Tumor Markers In Patients Research Articles

The prognostic importance of tumor markers in adenocarcinomas of the gastrointestinal tract.

Although no serological tumor marker has yet been shown to be sufficiently sensitive and specific to be used in screening for colorectal, gastric, or pancreatic cancers, elevated pre-operative levels of carcinoembryonic antigen and carbohydrate antigen (CA) 19-9 correlate with advances stages of disease and a poorer clinical outcome. Monitoring of serum carcinoembryonic antigen after primary resection of colorectal cancer may identify a small percentage of asymptomatic patients with recurrent disease who are amenable to a second surgical procedure with curative intent. A new class of tumor markers that recognize cytokeratins, which comprise the intermediate filaments of the cytoskeletons of epithelial cells, is being evaluated as a prognostic factor in colorectal and gastric cancer. Elevated mRNA expression of matrix metalloproteinases and their inhibitors in tumor tissue compared with normal intestinal mucosa predicts for a poor prognosis after primary management of colorectal cancer. CA 19-9, CA 72-4, and CYFRA 21-1 may convey prognostic information in gastric cancer. CA 19-9 appears to be the best overall tumor marker for pancreatic cancer, and a subsequent rise after postoperative normalization may precede clinical detection of recurrent disease. CAM 17.1 is a new marker that has similar sensitivity but better specificity than CA 19-9 in pancreatic cancer. The utility of serial monitoring of these tumor markers in patients with advanced disease is less well established but may become helpful if more effective systemic therapy is developed.

Diagnostic and prognostic value of Cyfra 21-1 compared with other tumour markers in patients with non-small cell lung cancer: a prospective study of 116 patients

The diagnostic value of Cyfra 21-1 in non-small lung cancer (NSCLC) has been established, but few studies have focused on its prognostic value. The aim of this study was to compare that of carcinoembryonic antigen (CEA), carbohydrate antigen (CA) 19-9, CA 125, neuron-specific enolase and squamous cell carcinoma antigen. 116 patients with unresectable ( n = 88) or resectable ( n = 28) NSCLC were prospectively monitored from diagnosis, for a median of 14.4 months. All patients underwent tumour-marker determinations before treatment, then every 3 months. Their diagnostic value was studied using ROC (receiver operating characteristic) curves, based on control measure in 23 patients with benign lung diseases. The prognostic analysis was based on overall survival as the main endpoint. The diagnostic value of Cyfra 21-1 was confirmed, with a sensitivity of 54% and a specificity of 96% at a cut-off value of 3.3 ng/ml. At diagnosis, in the 88 non-surgical NSCLC, besides the presence of metastases ( P = 0.017), Cyfra 21-1 ( P = 0.017) and CA 125 ( P = 0.03) were related to outcome. Elevated levels of Cyfra 21-1 at any time during the disease course was selected by multivariate analysis as additional predictors of poor survival.

Diagnostic and Prognostic Value of Cyfra 21-1 Compared with Other Tumour Markers in Patients with Non-small Cell Lung Cancer: A Prospective Study of 116 Patients

The diagnostic value of Cyfra 21-1 in non-small lung cancer (NSCLC) has been established, but few studies have focused on its prognostic value. The aim of this study was to compare that of carcinoembryonic antigen (CEA), carbohydrate antigen (CA) 19-9, CA 125, neuron-specific enolase and squamous cell carcinoma antigen. 116 patients with unresectable (n = 88) or resectable (n = 28) NSCLC were prospectively monitored from diagnosis, for a median of 14.4 months. All patients underwent tumour-marker determinations before treatment, then every 3 months. Their diagnostic value was studied using ROC (receiver operating characteristic) curves, based on control measure in 23 patients with benign lung diseases. The prognostic analysis was based on overall survival as the main endpoint. The diagnostic value of Cyfra 21-1 was confirmed, with a sensitivity of 54% and a specificity of 96% at a cut-off value of 3.3 ng/ml. At diagnosis, in the 88 non-surgical NSCLC, besides the presence of metastases (P = 0.017), Cyfra 21-1 (P = 0.017) and CA 125 (P = 0.03) were related to outcome. Elevated levels of Cyfra 21-1 at any time during the disease course was selected by multivariate analysis as additional predictors of poor survival.

