Multiple biochemical tumor markers in seminoma. A double-blind study

The value of certain biochemical tumor markers have been well established in nonseminomatous testicular cancer. However, the lack of frequent tumor markers in the sera of patients with seminoma has prompted us to embark on this double blind study. The authors studied 89 patients with the histologic diagnosis of seminoma utilizing placental alkaline phosphatase (PLAP), gamma-glutamyl transpeptidase (gamma GT), human chorionic gonadotropin (hCG), and alpha-fetoprotein (AFP). It was found that 12/30 patients (40%) with active tumor had elevated serum PLAP and 10/30 (33%) of these patients had elevated serum levels of GGT. Eighty percent of the patients with clinically active tumors had detectable serum levels of one or more of these biochemical markers. Since the frequency of the previous tumor markers have been scarce in seminoma, these serial utilization of these biochemical markers should assist the clinician to detect and monitor seminoma patients more efficaciously. However, the false-positive, false-negative, rates, and biologic half lifes of these markers should be taken in account.     

Tumor markers in mammary carcinoma. An evaluation of carcinoembryonic antigen, placental alkaline phosphatase, pseudouridine and CA-50

In 104 patients with breast cancer, carcinoembryonic antigen (CEA), placental alkaline phosphatase (PLAP) and the carbohydrate antigen CA-50 were analysed in serum. Excretion of the modified nucleoside, pseudouridine, was analysed in urine. The patients were subdivided in three different clinical stages according to disease manifestations. Levels of CEA and pseudouridine correlated to clinical stage and 58 per cent of the patients with distant metastases had elevated levels of CEA, compared with 36 per cent for pseudouridine. For PLAP and CA-50, the levels did not show any clear correlation to clinical stage. Increased activity of PLAP correlated strongly to tobacco smoking. A decrease in the level of CEA was observed following radical mastectomy. Increase in CEA levels predicted relapse in 5 out of 14 patients within about 3 to 6 months. In patients with tumor manifestations, elevated CEA levels predicted an inferior prognosis compared to those with ordinary levels.     

Placental alkaline phosphatase as a tumour marker in seminoma using the H17 E2 monoclonal antibody assay

Serum samples from 62 patients with seminoma were assayed for placental alkaline phosphatase-like activity using the monoclonal antibody H17 E2, in order to evaluate its utility as a serum tumour marker. Fifteen of 16 patients (94%) with active seminoma had elevated serum PLAP levels. Sixteen of 46 (35%) of patients considered to be in remission had elevated PLAP levels (false positive rate 35%). Fifteen false positive results were considered attributable to concomitant smoking, and if these patients are excluded, only one false positive case was detected. In 7 out of 7 patients sequential PLAP assays reflected clinical response to treatment.     

Tumor Markers in Colorectal Carcinoma An Evaluation of Carcinoembryonic Antigen, Tissue Polypeptide Antigen, Placental Alkaline Phosphatase and Pseudouridine

The biologic markers carcinoembryonic antigen (CEA), tissue polypeptide antigen (TPA), placental alkaline phosphatase (PLAP) and pseudouridine were analysed in 37 patients with colorectal carcinoma. CEA, TPA and PLAP were derived from the serum and pseudouridine from the urine. The incidence of all four markers increased with advancing stages of the disease. Patients with distant metastases had elevated levels of CEA, TPA, PLAP and pseudouridine in 85, 27, 18 and 33 per cent of the total cases, respectively. When survival was compared, patients with 2 to 4 elevated markers had shorter survival than those with none or only one elevated marker.     

Serum marker potential of placental alkaline phosphatase-like activity in testicular germ cell tumours evaluated by H17E2 monoclonal antibody assay

A monoclonal antibody (H17E2) was used in a solid-phase localisation of enzyme activity (ILEA) assay to evaluate placental-like alkaline phosphatase (PLAP) as a serum marker of testicular germ cell tumours. Single or repeated assays were performed on 213 normal blood donor and a smaller number of term pregnancy and testicular cancer sera. The detection limit of PLAP by this system was 0.14 O.D. units equivalent to 0.04iul-1. Of 50 patients with established metastatic disease tested before treatment, 88% of 16 with seminoma, 54% of 13 with mixed seminoma and malignant teratoma and 33% of 21 with malignant teratoma had serum PLAP greater than 0.2 O.D. units. This compared to an incidence of 2% in non-smokers and of 29% in smokers who had been free of disease for more than 12 months. In 15 of 22 successfully treated patients, pre-treatment serum PLAP exceeded 0.2 O.D. units (mean 0.69 O.D.) and varying (53-97%) reductions in the initial levels occurred with treatment. These results with monoclonal antibody ILEA assay suggest that measurement of PLAP levels will be useful in the management of patients with germ cell tumours, particularly seminoma.     

