Search

1. The concentration of total protein in c.s.f. and plasma has been measured in fetal sheep of different gestational ages and in the adult. In c.s.f. it was highest (approximately 840 mg/100 ml.) in the youngest fetuses (35 days) and declined steeply by 60 days (260 mg/100 ml.). It decreased less markedly in the last half of gestation to reach about 50 mg/100 ml. at 125 days which is twice the adult value. Protein concentration in plasma was lowest in the youngest fetuses and did not rise much until the second half of gestation during which time it doubled. There was a further rather larger increase between the late fetal (125 days) stage and the adult.2. Albumin, fetuin, alpha-fetoprotein (AFP), transferrin and lipoprotein were identified in c.s.f. and plasma.3. The concentrations of albumin, AFP and fetuin in c.s.f. and plasma at different gestational ages were measured using immunodiffusion and radioimmuno-assays.4. Albumin was the major protein in plasma at all ages studied (35-128 days gestation and adult). Its concentration increased throughout gestation whereas that of fetuin was similar at all fetal ages and that of AFP declined markedly during the second half of gestation. In the adult, fetuin was only about 0.1% of the total protein in plasma and AFP was not detectable.5. In 35-40 day fetuses albumin, AFP and fetuin accounted for 90% of the total protein concentration in plasma and for about 70-80% of the total protein concentration of c.s.f. As gestation progressed the concentration of all three proteins in c.s.f. declined. But the concentration of AFP and fetuin fell more rapidly and to a greater extent than that of albumin; neither AFP nor fetuin could be detected in adult c.s.f.6. The c.s.f.: Plasma ratio was above 50% for AFP and above 60% for fetuin at 35 days compared with about 25% for albumin at the same fetal age. The c.s.f.: plasma ratios for all three proteins declined with increasing fetal age and were not significantly different from each other by about mid gestation.

Second-trimester diagnosis of most open neural tube defects has become clinically feasible in the past five years. Alpha fetoprotein is the common denominator in a two-stage screening program. Maternal serum analysis identifies high-risk pregnancies; amniotic fluid analysis is diagnostic. The biochemical basis for these tests and findings of clinical studies are reviewed.

Proteins, which are elevated in their blood concentration in pregnant women and patients suffering from malignant tumours, are reported because of their growing significance for the clinical practice. At present mainly are the following "pregnancy" proteins of clinical relevance: human chorionic gonadotrophin (HCG), human placental lactogen (HPL), placental heatstable alkaline phosphatase (HAP), pegnancy-associated alpha2-glycoprotein ("pregnancy zone" protein, PZ), socalled pregnancy-specific beta1-glycoprotein (SP1) and alpha1-fetoprotein (AFP). Applications to the clinical practice may be the surveillance of normal pregnancy, the recognition of dangerous conditions for mother and fetus, the inhibition of graft rejection, the induction of abortion, antibodies against pregnancy proteins as abortifacient and antifertilizer as well as the tumour diagnosis including the control of treatment and recognition of recidives.

Open neural tube defects have been associated with elevated levels of alpha-fetoprotein in maternal serum and amniotic fluid. In the present study, specimens were collected during 14 to 20 weeks of gestation in five laboratories. Dates were confirmed by ultrasound. Among 1,049 subjects included in total serum alpha-fetoprotein determinations, 25 infants were born with neural tube defects, one patient had intrauterine fetal demise, and 1,023 delivered normal infants. The false-negative rate for serum alpha-fetoprotein was 0.38%, and the false-positive rate was 0.29%. For the total 1,211 amniotic fluid specimens, 23 infants were born with neural tube defects. The false-negative rate for amniotic fluid alpha-fetoprotein was 0.25%. All were closed lesions. There were no false-positive values. This study reaffirms the value of screening for neural tube defects by means of serum alpha-fetoprotein RIA. It also negates the argument that serum screening would promote excessive amniocentesis.

We measured tumor cytosol and peripheral serum levels of human chorionic gonadotropin (HCG) and alphafetoprotein (AFP) simultaneously in 40 patients with nonseminomatous testicular cancer. Serum AFP levels were normal in 13 of 40 patients and were elevated in 7 of 13 (54%) tumor cytosols from these patients. Serum HCG levels were normal in 16 of 40 patients and were elevated in 13 of 16 (81%) tumor cytosols from these patients. Correlation of simultaneous serum and tumor cytosol levels of HCG and AFP indicates that tumor cytosols are more likely to have elevated levels of HCG and/or AFP than are sera in patients with testicular tumors.