Circulating tumor markers in gynecological tumors.

The correct use of tumor markers in patients with gynecological cancers can make an impact on management of the disease. Their value in screening for ovarian cancer cannot be recommended outside trials. Increased levels of various markers give prognostic information that could alter therapy for patients with early ovarian or cervical carcinoma. Their main value is in monitoring the response to therapy and confirming relapse, for which they may be superior to scans.

Tumor markers in patients with malignant pleural mesothelioma

Tumor markers in patients with malignant pleural mesothelioma

Selective decrease in serum immunoglobulin G1: A tissue nonspecific tumor marker detecting early stages of gynecologic malignant disease with high efficiency

Malignant diseases of various tissue origin have previously been found to be associated with a characteristic shift in the serum pattern of IgG subclasses, i.e., a highly significant reduction of the percent of IgG1 and an increase of the percentage of IgG2 relative to the total IgG. In the present study we examined the diagnostic performance of this indirect tumor marker in patients with carcinomas of various sites within the female reproductive tract. Using quantitative affinity chromatography, the percents of IgG1 and IgG2 in the total IgG were determined for 207 patients with carcinoma of the ovary, cervix, or corpus uteri, prior to any treatment. The data were compared with those of 135 age matched healthy females and 52 patients with benign gynecologic diseases. It was found that (1) mean values for the percents of IgG1 and IgG2 of all of the cancer patients differed significantly from those of the patients with benign disease and healthy controls; (2) no differences were noted between carcinomas of the ovary, corpus or cervix uteri; (3) early stages of carcinoma exhibited the effect to the same extent as late stages; (4) the specificity of the percent of IgG1 to discriminate between controls and cancer patients ranged between 90 and 100%, regardless of localization and stage of tumor; and (5) whereas with ovarian cancer CA 125 showed a slightly greater sensitivity, the percent of IgG1 was by far more sensitive than the conventional markers CA 125, TPA, CEA, Ferritin, and SCC to diagnose carcinoma of the cervix and corpus uteri, notably at early stages. Combined analysis of the percent of IgG1 and CA 125 and/or TPA led to an increase in sensitivity with tumors of all three sites. Thus, the determination of the percent of IgG1 by itself and/or in combination with conventional markers may provide relevant information regarding the noninvasive detection of early stages of gynecologic carcinoma.

The clinical value of tumour markers in the management of ovarian cancer.

It is now clear that the correct use of tumour markers in patients with epithelial ovarian carcinoma can make an impact on clinical management. Those of greatest current interest are CA 125, tissue polypeptide specific antigen, tumour-associated trypsin inhibitor and the mucins cancer-associated serum antigen and DVXl. Their value in screening is unproven and cannot be recommended outside clinical trials. They are all too non-specific for use in the diagnosis of ovarian carcinoma but when used with ultrasound can help in the differential diagnosis of a pelvic mass. Although elevated CA 125 concentrations prior to the third course of chemotherapy is the best indicator of poor prognosis, this information is not accurate enough to justify a change in therapy. The major role of CA 125 is in monitoring response to therapy. A precise definition of tumour response based on a 50% or 75% fall in serial CA 125 concentrations that is maintained for 1 month is as accurate as serial measurements of tumour size for determining whether a particular treatment is active. The role of regular CA 125 measurements during follow up is unclear. However, if there is suspicion of relapse and it is found that the CA 125 concentration has risen to over twice the upper limit of normal, the probability of relapse is so high as to make other investigations unnecessary. Patients with cancer of the ovary typically present with non-specific symptoms and advanced disease. Such cancer usually metastasizes as peritoneal deposits which are rarely palpable or detectable on scans. Tumour markers, therefore, have a greater potential value in the management of ovarian cancer than in virtually any other cancer type except choriocarcinoma.