Levels of alkaline phosphatase isozymes in human seminoma tissue

The three human isozymes of alkaline phosphatases were quantitatively determined in normal testis and seminoma tissues. The highly selective assays were based on isozyme specific monoclonal antibodies. In the normal testis approximately 90% of the catalytic activity originates from the tissue unspecific alkaline phosphatase, and the remaining activity was due to trace expression of both intestinal (approximately 5%) and placental alkaline phosphatase (PLAP) or PLAP-like isozyme (approximately 5%). In homogenates of seminoma tissues, highly increased levels of all three isozymes were identified. Both the tissue unspecific alkaline phosphatase and PLAP-like enzymes displayed relative increases of 10- to 100-fold and intestinal alkaline phosphatase 2- to 10-fold compared with normal testis. This finding indicates that the entire genome coding for alkaline phosphatases may be activated in seminomas. The PLAP-like enzyme from seminoma cells comprises a heterogenous population of molecules demonstrating partial heat sensitivity and microheterogeneity upon starch gel electrophoresis in contrast to the pregnancy related PLAP. These findings have implications for the different PLAP assays used in the clinical monitoring of seminoma patients.     

Monoclonal antibody assay of serum placental alkaline phosphatase in the monitoring of testicular tumours

A monoclonal antibody (H17E2) recognising both placental alkaline phosphatase (PLAP) and testicular PLAP-like alkaline phosphatase was incorporated in a solid phase immunoassay. This was used to measure levels of PLAP in 257 sera from 148 patients with germ cell neoplasms of the testis. High levels of PLAP were found in all patients with active seminomas (mean 0.85 O.D.) compared to those in clinical remission (mean 0.20 O.D.) (P less than 0.0001). More importantly, changing levels of PLAP correlated with the course of disease in 79 samples from 33 patients with seminoma (P less than 0.0001). Elevated PLAP levels were also noted in patients in remission who were smokers (mean 0.32 O.D.) compared to non-smokers (mean 0.15 O.D.) (P less than 0.001). These data demonstrate that determination of PLAP levels using this sensitive immunoassay is an important new adjunct in the monitoring of the response to treatment in patients with seminoma.     

Is placental alkaline phosphatase (PLAP) a useful marker for seminoma?

The usefulness of placental alkaline phosphatase (PLAP) as a tumour marker was assessed in 1578 serum samples from 236 patients with seminoma. Smoking habits were known for all but 7 patients (22 samples). Smoking was associated with significantly higher mean levels of PLAP in disease-free patients (28.8[S.E. 2.1]U/l vs. 15.9[1.3] U/l in non-smokers). Mean PLAP levels were higher in patients with active disease (78.6 [23.5] U/l in non-smokers and 47.2 [18.5] U/l in smokers). The median values showed a similar trend. However, there was considerable overlap between the various groups and differences between mean and median values indicated that PLAP values were distributed asymmetrically. The predictive value of PLAP as a tumour marker was consequently much less than superficial inspection of these values might suggest. In 97 patients on surveillance, only 2 out of 11 patients who relapsed had elevated PLAP at the time of clinically detectable relapse. With the upper limit of normal PLAP quoted by our laboratory (35 U/l), specificity and sensitivity were, respectively, 88% and 45% (all patients) and 96% and 47% (non-smokers). The sensitivity and specificity of PLAP were assessed in more detail for a series of threshold values (normal vs. abnormal) with a graphical method. Only in non-smokers did PLAP seem useful and even in this group the positive predictive value of an “abnormal” test may be low; less than 50% in clinically relevant circumstances. Serum PLAP assay cannot usefully stand alone as a marker for seminoma and its routine estimation contributes little to follow-up.     

