinically important liver AFP differentiation by means of LCA and E-PHA differentiation. In conclusion the investigation of AFP lectin binding in cases unexplained AFP elevation may be helpful in the differentiation of AFP source and malignancy.

Value of HAMA--determination in clinical practice--an overview.

HAMAs (human anti mouse antibodies) in serum of patients may be stimulated as an immunologic reaction to the application of animal protein. They are differentiated into heterophilic (species-unspecific) and mouse-specific anti-isotypic (often IgG1) or-idiotypic antibodies as well as according to the human immunoglobulin type (IgG or IgM). Besides infrequent clinically immunologic side effects (anaphylactic reaction), their influence on tumor marker measurements predominates in the follow-up of tumor patients by the possible impairment of newly applied monoclonal antibodies. The detection of HAMAs is performed on suspicion of a tumor by self-made or commercial tests, however, there is no consensus standardization of tests, standards and controls. Experiences are reported herein on the occurrence of heterophilic antibodies, reasons of suspicion, detection and removal, comparative determination of HAMAs by two commercially available tests (ImmunoSTRIP HAMA, Enzygnost) and HAMA induction in patients following immunoscintigraphy (OC-125, n = 27). Besides the rare appearance of HAMAs in patients without pretreatment or following biologic therapy, i.v. application of monoclonal antibodies for diagnosis (immunoscintigraphy) or therapy represents the most frequent reason for HAMA development that should always be ruled out in case of anamnestic indication and clinically unexplained tumor marker changes.

Use of serum PIVKA-II (DCP) determination for differentiation between benign and malignant liver diseases.

BACKGROUND: The concentration of Des-gamma-carboxy-prothrombin (DCP) or PIVKA-II has been described to be increased in patients with hepatocellular cancer, along with its elevation in vitamin K deficient states by warfarin or dicoumarol treatment. The aim of the study was to investigate its clinical value in HCC in comparison with alpha-fetoprotein. PATIENTS AND METHODS: Measurements were performed in duplicate in serum of 87 patients with benign (acute/chronic hepatitis B/C/autoimmune, liver cirrhosis B/C/alcoholic) and 154 patients with highly probable (CT, MRT imaging) or histologically proven HCC. Two commercial or research ELISA tests (1: Eitest MonoP-II, Eisai, Tokyo, Japan; 2: Asserachrom PIVKA-II, Stago, France) using murine monoclonal anti-PIVKA-II antibodies were used comparatively and compared with a laboratory-developed conventional AFP-RIA. RESULTS: By ROC analysis, a highly significant discrimination (p < 0.0001) was found at cutoffs of 0.09 AU/ml (Eisai) or 0.8 ng/ml (Stago) at a specificity of about 90% (Eisai: s = 78.6%, ppv = 0.92, npv = 0.70; Stago: s = 77.9%, ppv = 0.93, npv = 0.70) compared with the AFP-test at a cutoff of 45 ng/ml (sp = 91%, s = 58.4%, ppv = 0.92, npv = 0.55). A higher significant correlation was seen between both DCP tests in malignant (rS = 0.89, p < 0.0001) than benign groups (rS = 0.41, p < 0.001) and a lower correlation between the AFP and Eisai (rS = 0.27/0.36, p < 0.01) and Stago test for the malignant group (0.16; p < 0.05). CONCLUSION: DCP determination in serum/plasma adds significantly in the discrimination between benign and malignant liver diseases.

Is serum tumor marker half-life a guide to prognosis in metastatic nonseminomatous germ cell tumors?