Tumor markers in patients with pancreatic carcinoma

Tumor markers are putative prognostic indicators for patients with carcinoma, but have not heretofore been evaluated in patients with Stage II and III pancreatic carcinoma. Patients with Stage II (n=9) and Stage III (n=25) unresectable regional adenocarcinoma of the pancreas were treated with combined modality therapy. Treatment consisted of split course radiotherapy and simultaneous combination chemotherapy with fluorouracil infusion, streptozotocin, and cisplatin. Prior to treatment, patients free of both infection and jaundice provided blood for CA 19-9, carcinoembryonic antigen (CEA) and CA 125 assays. The overall median survival of Stage II patients was 21.1 months. Due to the small number of Stage II patients with markedly abnormal assays, it was not possible to test for a statistically significant association between pretreatment tumor assays and survival. Among patients with Stage III pancreatic carcinoma, a CA 19-9 assay of 2000 u/ML or less identified a group of 16 patients with a median survival of 12.8 months. In contrast, 8 Stage III patients with a CA 19-9 assay of greater than 2000 u/mL had a median survival of 8 months and only 1 patient survived for 1 year (P=0.020, log rank test; P=0.010, Wilcoxon test). Among Stage III patients, a comparison of those with a normal assay versus any degree of abnormal assay failed to provide prognostic information. Analyses based on a combination of CA 19-9 and CA 125 assays provided additional powerful prognostic information: (P=0.002, log rank test; P=0.005, Wilcoxon test). CEA assays failed to provide information alone or in combination with the CA 19-9 assay. After adjusting for the CA 19-9 assay in multivariate analyses, neither performance status nor tumor size were significant prognostic variables for patients with Stage III cancers. Pretreatment CA 19-9 assays provide powerful independent and objective prognostic information.

Diagnostic usefulness of five tumor markers in patients with primary lung cancer

Diagnostic usefulness of five tumor markers in patients with primary lung cancer

Pro-gastrin-releasing peptide(31-98) as a tumour marker of small-cell lung cancer: comparative evaluation with neuron-specific enolase

We attempted to clarify whether serum levels of a carboxy-terminal fragment of ProGRP, ProGRP(31-98), could serve as a more accurate tumour marker in patients with SCLC than neuron-specific enolase (NSE). ProGRP(31-98) and NSE were measured retrospectively in 101 newly diagnosed untreated patients with SCLC, 111 with non-small-cell lung cancer (NSCLC) and 114 patients with non-malignant lung diseases. ProGRP(31-98) and NSE levels were determined using a sandwich enzyme-linked immunosorbent assay. Sensitivity in SCLC patients was 72.3% for ProGRP(31-98) and 62.4% for NSE. Comparing the area under curve (AUC) of 'receiver operator characteristics' of ProGRP(31-98) with that of NSE, ProGRP(31-98) was the more powerful marker in the diagnosis of SCLC (P = 0.0001). Serum levels of ProGRP(31-98) were higher in the 40 patients with extensive disease than in the 61 patients with limited disease (P = 0.0082). ProGRP(31-98) was significantly higher in patients with pure small-cell carcinoma than in patients with mixed small-cell/large-cell carcinoma (P = 0.02). In serial measurement in 16 patients responding to treatment, a high degree of correlation was noted between the decrease in serum ProGRP(31-98) levels and clinical response during the second week after treatment (P = 0.0045). These results indicate that the determination of serum ProGRP(31-98) levels plays an important role in the diagnosis and treatment of SCLC patients.