The most important aspects of the clinical findings, endocrinology and course of the disease in 14 patients with the histological diagnosis of pure seminoma of the testis and positive serum beta-HCG levels are presented. In addition to routine investigations, immunohistochemical examination of the formalin-fixed specimen by the immunoperoxidase technique was also carried out. Serum alpha-feto-protein levels remained normal in all 14 patients. There was no correlation between serum beta-HCG level and tumour stage. The longest period of observation was 76 months and the shortest 8 months. Patients with no evidence of metastatic disease (5 patients, N0M0) or patients with low-stage disease (5 patients, N1 or N2M0) respond very well to radiotherapy. 4 patients (N3-N4M0) with large retroperitoneal metastases or distant lymph node metastases underwent different combinations of treatment.

Morphometric, immunohistochemical, and electron-microscopic studies were undertaken in an attempt to identify the types of hepatoblastoma cellular elements responsible for the synthesis of alpha-fetoprotein (AFP), and to see how they may relate to serum AFP levels and the metastatic spread and prognosis of the hepatoblastoma. Morphometric studies of 21 hepatoblastomas with statistical treatment of the results revealed a moderately strong reliable correlation of the AFP serum titer with the volume ratio of embryonal tumor cells and with that of those tumor elements of endodermal hepatic diverticulum which are similar to the latter cells with regard to degree of differentiation. Also, a consistent, reliable negative correlation was demonstrated between serum AFP titer and the volume of fetal hepatoblastoma cells. The volume ratio of stromal elements was found to be subject to chance variations and not to correlate with serum AFP level. Immunohistochemical and electron-microscopic studies confirmed the morphometric findings and showed AFP synthesis to be effected by poorly differentiated hepatoblastoma cells--by endodermal hepatic diverticulum elements at first and by embryonal and intermediate tumor cells later--and to decrease as the liver tumor cells differentiate further. It is concluded that a high serum AFP level is, generally, an indication that the hepatoblastoma is an extensive one and consists of poorly differentiated cells so that the prognosis is unfavorable.

With maternal serum alpha-fetoprotein testing, large numbers of previously "low-risk" patients are now considered high risk and are offered genetic testing. Anecdotally, these patients have been perceived as more highly anxious than other second-trimester patients referred for genetic testing because of advanced maternal age. Thus we have studied patient demographics, true genetic risks, the perceptions of risk, and state (situational) and trait (constitutional) anxiety for these patients and their partners. Significantly increased state anxiety was noted for mothers as compared with fathers both in the group of women referred for testing because of maternal serum alpha-fetoprotein levels and in those referred due to advanced maternal age. State anxiety was increased in the women referred for maternal serum alpha-fetoprotein levels as compared with women referred for advanced maternal age. True genetic risks were comparable between the groups. Some critics have argued that maternal serum alpha-fetoprotein screening engenders unnecessary anxiety. Our data show that patients undergoing genetic testing due to maternal serum alpha-fetoprotein levels have higher state anxiety than women undergoing testing because of advanced maternal age, but that indication is much less a factor than are partner differences. Therefore, increased anxiety after abnormal maternal serum alpha-fetoprotein testing results cannot be reasonably used as an argument against such testing.

Carcinoma of the testicle is a unique clinicopathologic entity in that an optimistic attitude toward cure can be entertained even in the presence of distant metastasis. Chemotherapy, followed by an aggressive surgical resection of residual disease, can result in eradication of this neoplasm. This tumor is capable of being monitored by the use of serum markers, namely, human chorionic gonadotropin and alpha-fetoprotein. After induction chemotherapy, the transition of elevated serum tumor markers to normal levels suggests that malignant disease has been eliminated or converted to teratoma. Elevated markers indicate persistent or recurrent carcinoma and mandate further chemotherapy. If normalization of tumor markers occurs, any residual mass in the abdomen, chest, or neck should be surgically resected. The otolaryngologist plays a role in the diagnosis and the resection of residual neck disease. Metastatic testicular carcinoma can present as a supraclavicular neck mass and must be considered in the differential diagnosis of a mass in this area. Large residual neck lesions are best removed through the exposure afforded by modified neck dissection.