Placental-like alkaline phosphatase and DNA flow cytometry in spermatocytic seminoma.

Immunohistochemical analysis was done on 7 testicular tumors classified as spermatocytic seminoma (SS) and 25 classic seminomas. Except for a few scattered cells, the spermatocytic seminomas were negative for placental-like alkaline phosphatase (PLAP); the classic seminomas were all positive for this enzyme. The SS also were negative for alpha-fetoprotein (AFP), beta-human chorionic gonadotropin (hCG), and leukocyte common antigen (LCA). The ploidy of the seven tumors of SS was as follows: two, diploid; two, near-diploid; one, tetraploid; one, aneuploid; and one, uninterpretable. The essentially negative staining of SS for PLAP was strikingly different from the pattern in classic seminoma. Thus, staining for this enzyme is useful for making the differential diagnosis between classic seminoma and SS. To differentiate between malignant lymphoma and SS, staining for leukocyte common antigen is helpful.     

Patterns of seminoma tissue markers and deletions

Seminomas and control tissues were analyzed for several tumor markers. Very high levels of placental alkaline phosphatase (PLAP)-like enzyme levels were found in all 18 seminomas studied. The majority of the seminomas were of phenotype I, thus differing from palcental PLAP. The mean amount of enzyme protein as measured by monoclonal antibodies, was 100 times higher than in non-malignant tissues and 10 times lower than in placental tissue. The specific enzymatic activity in seminomas was about half of that observed in placenta. Similarly, the specific activity of PLAP-like enzymes in sera of patients with seminoma was only about half of that found in pregnancy sera. HCG was strongly elevated in 3 seminomas, but not obviously related to PLAP. Thirteen of the 17 pure seminomas had HCG over 100 IU/g, which was not seen in normal testes. Liver alkaline phosphatase (LAP) and intestinal alkaline phosphatase (IAP) were high in seminomatous tissues, the mean increases being 60-fold and 20-fold, respectively. The highest IAP levels were found in 2 yolk-sac tumors. Ferritin was moderately elevated in seminomas, but high in several control tissues. Carcinoembryonic antigen (CEA) was not elevated and alpha-fetoprotein (AFP) was not detected at all in pure seminomas. A decrease in carbohydrate antigen 50 (CA-50) content was noted in seminomas as compared to normal testes, yolk-sac tumors and choriocarcinomas. Defects in tumor-related enzymes may account for increase of PLAP and decrease of CA-50.     

Detection in colorectal carcinoma patients of antibody cytotoxic to established cell strains derived from carcinoma of the human colon and rectum.

Sera from eight of 15 patients with colonic carcinoma exhibited demonstrable cytotoxicity against an established cell strain derived from adenocarcinoma of the ileocecum, HCT-8. Sera from 12 of 16 patients with rectal carcinoma were cytotoxic for an established cell strain derived from an adenocarcinoma of the rectum, HRT-18. Patients with colonic carcinoma exhibited serum cytotoxicity against only the colonic target cells, whereas patients with rectal carcinoma gave significant cytotoxicity against both cell strains. This cytotoxicity was shown to be complement-dependent and appeared to be specific for colonic and/or rectal carcinoma cells. Although the cells produced carcinoembryonic antigen (CEA) in vitro, the cytotoxic antibody response in these patients did not appear to be directed against CEA. Serum cytotoxicity was not demonstrated against two other cell strains, HCT-48 and HT-29, derived from adenocarcinomas of the human colon, except for a reaction against a blood-group-related antigen. These cell strains had comparable levels of cell-associated CEA. The routine titration of cytotoxic antibody against these established cell cultures may provide meaningful information on the host's immune response to colorectal neoplasms.     