Elevation of Cerebrospinal Fluid Soluble CD27 Levels in Patients With Meningeal Localization of Lymphoid Malignancies

Abstract Diagnosis of meningeal localization of lymphoid malignancies by means of cytologic examination of the cerebrospinal fluid (CSF) can be difficult. Thus far no reliable CSF tumor markers have been identified. CD27 is a transmembrane disulfide-linked 55-kD homodimer present on most peripheral blood T cells and on a subset of B cells. CD27 is also expressed on human malignant B cells and high levels of soluble CD27 can be present in the serum of patients with B-cell malignancies. The aim of this study is to determine prospectively the diagnostic value of CSF sCD27 as a tumor marker in patients with meningeal localization of lymphoid malignancies. CSF sCD27 levels were determined by sandwich enzyme-linked immunosorbent assay. The optimal cut-off value using receiver operator characteristics curves was found to be 10 U/mL. sCD27 levels were normal in all 50 control patients (lumbar disc protrusion) and in 39 of 40 samples obtained from patients with either solid tumors or acute myeloid leukemia. Of 104 CSF samples from 70 children with acute lymphoblastic leukemia (ALL) or non-Hodgkin's lymphoma (NHL) undergoing routine central nervous system (CNS) staging, sCD27 was false positive and false negative in only one sample each. In 70 samples from 45 patients suspected of meningeal localization of ALL or NHL, the sCD27 test had an excellent sensitivity (100%) and specificity (82%). In 7 patients with positive CSF studied longitudinally, sCD27 levels correlated very well with remission and relapse. sCD27 levels were not nonspecifically increased by the administration of cytostatic drugs. Finally, sCD27 was also elevated in the 4 patients studied with primary central nervous system lymphoma (PCNSL). CSF sCD27 is a promising tumor marker in patients with either meningeal localization of lymphoid malignancies or PCNSL, and can be useful in the differential diagnosis of CNS involvement by either lymphoid malignancies or solid tumors.

Elevation of cerebrospinal fluid soluble CD27 levels in patients with meningeal localization of lymphoid malignancies [published erratum appears in Blood 1996 Oct 1;88(7):2818

Diagnosis of meningeal localization of lymphoid malignancies by means of cytologic examination of the cerebrospinal fluid (CSF) can be difficult. Thus far no reliable CSF tumor markers have been identified. CD27 is a transmembrane disulfide-linked 55-kD homodimer present on most peripheral blood T cells and on a subset of B cells. CD27 is also expressed on human malignant B cells and high levels of soluble CD27 can be present in the serum of patients with B-cell malignancies. The aim of this study is to determine prospectively the diagnostic value of CSF sCD27 as a tumor marker in patients with meningeal localization of lymphoid malignancies. CSF sCD27 levels were determined by sandwich enzyme-linked immunosorbent assay. The optimal cut-off value using receiver operator characteristics curves was found to be 10 U/mL. sCD27 levels were normal in all 50 control patients (lumbar disc protrusion) and in 39 of 40 samples obtained from patients with either solid tumors or acute myeloid leukemia. Of 104 CSF samples from 70 children with acute lymphoblastic leukemia (ALL) or non-Hodgkin's lymphoma (NHL) undergoing routine central nervous system (CNS) staging, sCD27 was false positive and false negative in only one sample each. In 70 samples from 45 patients suspected of meningeal localization of ALL or NHL, the sCD27 test had an excellent sensitivity (100%) and specificity (82%). In 7 patients with positive CSF studied longitudinally, sCD27 levels correlated very well with remission and relapse. sCD27 levels were not nonspecifically increased by the administration of cytostatic drugs. Finally, sCD27 was also elevated in the 4 patients studied with primary central nervous system lymphoma (PCNSL). CSF sCD27 is a promising tumor marker in patients with either meningeal localization of lymphoid malignancies or PCNSL, and can be useful in the differential diagnosis of CNS involvement by either lymphoid malignancies or solid tumors.

Changing incidence and delay of testicular cancer in southern Norway (1981-1992).