Maternal serum alpha fetoprotein (MSAFP) measurement between 16 and 21 weeks gestation is used to define a group of women with an increased risk of fetal abnormality, particularly open neural tube defect. The test is strongly gestation dependent and women with high MSAFP levels require sonar scan examination to define gestation, exclude twins and examine the fetus for obvious malformation or death. It has been reported that women with no primary cause for raised MSAFP have an increased incidence of low birth weight babies. Conflicting reports have separately ascribed these to premature delivery and to intra-uterine growth retardation. We have studied the relationship between MSAFP and low birth weight infants with respect to both prematurity and retarded fetal growth. MSAFP values were expressed as multiples of the appropriate weekly median (MOM) values relating to normal pregnancies with normal outcomes at term. For our normal population an MSAFP value of 2 MOM is the 95% centile, i.e. 5% of normal outcome pregnancies of sure gestation will have MSAFP values in the second trimester which are at or above 2 MOM. Information was available on 389 women whose infants were liveborn singletons weighing 2.5 kg or less. 33 (8.5%) of these women had MSAFP greater than 2 MOM (p less than 0.005) and of the 145 women whose babies weighed less than 2 kg, 17 (11.7%) had MSAFP at this level (p less than 0.001) Tab. I).(ABSTRACT TRUNCATED AT 250 WORDS)

The use of amniotic fluid alpha-fetoprotein measurements in the early prenatal diagnosis of central nervous system disorders is reviewed. The method is sensitive and specific and has had a pronounced effect on the counselling of mothers who have already had a child with spina bifida or anencephaly. Measurement of other amniotic fluid components, such as fetal macrophages, beta-trace protein, alpha2-macroglobulin or fibrinogen-degradation products, is less reliable and these should not be used in making primary diagnoses. The assay of alpha-fetoprotein in maternal serum shows signs of being a useful preliminary screening process, and may identify pregnancies where amniocentesis and confirmatory amniotic fluid determinations should be offered.

In the congenital nephrotic syndrome (Finnish type), concentrations of alpha fetoprotein in the amniotic fluid and maternal serum are markedly elevated in the second trimester of pregancncy. Demonstration of this alteration allows early prenatal diagnosis of this fatal condition and elective termination of the pregnancy before 20 week's gestation.

125I-labelled asialofetuin injected intravenously into chars (Salmo alpinus L.) was cleared from the blood with a half-time of approx. 60 min. The injected asialofetuin was found to be accumulating in the liver when different organs were dissected out at various time points after the injection. 125I-labelled fetuin was not concentrated in the liver, and the uptake of 125I-labelled asialofetuin was almost completely inhibited by simultaneously injecting an excess unlabelled asialofetuin, indicating a specific uptake mechanism for the galactose-terminated glycoprotein in char liver. Denatured human serum albumin was not taken up by the liver, but was recovered in the kidneys, suggesting that the liver in char is devoid of macrophages. It therefore seems reasonable to assume that the injected 125I-labelled asialofetuin is taken up by the parenchymal cells of the char liver.

We report a prospective study aimed at assessing the value of serum marker determinations in a supposedly healthy population to detect cancer and to identify individuals at high risk. We analyzed a group of 1,611 supposedly healthy subjects attending a cancer detection center, over a 1-5 year period and a control group of 100 cancer patients. Repeated determinations of the following markers were made: CEA, AFP, HCG, beta-HCG, beta 2-M, ferritin, beta 1-SP, all by radioimmunoassay. In the literature, marker determinations are considered not to be useful for cancer screening; in spite of this, we determined "normal" and "suspicious" levels for each marker and were able to define a group "at risk" that may harbor an early cancer (representing 23.6% of the total) and a "normal" group. The cancer detection rate was 45 0/00 (17/378) in the risk group and 3.2 0/00 in the "normal" one (4/1233). Our data show that markers could play a role in cancer screening.