Electrophoretic heterogeneity of alkaline phosphatase isozymes in seminoma and normal testis

Electrophoretic patterns of seminoma- and normal-testis-derived alkaline phosphatase isozymes, the placental alkaline phosphatase (PLAP)-like enzyme and the tissue-nonspecific (liver) alkaline phosphatase (LAP), were studied on starch gel and isoelectric focusing (IEF). Different migration patterns of the PLAP-like enzyme were observed with respect to both seminomas and normal testes on starch gel electrophoresis. On IEF, seminomas showed different staining patterns among different tumors; however, a common main activity was focused at pIs of 4.3-4.6, corresponding to pIs of PLAP. Normal testes showed two enzyme-staining regions, at pIs of 4.1 and 5.0-5.2, which were discriminated from pIs of PLAP and the PLAP-like enzyme in seminoma. The PLAP-like enzyme in seminoma was differentiated from PLAP by digestion with neuraminidase. Neuraminidase treatment simplified the distribution patterns of the PLAP-like enzyme in normal testis, but did not alter the pattern of microheterogeneity in seminoma. Two factors other than sialylation, namely structural modification of the carbohydrate moiety and variation of hydrophobicity, were shown to contribute to the microheterogeneity of the PLAP-like enzyme in seminoma. LAP in seminoma and in normal testis also showed marked electrophoretic heterogeneity and differences in pI distributions from LAP of liver. However, the migration patterns after desialylation were very similar to each other. The findings imply that electrophoretic heterogeneity demonstrated in LAP in seminoma and in normal testis is caused by a difference in sialic acid content in the molecule, and the heterogeneity of the PLAP-like enzyme in seminoma is considerable.     

Hydrophobicity and lectin affinity of alkaline phosphatase isozymes in seminoma and normal testis

Isozymes of alkaline phosphatases (ALP) in seminoma and normal testis were separated by use of high-performance liquid chromatography and a TSK-gel phenyl-5PW column. The tissue-nonspecific (liver) ALP (LAP) was the dominating isozyme, consisting of more than 90% ALP activity. The placental ALP (PLAP)-like enzyme contributed to 4-8% of the total ALP activity. The intestinal isozyme (IAP) could not be identified. The glycosylation patterns of the isozymes were studied using concanavalin A (Con A) affinity chromatography and batch elution with competing sugar. All PLAP activity in placental extracts and LAP activity in liver extracts was bound to Con A-Sepharose. In the tumor extracts, only 50-70% of the PLAP-like enzyme and 20-50% of the LAP activity from seminomas were bound to Con A-Sepharose. A similar binding pattern of the PLAP-like enzyme and LAP was also seen in the normal testes. This variability in Con A reactivity with PLAP or the PLAP-like enzyme was also reflected in serum of seminoma patients and of pregnant women. Thus, ALP expressed in seminoma has different lectin affinity characteristics compared with the same isozyme from placenta and liver, but almost identical to ALP in the normal testes. These findings imply that the PLAP-like enzyme and LAP in the testis can be discriminated from PLAP of placenta and LAP of liver by carbohydrate lectin affinity. It also supports the concept that the increased amounts of ALP in seminomas result from the enhanced eutopic expression of enzymes normally expressed in the testis.     

Fetal proteins occurring in testicular teratomas

Alpha-fetoprotein (AFP) and carcinoembryonic antigen (CEA) were determined simultaneously in serial samples from patients with teratocarcinomas, testicular tumors of so-called embryonic origin. These tumor-associated fetal substances were seen together in four cases, AFP alone was seen in five, CEA in three and neither antigen in two. The embryonic cancer component was present in all primary tumors in which either fetal antigen was detected. AFP was clearly related to progression or regression of tumor while CEA occurred transiently. Four primary teratocarcinomas and four seminomas were established in short-term tissue culture. Only one of these particular patients had raised serum AFP levels. When, however, the original tumor-cell suspensions were assayed by radioimmunoassay or indirect immunofluorescence, AFP was detected in three out of the four teratocarcinomas. CEA was seen in a low percentage of cells from two primary teratocarcinomas. No seminoma cells contained either of the antigens. This means that perhaps most teratocarcinomas do indeed produce AFP and/or CEA, although serum assays may be negative. No relationship was found between the occurrence of the two fetal substances.     