BACKGROUND: The value of serum tumor marker kinetics of AFP and HCG assessed by marker half-life (MHL) analysis for diagnosis and in the follow-up of patients with nonseminomatous germ cell tumors (NSGCT) is still debated controversally. The aim of this retrospective study was therefore to investigate the influence of serum MHL during the first two cycles of chemotherapy on the long-term outcome in metastatic NSGCT. MATERIAL AND METHODS: 147 patients with at least 2 abnormal marker values > 7 days after start of chemotherapy were investigated for HL analysis (HL cut off 3.5 days for HCG and 7 days for AFP). HCG and AFP determinations were performed by a double monoclonal IRMA (HCG) and conventional RIA (AFP) developed by our laboratory. RESULTS: According to these cut offs 35/108 patients (32.4%) had a prolonged HCG HL and 41/114 patients (36%) a prolonged AFP HL. Patients with either MHL prolonged had a significantly inferior overall survival (OS; 10 year OS 37% vs. 75%, p = 0.0005) and progression-free survival (PFS; 10 year PFS 29% vs. 69%, p < 0.0001) than those with a prolonged HCG MHL (OS; 10 year OS 36% vs. 65%, p = 0.003; 10 year PFS 28% vs. 56%; p = 0.001) and even more than those with a prolonged AFP MHL (10 year OS 39% vs. 70%, p = 0.02; 10 year PFS 32% vs. 56%, p = 0.01). CONCLUSION: The remarkable prognostic information assessed by MHL analysis in this retrospective study merits further confirmation by a prospective study for its appropriateness in selecting patients for high dose chemotherapy.

Multivariate analysis of the prognostic value of CEA and CA 19-9 serum levels in colorectal cancer.

UNLABELLED: Carcinoembrionic Antigen (CEA) and carbohydrate antigen 19-9 (CA 19-9) are the most commonly used tumor-associated antigens in the management of patients with colorectal cancer. The aim of this study was to evaluate the prognostic value of preoperative serum levels of CEA and/or CA 19-9 and the classical prognostic factors (age, sex, tumor infiltration and staging) in 495 patients. PATIENTS AND METHODS: The retrospective study was performed on frozen sera (stored at -70 degrees C) of patients with histologically proven colorectal cancer. Survival function estimates were calculated (Kaplan-Meier). The patients were separated into two groups according to the preoperative marker levels. Cut-off levels calculated at a specificity of 100% versus healthy individuals were used: < 4 ng/mL versus > or = 4 ng/mL for CEA and < 60 U/mL versus > or = 60 U/mL for CA 19-9. Survival curve differences were assessed using the log-rank-test. Mulivariate Cox's proportional hazard regression analysis was performed to examine the association between tumor marker levels and survival time. Classical prognostic factors such as age, sex, tumor infiltration, tumor stage (Dukes' classification) were included as covariants. The mantel-Haenszel method was used to assess the survival rate of patients with colorectal carcinoma and high versus low levels of tumor-associated antigens according to tumor stages. RESULTS: The Dukes' stages (log-rank chi-square = 231.9; p < 0.0001) represent the best prognostic factor besides the preoperative values of CA 19-9 (log-rank chi-square = 162.5). CEA shows a log-rank chi-square of 71.4. Thus, CEA and CA 19-9 can be used to discriminate two groups of patients with significantly different survival times (p < 0.0001). The importance of different parameters in providing additional prognostic information was evaluated by multivariate analysis. Only items of statistically significant prognostic relevance (univariate analysis) were used for this analysis. Estimated relative risks of death adjusted for tumor stage were 5.5 considering Dukes' stage A versus Dukes' stage B/C and Dukes' stage B/C versus Dukes' stage D, respectively, and an increasing relative risk of 27.5 for Dukes' stage A versus Dukes' stage D (p < 0.001). The relative risk for preoperative CA 19-9 serum concentrations (> or = 60 U/mL versus < 60 U/mL) was 2.3 (p < 0.001) and for preoperative CEA concentrations (> or = 4 ng/mL versus < 4 ng/mL) 1.4 (p < 0.07). For CEA the 2-year survival rates in the group of patients with preoperative serum concentrations > 4 ng/mL versus < 4 ng/mL and Dukes' stage D were 16% versus 38%, in Dukes' stage B/C 73% versus 91% and in Dukes' stage A 100% versus 98%. For CA 19-9 the 2-year survival rates in the group of patients with preoperative serum concentrations > or = 60 U/mL versus < 60 U/mL and Dukes' stage D were 10% versus 39%, whilst in Dukes' stage B/C 58% versus 87%. In the group of patients with Dukes' stage A with serum levels > or = 60 U/mL a 2-years survival rate of 100% was found. CONCLUSION: The postoperative Dukes' classification represents the best prognostic information besides the preoperative values of CA 19-9. The predictive information provided by preoperative CA 19-9 serum levels is independent from that obtained by the other factors investigated. Only Dukes' classification and CA 19-9 levels showed statistical significance (p < 0.001).