In rats bearing a solid form of AH-109A hepatoma, serum asialofetuin sialyltransferase activity was significantly increased. In order to identify the source of the increased serum sialyltransferase, the asialofetuin sialyltransferase activities of normal and host liver, tumor, and normal and host serum were studied by phosphocellulose column chromatography. While normal and host (day 17) livers exhibited two peaks, namely, transferases I and II, which were previously shown to be the sialylated and unsialylated species, respectively, of beta-galactoside alpha 2 leads to 6 sialyltransferase, the tumor exhibited a single peak of the enzyme, which was a sialylated species (as was transferase I) but was eluted at a position clearly distinguishable from that of either transferase I or transferase II. Under these conditions, both normal and host (day 17) sera were found to contain transferase I but not the tumor-type enzyme. The results have been interpreted as indicating that in rats with AH-109A, the tumor is not the source of the increased serum sialyltransferase. In these rats as well as in normal rats, serum sialyltransferase appears to originate mainly from the liver, whose sialyltransferase activity was also increased in rats with AH-109A.

Serum samples from 20 patients with acute exacerbation of chronic hepatitis due to hepatitis B virus and 20 patients with hepatocellular carcinoma arising from B viral cirrhosis with elevated levels of alpha-fetoprotein (AFP) were analyzed by affinity column chromatography for concanavalin A binding. Serum AFP was tested at regular intervals in all of these patients. Acute exacerbation was defined as elevation of serum transaminase greater than 300 IU/L in patients with chronic hepatitis B. In hepatocellular carcinoma, serum AFP levels fluctuated but remained higher than 92 ng/ml, whereas, in acute exacerbation of chronic hepatitis B, serum AFP levels returned to normal within 3-12 months of follow-up. The results of concanavalin A-binding assay revealed that AFP from both these groups had a high affinity for concanavalin A, and this assay could not be used to discriminate between the two conditions.

The value of ultrasonography and serum alpha foetoprotein assay in the detection of hepatocellular carcinoma was evaluated in a retrospective study of 217 patients, (135 men and 82 women; mean age 58 years) who had been hospitalized for 18 months. Cirrhosis was of alcoholic origin in 76 per cent of the cases and complicated by ascites in 54 per cent. Twenty-four patients (11 per cent) had hepatocellular carcinoma diagnosed at ultrasonography in 20 cases. The carcinoma was diffuse in 7 cases, focal in 13 cases and less than 3 cm in diameter in 3 cases. A second ultrasonography diagnosed hepatocellular carcinoma in 4 cases. In 25/217 patients (11.5 per cent) the focal lesion seen at sonography was obviously not a carcinoma. Among the 207 patients who had a serum alpha foetoprotein assay, the result was normal in 167 (81 per cent) and 4 of these had a carcinoma. Among the 40 patients with an abnormal level of alpha foetoprotein, 19 had a carcinoma. This study confirms the good sensitivity of ultrasonography in the detection of hepatocellular carcinoma, but it also shows its low specificity, especially in cases with focal lesions. Ultrasonography and serum alpha foetoprotein assay should be used to detect hepatocellular carcinomas in cirrhotic patients who may benefit from surgical treatment. In the other patients, ultrasonically-guided biopsy should provide a firm diagnosis of carcinoma before undertaking other treatments, such as chemoembolization.

Using affinity chromatography on concanavalin A Sepharose, the authors studied the molecular heterogeneity of the serum alpha-fetoprotein of 53 patients with hepatocellular carcinoma, 16 patients with metastatic tumors to the liver, and 16 patients with germ cell tumors. Mean concanavalin binding of alpha-fetoprotein in the sera of patients with hepatocellular carcinoma was 79%, whereas the mean binding in metastatic tumors was 52% and that of germ cell tumors was 45%. This striking molecular variation of the alpha-fetoprotein produced by these different tumors is helpful in the clinical distinction of these tumors.

Serum alphafetoprotein (AFP) levels were measured using a sensitive radioimmunoassay in 108 hepatitis B surface antigen (HBsAg)-positive subjects and 695 controls. The concentrations were significantly higher in the HBsAg-positives. Within this group, the highest levels were found in those with active HBV infection. In those without evidence of acute infection, the levels were higher in the high-risk than in the low-risk subjects. It is concluded: 1) that measurement of serum AFP might be a useful additional index of infectivity and prognosis in HBsAg-positive subjects; and 2) that in the light of the association between chronic HBV infection, hepatocellular carcinoma, and raised AFP in non-European populations, consideration should be given to regular monitoring of AFP levels in HBsAg-positive subjects in the United Kingdom.