The medical records of 352 patients with newly diagnosed testicular cancer were reviewed. Patients were orchiectomized during three 2-year periods (1981/82, 1986/87, 1991/92) and were referred for further treatment to the Norwegian Radium Hospital. They represented 96% of all cases with unilateral testicular cancer occurring within a defined area in the southern part of Norway. An increase in testicular cancer patients was registered, mainly between the second and third time periods (61% increase). Gynaecomastia was recorded in 7% of all patients (seminoma: 6%; non-seminoma: 8%). Serum tumour markers (alpha-fetoprotein and/or human choriogonadotropin) were elevated before orchiectomy in 51% of the evaluated patients. During the studied 12-year period, considering seminoma and non-seminoma patients combined, the overall median delay decreased from 18 to 14 weeks (p = 0.006), the overall median diagnostic delay decreased from 14 to 10 weeks (p = 0.04) and the median treatment delay decreased from 37 to 28 days (p = 0.002). Due to increased frequency of stage I patients, introduction of an outpatient-based surveillance policy and improved administrative routines of the Health Care System, the median time of hospitalization was reduced from 37 (1981/82) to 9 days (1991/92). In seminoma, but not in non-seminoma patients, an overall delay of less than 16 weeks from the onset of symptoms was correlated with the incidence of stage I disease. The cancer-related 5-year survival rate for all 352 patients was 99%, without significant difference between the three periods under investigation. A patient's delay of more than 3 months was correlated with a significantly decreased 5-year survival rate if all patients are considered (p = 0.02). (1) The significant increase of the incidence of testicular cancer in the southern part of Norway remains unexplained and warrants intensified search for aetiological factors of this malignancy. (2) The Health Care Service is challenged to make available sufficient resources for the rapid diagnosis, treatment and follow-up of the increasing number of new patients with testicular cancer, following modern principles of toxicity-reduced and resource-saving treatment. (3) Attempts should be made to shorten the patient's and doctor's delay by awareness campaigns and postgraduate education of general practitioners. The importance of the determination of serum tumour markers in patients with testicular masses should, in particular, be emphasized together with the significance of gynaecomastia in the young adult male.

頭頚部癌患者におけるＣＹＦＲＡ２１−ｌの意義

頭頚部癌患者におけるＣＹＦＲＡ２１−ｌの意義

Combination assay for serum tumor markers in patients with hepatic carcinoma

Combination assay for serum tumor markers in patients with hepatic carcinoma

1264 Cytokeratin 19 soluble fragments (CK19) determination in patients with non-small cell lung cancer (NSCLC): Comparison with TPA, CEA, SCC and NSE

Preliminary studies have shown a correlation between CK 19 high levels and NSCLC. In order to evaluate the clinical role of this tumour marker, we have compared CK 19 with TPA, CEA, SCC and NSE in a series of 72 patients with newly diagnosed, histologically proven NSCLC (39 squamous cell, 33 adenocarcinoma); all patients underwent surgical resection. CK 19 serum levels were determined by means of the Enzymun Test Cyfra-21.1 (Boehringer Mannheim). ROC curves were defined for each tumour marker; Youden test, Mann-Whitney U test and the Kruskal-Wallis test were used for statistical analysis. Our data show that CK 19 is an accurate tumour marker in patients with NSCLC and it displays a close association with the squamous cell histotype. However, CK 19 does not offer better informations than CEA in adenocarcinoma and TPA in squamous cell carcinoma. Preliminary studies have shown a correlation between CK 19 high levels and NSCLC. In order to evaluate the clinical role of this tumour marker, we have compared CK 19 with TPA, CEA, SCC and NSE in a series of 72 patients with newly diagnosed, histologically proven NSCLC (39 squamous cell, 33 adenocarcinoma); all patients underwent surgical resection. CK 19 serum levels were determined by means of the Enzymun Test Cyfra-21.1 (Boehringer Mannheim). ROC curves were defined for each tumour marker; Youden test, Mann-Whitney U test and the Kruskal-Wallis test were used for statistical analysis. Our data show that CK 19 is an accurate tumour marker in patients with NSCLC and it displays a close association with the squamous cell histotype. However, CK 19 does not offer better informations than CEA in adenocarcinoma and TPA in squamous cell carcinoma.