Eutopic expression of placental-like alkaline phosphatase in testicular tumors

Very high levels of placental-like alkaline phosphatases (PLAP-like enzymes) were observed in tissues from 13 typical seminomas. Four tumors with seminoma components contained these enzymes to varying degrees, and other testicular tumors had them in smaller or non-detectable amounts. Analysis using monoclonal antibodies produced against the common placental alkaline phosphatase (PLAP) phenotypes and enzyme inhibition studies with amino acids and peptides showed the PLAP-like enzymes present in seminoma to be similar to those PLAP-like enzymes which are expressed in lower amounts in two embryonal carcinomas and in trace amounts in normal testicular tissue. These similarities suggest that the increased expression of PLAP-like enzymes in seminomas results from enhanced eutopic expression of enzymes found in normal testis.     

Alkaline phosphatase isozymes in human testicular germ cell tumors, their precancerous stage, and three related cell lines

The four known isozymes of the human alkaline phosphatase (ALP) were detected by isoelectric focusing in extracts of various types of germ cell tumors, three related cell lines, and their precancerous elements (atypical germ cells). In seminoma, placental alkaline phosphatase (PLAP) and germ cell alkaline phosphatase (PLAP-like) could be separated by isoelectric focusing following isolation by immunoaffinity. The occurrence of both isozymes in seminoma could explain partial heat sensitivity and variation in electrophoretic patterns of the seminoma isozyme frequently observed upon starch gels, in comparison to the normal placental phenotype. The four ALP isozymes are produced not only in germ cell tumors, but already in precancerous tissues. Quantitative analysis showed that the amount of the four isozymes varies in parallel in the tumors tested. Maximal expression was found in seminoma. The relation between ALP gene overexpression and gene amplification by polyploidy of chromosomes 1 and 2 in these lesions is discussed. On the other hand, the ectopic expression of intestinal alkaline phosphatase and PLAP associated with overexpression of PLAP-like in tumor cells as well as in their precancerous stage indicates gene activation by some unknown mechanisms, probably a regulatory process affecting the three tissue-specific ALP genes simultaneously.     

An assessment of combined tumour markers in patients with seminoma: placental alkaline phosphatase (PLAP), lactate dehydrogenase (LD) and beta human chorionic gonadotrophin (beta HCG).

We have assessed the tumour markers placental alkaline phosphatase (PLAP), lactate dehydrogenase (LD), and human chorionic gonadotrophin (beta HCG) using 2,000 serum samples from 286 patients with seminoma. The ROC curves show that no one marker performs adequately for the detection of disease either at initial staging or during follow-up. We used a Markov model heuristically to devise strategies, in which marker results were assessed in combination, which might be useful in clinical practice. We found that the best strategy was to consider a test result abnormal only if either the beta HCG was greater than 6 Ul-1 or the LD was greater than 400 U l-1 and the PLAP level was greater than 60 U l-1. This will detect about 50% of patients with disease and the false-positive rate is 2%. In practical terms this means that PLAP need only be estimated in patients whose beta HCG is less than 6 IU l-1 and whose LD is greater than 400 U l-1.     

Tumor-associated glycoprotein-72 serum levels complement carcinoembryonic antigen levels in monitoring patients with gastrointestinal carcinoma. A longitudinal study

Eighty-two patients diagnosed with gastrointestinal (GI) adenocarcinoma were evaluated before and for 26 months after primary tumor resection for the presence of two serum tumor markers: tumor-associated glycoprotein-72 (TAG-72) and carcinoembryonic antigen (CEA). Elevated TAG-72 and CEA serum levels were found preoperatively in 32 (39%) and 34 (41.5%) of the 82 patients, respectively. The percentage of patients with elevated serum levels of either TAG-72 or CEA was 56.1% (46 of 82). Twelve (15%) patients who had normal CEA serum levels had elevated TAG-72 serum levels, and conversely, serum from 14 (17%) patients who were TAG-72 negative were CEA positive. Forty-five of the 82 patients were diagnosed with advanced disease (i.e., Stages C and D for colorectal, Stages III and IV for stomach), and 29 (64.4%) and 26 (57.8%) of those patients had elevated serum levels of TAG-72 or CEA, respectively. Elevated levels of either TAG-72 or CEA, however, were found in sera of 82.2% of patients with advanced GI cancer, which is an increase of 24.4% over the use of CEA antigen alone as a marker of disease. The measurement of both TAG-72 and CEA may improve the diagnosis of patients with GI malignant disease due to the apparent complementary association which exists between these tumor markers. Serum TAG-72 and CEA levels were monitored in 31 patients for varying lengths of time after resection of the carcinoma; 11 patients developed recurrent disease. Sera from nine of 11 (81.8%) of these patients had elevated TAG-72 levels and six of 11 (54.5%) had elevated CEA levels. Tumor marker elevations were observed either before (35 to 166 days) or at the time of diagnosis of recurrence. The elevation of one or both markers correlated with the clinical status in ten of 11 (90.9%) patients with recurrence. In addition, 20 patients who were clinically free of disease after more than 700 days' follow-up had normal serum levels of both TAG-72 and CEA. These findings suggest that the combined use of serum TAG-72 and CEA measurements may improve detection of recurrence in patients with GI cancer and may be useful in the postsurgical management of GI adenocarcinoma patients.     