Fourteen cases of yolk sac carcinoma, 10 occurring in gonadal, and four in extragonadal sites, seen at the Indiana University Hospitals from 1949 to 1974, were analyzed with respect to pathologic features, laboratory findings, and clinical course. Their histologic appearance was similar regardless of the site of origin. Two basic histologic types were observed--the more common endodermal sinus pattern and the rare polyvesicular vitelline form. The prognosis is unfavorable, but three of our cases exhibited objective responses to chemotherapy. In our small series, the better prognosis of testicular yolk sac carcinomas in children found by some authors was not evident. Four of the 6 patients with yolk sac carcinoma in which serum alpha-fetoprotein determinations were performed showed positive results. Three of these cases had residual or metastatic disease clinically. The demonstration of alpha-fetoprotein in the serum of patients with yolk sac carcinoma lends further support to the yolk sac origin of these tumors and could also prove to be of prognostic value by indicating the presence of residual or recurrent disease.

The blood levels of tumour markers (alpha-foetoprotein and human chorionic gonadotropin) were determined and serial chest X-rays were obtained from 11 patients with germ cell tumours to study the growth of the tumours. Assuming an exponential growth pattern, the tumour marker levels and diameters of lung metastases were analysed. The use of a simple model made it possible to describe the time course of the change of tumour marker levels in blood before and during treatment. The analysis provided estimates of doubling times for marker-positive subpopulations, which were generally in accordance with the doubling times for lung metastases. The estimated doubling times had a log-normal mean value of 21.3 days, in agreement with the 19.5 days for embryonic tumours obtained in other studies.

A pregnant black woman was found to have extremely high serum alpha-fetoprotein (AFP) concentrations on several occasions between 18 and 28 weeks of gestation; the amniotic fluid total AFP level was borderline at 22 weeks, and the concanavalin-A non-reactive fraction was below normal. At this stage termination of pregnancy was considered, but qualitative acetylcholinesterase electrophoresis was negative and repeated echographic examinations revealed no fetal defect. The pregnancy was allowed to continue, and a normal premature female infant was delivered at 36 weeks. Possible explanations and the potentially serious implications of the results of laboratory tests in a case such as this are discussed.

Of the 118 patients with primary liver cancer treated in our clinic during a 6 year 7 month period, 11 patients had cholangiocellular carcinoma of the liver. Ten (90.9 percent) of the 11 patients underwent surgery, and 9 (90 percent) had major hepatic resection with no operative deaths. Seven were still alive at the time of the study 2 months to 4 years 1 month after surgery. Three patients died from recurrence 9 months, 11 months, and 1 year 4 months after surgery. The 1 and 3 year actuarial survival rates after hepatic resection were excellent (85.7 and 61.2 percent, respectively) compared with the 1 and 3 year rates of 58.8 and 0 percent in the Japanese series. Even though the cholangiocellular carcinomas in our 11 patients were in the advanced stages, good results were obtained by hepatic resection. If they were detected at a much earlier time, an even more remarkable survival rate could be expected.

Pregnancy specific beta 1 glycoprotein (PS beta 1G), human choriogonadotropin (hCG), alpha-foetoprotein (AFP) and carcinoembryonic antigen (CEA) were measured simultaneously in 58 patients with a progressive germinal testicular tumour, and in 20 patients in complete remission. All markers were negative in all the cases in remission, and in 1 case of progressive teratoma. The specificity of the association of these 4 markers was 76 p. 100, higher than any of them measured alone. There was no correlation between individual markers with respect to histological classification. The highest diagnostic specificity was obtained with hCG in all types of tumours. PS beta 1G was associated with hCG but with a lower specificity. AFP is mainly related to embryonic carcinoma but may be raised in all forms of tumour, except seminomas. CEA was positive less often, irrespective of the type of tumour, and never alone. The value of measuring these tumour markers is double: on the one hand it allows better characterisation of the tumour, and on the other, the clinical course and efficacy of treatment can be followed-up more closely and tumour recurrence may be diagnosed at an early stage.