Quantitation of MCA, CA 19.9, CA 125 and TSA levels in lung cancer

There is an accumulating body of evidence in the medical literature concerning the investigation of many biochemical markers for lung cancer [I]. The serum tumor markers which have been used in lung cancer have not shown a dmgnoshc ability h c h makes their routine clinical use recommeaded. Of these markers, carcinoembryonic antigen, the hest known, can only offer a sensitivity of 30 % with a specificity 90 YO [2]. Preliminary published data suggest that much-like carcinomaassociated antigen (MCA) is a rather specific and sensitive tumor marker in patients with breast carcinoma [3,4], but there is not yet cxmensus as to its posslble use in sera and in cytosol. Its utilw in lung cancer is still discussed 15). Carbohydrate antigen 19.9 (CA 19.9) is a carbohydrate antigenic determinant which has been identified as a sialylated lacto-Nfucopentaose 11, an oligosaccharide shanng structural f i r e s with L m i s blood group substances. It has been foynd on glycoproteins in the sera of cancer patients, more specifically in those with colorectal, gastric and pancreatic cancer [6,7]. Cancer a n t i p 125 (CA 125) is a high molecular mass glycoprotein expressed on the cell surface of some deriMtes of embryonic coelomic epithelium. This tumorassociated antigen widely used to monitor ovarian carcinomas has been suggested as a promising noninvasive test that could differentiate benign from malignant d t i o n s [8,9]. N-a&tybraminic acid or sialic acid is reMvely new marker which is under intensive evaluati00. Recent clinical andlaboratoryshuiieshaveshownthatavarietyoffancersare associated with elevated levels of sialic acid in serum and other biological fluids. Furthermore, it has been suggested that the measurement of sialic acid in the serum provides an additional test for dqnous and management of cancer [lo]. This investigation was effectuated to study and evaluate MCA, CA 19.9, CA 125 and TSA in different lung carcinoma cases and to m p a r e with normal cases. The experimental procedures were carried out in the Department of Biochemistry at Dokuz Eyllll Uniwmty Faculty of Medicine between September 1993 and August 1994. Th~s study was effectuated at two levels: tissue ( I ) and serum (2). (1) Study at the tissue level: Thirty-four tissues with histologically proven wn-snall cell lung cancer (NSCLC, n=27) and small cell lung cancer (SCLC, n=7) in comparison to normal lung cases(n=6) were investigated to quautitate the levels of MCA (much-like carcinomaassociated antigen), CA 19.9, and CA 125. Tissue samples were obtained from the patients undergoing fiberoptic bronchoscopy. MCA, CA 19.9, and CA 125 tissue levels were assayed in the supernatants of tissue homogenates by employing an enzyme-immunoassay (EIA). Using Kruskal Wallis and MaMBtwhrtney test, statistidy significant difference was found belween MCA levels in the group of overall lung cancer (0.49 f 1.0 Uhng protein) and the group of normal tissues (4.56 f 3.0 U/ mg protein), p< 0.01. No statistically significant hfference was detemuned . betweenthetotalmalignantandnormaltissue levels of CA 19.9 and CA 125. Also, significantly higher levels of MCA were obtained in the groups of NSCLC (0.6 f 1.1 Uhng protein) and SCLC (0.06 f 0.03 Uhng protein) than in n o d tissue samples @< 0.01), whereas CA 19.9 and CA 125 levels were observed not to be statistica!ly significant. In canclusion, MCA appears to be an additional tissue marker in the diagnosis of lung cancer. Table 1 represents tissue MCA, CA 19.9 and CA 125 levels. Tablel. Tissue MCA. CA 19.9 and CA 125 levels in lung carcinoma.