Multiple antigens as marker substances in germinal tumors of the testis.

Germinal cell tumors of the testis were studied for the presence of several tumor-associated antigens. Antisera were produced by immunizing rabbits with the purified antigens of alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), and hepatoma ferritin. Indirect immunofluorescence on embryonal carcinoma with or without teratoma components demonstrated that their staining range was 1--60 per cent with antiserum against AFP, 0--16 per cent with anti-serum against ferritin, and 0-40% with antiserum against CEA. Ferritin-like substances have not been described previously in germinal tumors of the testis. No staining was seen with seminoma cells or benign testicular tissues. Raised serum levels of AFP and the ferritin-like substance were related both to the presence of tumor and to dissemination of the disease. CEA occurred transiently in serum. Eleven patients with primary tumors had no antigen in their sera and have all survived, but the median survival time for 8 patients with either antigen in preoperative sera was 12 months. Five patients with advanced tumor in whom neither AFP nor ferritin was detected had a much longer median survival time (58 mo) than did 13 patients with high levels of serum AFP or ferritin (12 mo). The presence of either AFP or ferritin in sera of patients with primary or advanced disease, therefore, seemed to indicate a poor prognosis. The determination of both substances in serum may be useful in the follow-up of patients with certain types of testicular tumors. The proportion of cells containing each antigen varied in the different tumors. Similarly, each antigen could occur independently in serum. This suggested that certain germ cell tumors contained subpopulations of cells, which differed in their production and release of the antigens studied.     

Pathobioogy of intracranial germ-cell tumors: Immunochemical,immunohistochemical, and electron microscopic investigations

Summary In an attempt to clarify the pathobiology of intracranial germ-cell tumors, we studied 56 patients with these tumors, both clinically and pathologically. Our evaluations included immunochemical, immunohistochemical, and electron microscopic observations. Thirty tumors originated in the pineal region, ten in the suprasellar region, nine in the basal ganglia, and two in other regions. Five had double sites of origin. Histologically, there were 39 germinomas, six teratomas, four embryonal carcinomas, two choriocarcinomas and five mixed tumors. Among tumors tested immunochemically, serum carcinoembryonic antigen (CEA) was positive in 3.4%, serum alpha-fetoprotein (AFP) in 25.8%, and serum human chorionic gonadotropin (HCG) in 32.1%. In CSF, CEA was positive in 14.3%, AFP in 26.3%, and HCG in 75%. One third of germinomas were positive for serum HCG, and 88.9% for CSF HCG. One half of teratomas were positive for CEA of CSF. In immunohistochemical testing, the positive rates for CEA, AFP, HCG, and placental alkaline phosphatase (PLAP) were 26.9%, 11.5%, 24.1%, and 55.6%, respectively. Most teratomas were positive for CEA and the reactions were prominant in gland-like structures. HCG-positive syncytial cells were found in three germinomas and two embryonal carcinomas as well as in choriocarcinomas. The majority of germinomas were positive for PLAP and the reactive sites were tumor cell membranes and cytoplasm. On electron microscopy, germinomas were least differentiated, followed in order by embryonal carcinomas, choriocarcinomas, and teratomas. Further clinical and pathological studies will be necessary for a better understanding of the biology of these tumors.