Des-gamma-carboxy prothrombin [DCP], a protein induced by vitamin K absence or antagonist-II and also abbreviated PIVKA-II, was evaluated as a serologic marker for hepatocellular carcinoma (HCC). Its plasma levels were measured by enzyme immunoassay (E-1023) using an anti-DCP monoclonal antibody in 514 patients with various diseases. Of 120 patients with HCC, 76 (63%) had abnormal DCP levels greater than 0.1 arbitrary unit (AU)/ml and 58 (48%) showed levels greater than 0.3 AU/ml. When a diagnostic minimum level of 0.3 AU/ml was applied for DCP, false-positive cases of HCC were virtually eliminated. In some patients with HCC, plasma DCP levels normalized after surgical resection of the tumor. However, they rose again later with recurrence of the disease. The sensitivity of DCP in the diagnosis and monitoring of HCC was increased by serial and simultaneous determinations of alpha-fetoprotein (AFP), because high DCP levels were observed more often in low AFP-producing HCC patients. Elevated plasma DCP levels were not related to low vitamin K concentration in the serum. In fact, in many patients vitamin K administration resulted in only a moderate reduction of DCP levels. These results suggested strongly that DCP was synthesized by the hepatoma cells.

The incidence of epi-cirrhotic liver carcinoma is apparently on the increase. Between 1969 and 1973 there were 84 cases of primary liver carcinoma and 467 of cirrhosis among 10211 autopsies. In 74 (16%) the carcinoma had developed in a cirrhotic liver. Clincally the diagnosis of carcinoma had been made in only 14 cases. The mean survival time after the diagnosis was seven weeks, after onset of the terminal symptoms 13 weeks. Typical clinical features were decompensated, hypertrophic, often coarsely granular, liver cirrhosis. The best diagnostic method was apparently laproscopy and determination of alpha 1-fetoprotein and cholestatic enzymes.

We modified a one-step, two-site, double monoclonal immunoenzymometric assay (Abbott Laboratories) for serum alpha1-fetoprotein (AFP) to increase its sensitivity and improve test precision at the low end. We increased sample size, incubation interval, and reaction time and temperature, and decreased the final reaction volume. Interassay CVs for the modified method ranged from 6.2 to 8.0% for mean concentrations of AFP in serum of 5.2 to 34.2 micrograms/L--substantially better than those for the unmodified monoclonal method--and agreed well with those of the comparison method (modified Abbott polyclonal immunoenzymometric assay). AFP values by the modified monoclonal procedure (y) correlated well with results by the polyclonal method (x): y = 0.983x + 1.84 micrograms/L (r = 0.927, n = 59). The detection limit of the modified monoclonal test was 0.2 microgram/L, as compared with 1.0 and 1.4 micrograms/L, respectively, for the modified polyclonal and the unmodified monoclonal procedures. We recommend using the modified monoclonal method for monitoring cancer patients with low tumor burden.

Right hemi-hepatectomy was performed in two patients because of metastases in one and a malignant hepatoma in the other. After five and six months, respectively, a right hepatic lobe was again demonstrated, at first by ultrasound, then by isotope scanning. Further studies by coeliacography, 99Tc hepatobida, CT scan and liver-biopsy histology from the regenerated lobe revealed a functionally and morphologically normal picture of a right hepatic lobe, the external form of the regenerated right lobe being identical with the original one. A histologically and functionally unchanged residual liver is considered the prerequisite for satisfactory regeneration after resection. Liver regeneration, as far as regulating factors are concerned, seems to be under similar hormonal control mechanism as external wound healing.

Pregnancy-associated plasma protein-A (PAPP-A) is a macromolecular glycoprotein produced during pregnancy by the trophoblast and the decidua. It is not specific to pregnancy, since immunoreactive PAPP-A is in the circulation of normal women and men and in seminal fluid. Immunohistochemically, PAPP-A is localized in some Leydig cells, in the epididymis, in the rete testis, and in seminal vesicles. This study was conducted to learn whether circulating levels of PAPP-A can be used as a marker for the monitoring of human testicular cancer. PAPP-A was increased only in certain nonseminomatous germ cell tumor [malignant teratoma undifferentiated (MTU) + seminoma, malignant teratoma intermediate (MTI)], but the overlap with the normal values was too big for PAPP-A to have any clinical significance. After orchidectomy or prostatectomy, PAPP-A decreased but the levels were were similar to those of normal men. In men, PAPP-A can be produced by certain nonseminomatous germ cell tumors, but its origin in disease-free males is probably not the testis or the prostate.

Preliminary findings of an apparently good correlation between computed tomography and serum tumor markers--human chorionic gonadotropin and alpha fetoprotein--are presented. These findings are from an ongoing prospective randomized study evaluating the efficacy of intensive chemotherapy and cytoreductive surgery in metastatic retroperitoneal testicular germ cell tumors.