1271 The role of soluble interleukin-2 (SIL-2R) receptors as a tumor marker in patients with advanced colorectal cancer (ACC)

The purpose of this study was to evaluate sIL-2R as a tumor marker in pts with ACC and to correlate it with CEA and CA 19–9. <h3>Results</h3> The sera of 52 pts and of 25 normals were studied. 75% of the pts had increased sIL-2R levels (L) mean 1539±155U/ml (CEA 692±259ng/ml, CA 19–9 10.948±5222 U/ml). Relationship of sIL-2R L with type of metastases was not significant (P<0.34). sIL-2R L had linear correlation with CEA (<i>P</i><0.05). Paired t-test between 1st and 2nd measurement of sIL-2R in non-responded pts was significant (<i>P</i> O.032). Prognostic value between pts with PROG or SD disease were highly significant (<i>P</i> 0.0008). Also, prognostic value for survival was highly significant (<i>P</i> 0.049). In conclusion, serum sIL-2R L in pts with ACC: (a) are increased in 2/3 of the pts, (b) are an indicator for disease progression, (c) they correlate with CEA L and (d) have prognostic value for survival.

Differential suppression of serum prostatic acid phosphatase and prostate‐specific antigen by 5‐alpha‐reductase inhibitor

To evaluate the effect of finasteride (Proscar) on the serum levels of prostatic acid phosphatase (PAP) and prostate-specific antigen (PSA) in patients with benign prostatic hyperplasia (BPH). Thirty patients on finasteride therapy for BPH formed the study group. Serum PSA and PAP levels were monitored for 2 years while the patients were receiving finasteride. During 12 months of finasteride therapy the serum PSA was suppressed but serum PAP was unaffected. The baseline mean PAP value was 1.303 ng/mL prior to finasteride therapy; this changed to 1.510 ng/mL (P = 0.195) at 6 months and 1.166 ng/mL (P = 0.383) at 12 months. The serum PSA was 2.630 ng/mL at baseline, 1.757 ng/mL (P < 0.001) at 6 months and 1.545 ng/mL (P = 0.001) at 12 months. Further studies are warranted to determine if PAP has a role as a tumour marker in patients whose PSA is suppressed as a result of finasteride therapy.

Clinical evaluation of a monoclonal antibody to serum KM01 for the diagnosis of hepatocellular carcinoma

In order to evaluate a monoclonal antibody KM01 which was developed in mice immunized against a human colon carcinoma cell line, serum levels of KM01 and other tumor markers were studied in patients with both hepatocellular carcinoma and liver cirrhosis and in patients with liver cirrhosis alone. The KM01 levels in the sera of 50 patients with hepatocellular carcinoma plus liver cirrhosis and 50 patients with liver cirrhosis were measured using an enzyme immunoassay method and compared with various tumor markers including α-fetoprotein (AFP), DUPAN-2,and protein induced vitamin K absence or antagonist-II (PIVKA-II). The mean serum level (±MSD.) and sensitivity of KM01 in patients with hepatocellular carcinoma plus liver cirrhosis were 734 (± 716) units/ml and 64%, respectively, they were significantly higher than those of liver cirrhosis patients ( P < 0.001). Three out of 9 cases showing negative serum AFP levels had positive serum KM01 levels. Although the sensitivity of serum KM01 level for hepatocellular carcinoma was inferior to serum AFP and plasma PIVKA-II values, the sensitivity of a combination assay of serum KM01 or AFP was increased to 88%. Clinical data of the patients with markedly elevated serum KM01 levels (more than 1000 units/ml) were compared with patients with moderately elevated levels (530–1000 units/ml); serum bilirubin and alkaline-phosphatase were statistically higher in the former group ( P < 0.01). These data suggest that the serum KM01 assay is a useful tumor marker for patients with hepatocellular carcinoma, especially in AFP negative hepatocellular carcinoma patients, and that biliary tract involvement may contribute to increased serum KM01 value.