The serum expression of the tumour markers hydroxybutyric dehydrogenase (HBD), beta human chorionic gonadotrophin (beta HCG), alphafetoprotein (alpha FP) and placental alkaline phosphatase (PLAP) has been assessed in a consecutive series of 50 patients with pure seminoma. In pre-orchiectomy sera from 15 patients with stage I disease, HBD was elevated in 5, beta HCG in 3, alpha FP in 1 and PLAP in 1 (of 3 only). In 27 patients with stage 2 or 3 disease, prior to chemotherapy, serum HBD was elevated in 19, beta HCG in 11, alpha FP in 2 and PLAP in 10 (of 18). The presence of beta HCG was investigated immunocytochemically in primary tumour tissue from 39 patients. It was demonstrated in 6 of 21 tumours from patients with stage I disease and 5 of 18 with stage 2 or 3 disease.

By means of polyacrylamide gel stage electrophoresis designed by us, serum gamma-glutamyl transferase (GGT) and alkaline phosphatase (ALP) were separated into 9 (I to IX) and 7 (I to VII) isoenzyme bands. The GGT II and ALP I were found only in the sera of patients with primary hepatic carcinoma showing positive rates of 29.5% and 8.5%, respectively. The combined positive rate of the two kinds of isoenzymes was 36.2% which is lower than that of alpha-fetoprotein (AFP) in the liver cancer but it has a higher specificity compared to the latter test. Furthermore, positive GGT II and ALP I may also be present in patients with liver cancer who were negative for AFP. It is obvious that the simultaneous determinations of GGT II, ALP I and AFP are mutually supplementary in the diagnosis of liver cancer.

Assays of maternal serum alpha-fetoprotein are subject to the phenomenon of assay drift, which may be defined as incorrect increase or decrease of alpha-fetoprotein values from their true values. Low maternal serum alpha-fetoprotein weekly volume (for example, fewer than 500 specimens per week) will result in a greater than 47% probability that 10% assay drift will not be recognized. Further, laboratory reports to clinicians may lead to either misdirecting 43% more pregnant women (with positive drift) into further (possibly invasive) diagnostic procedures or the offer of further diagnostic services to 32% fewer gravidas at increased risk (with negative drift) than should be so managed. We address the problem outlined above and present the reasons for establishment of regional maternal serum alpha-fetoprotein screening programs operating at sufficient volume to permit the identification and control of assay drift.

The AFP concentration in the amniotic fluid is elevated after contamination of the sample with fetal blood during a traumatic amniocentesis. Though the calculation of AFP due to fetal hemorrhage is possible by counting the fetal red cells there may be execssive errors resulting from the variability of fetal red cell count, amniotic volume, fetal serum AFP concentration, AFP concentration in the amniotic fluid changing with fetal age. Therefore it is impossible to interpretate elevated AFP values from a sample stained with fetal blood. A repeated amniocentesis is unavoidable. The correct time to do a second punction is still a matter of question.

An autopsy was performed of a patient who succumbed to pancreatic carcinoma and in whom an unusually high level of serum alpha-fetoprotein (AFP) (84,000 ng/ml) was obtained. An immunohistochemical approach, utilizing the peroxidase-antiperoxidase method, proved that AFP was confined within pancreatic carcinomatous cells of the primary lesion and liver metastases. No AFP-positive cells were ever observed in the hepatocytes adjacent to the metastatic lesions in the liver. It may be inferred that, in this particular case, the carcinomatous cells themselves synthesize AFP. The sites of production of AFP in AFP-positive pancreatic carcinoma were discussed in a review of the literature.

Between 1979 and 1982, 69 men with metastatic malignant teratoma completed sequential combination chemotherapy (POMB/ACE). Although two-thirds of these patients initially had advanced and bulky disease, life-table analysis projects a survival of 83%. Multivariate analysis indicates that the strongest predictor of survivals is not the clinical and radiological extent of disease but the initial serum concentration of human chorionic gonadotropin (HCG) and alpha-fetoprotein (AFP). In patients presenting with HCG levels below 50 000 IU/l and AFP levels below 500 kU/l the survival in 47 patients was 96%. This contrasts with the survival projected at 56% in 22 patients presenting with tumour markers at higher concentrations.