tumor markers in urologic cancer

10
TUMOR MARKERS IN UROLOGIC CANCER NASSER JAVADPOUR, M.D. From the Surgery Branch, Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Bethesda, Maryland Although the recent advances in technology, including radiologic, endoscopic, and pathologic techniques, have had significant impacts on earlier diagnosis and improved the care of cancer patients, the limits of sensitivity of these methods still preclude the detection of cancers smaller than 10’ cells (1 Gm.). During the past several years, a number of biochemical and im- munologic assays for tumor markers have made a dramatic improvement not only in sensitivity but also in specificity of detecting certain tumors. Tumor markers may be measured in sera and localized in cancer cells of cancer pa- tients utilizing sensitive and specific radioim- munoassay and immunocytochemical tech- niques. Among the important markers that have been found in patients with urologic cancers are steroid receptors, acid phosphatase, alpha- fetoprotein (AFP), h uman chorionic gonado- tropin (HCG), pregnancy specific beta, glyco- protein (SP,), calcitonin and steroid end prod- ucts which are specific for certain cancers. However, a number of nonspecific tumor markers, such as lactic dehydrogenase (LDH), carcinoembryonic antigen (CEA), placental lactogen (PL), pl acental alkaline phosphatase, and other placental proteins have been helpful occasionally. This review is a commentary on the role of biologic markers in urologic cancer; special emphasis is given to the more clinically useful and established tumor markers. This report con- tains the updated data generated in the National Cancer Institute laboratories and clinical pro- grams and other clinical and laboratory re- search centers. Criteria for Reliable Tumor Marker A number of criteria are essential for a marker to be useful in diagnosis and management of cancer patients. Among these criteria are specificity for malignant disease, detectability in body fluids and tissue extracts, a relatively short biologic half-life, and correlation of a posi- tive marker assay result with the presence of tumor. The assay result should reflect the tumor-bearing status and prognosis of the patients. It also is desirable to reflect the amount of tumor burden and to correlate with the effectiveness of anticancer therapeutic modalities. Among the available markers, only a limited number, such as AP, AFP, HCG, calci- tonin, and possibly SP,, may fulfill these criteria with an acceptable level of false positive and false negative results. Problem of false positive and false negative Clinical evaluation of the efficacy of diag- nostic tests often has produced misleading results, so that tests that were initially regarded as valuable were later rejected as worthless. The initial optimism and subsequent disillusion- ment of carcinoembryonic antigens as a marker for different cancers serve as a classic example of such disappointment. The sensitivity and specificity of a diagnostic test may be detected by a simple algebraic formula (Table I). Sensi- tivity (a/[a+b]) refers to the question: is this test sensitive enough to detect the minimal amount of disease, since failure to do so produces false TABLEI. Efficacy of tumw markers’ Test ----Cancer-- Results Present Absent Total Positive a b a+b Negative C d c+d TOTAL a+c b+d a+b+c+d (Patients) UROLOGY / AUGUST 1980 / VOLUME XVI. NUMBER 2 127

Upload: nasser-javadpour

Post on 18-Nov-2016

218 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Tumor markers in urologic cancer

TUMOR MARKERS IN UROLOGIC CANCER

NASSER JAVADPOUR, M.D.

From the Surgery Branch, Division of Cancer Treatment, National Cancer Institute, National Institutes of Health, Bethesda, Maryland

Although the recent advances in technology, including radiologic, endoscopic, and pathologic techniques, have had significant impacts on earlier diagnosis and improved the care of cancer patients, the limits of sensitivity of these methods still preclude the detection of cancers smaller than 10’ cells (1 Gm.). During the past several years, a number of biochemical and im- munologic assays for tumor markers have made a dramatic improvement not only in sensitivity but also in specificity of detecting certain tumors. Tumor markers may be measured in sera and localized in cancer cells of cancer pa- tients utilizing sensitive and specific radioim- munoassay and immunocytochemical tech- niques. Among the important markers that have been found in patients with urologic cancers are steroid receptors, acid phosphatase, alpha- fetoprotein (AFP), h uman chorionic gonado- tropin (HCG), pregnancy specific beta, glyco- protein (SP,), calcitonin and steroid end prod- ucts which are specific for certain cancers. However, a number of nonspecific tumor markers, such as lactic dehydrogenase (LDH), carcinoembryonic antigen (CEA), placental lactogen (PL), pl acental alkaline phosphatase, and other placental proteins have been helpful occasionally.

This review is a commentary on the role of biologic markers in urologic cancer; special emphasis is given to the more clinically useful and established tumor markers. This report con- tains the updated data generated in the National Cancer Institute laboratories and clinical pro- grams and other clinical and laboratory re- search centers.

Criteria for Reliable Tumor Marker

A number of criteria are essential for a marker to be useful in diagnosis and management of

cancer patients. Among these criteria are specificity for malignant disease, detectability in body fluids and tissue extracts, a relatively short biologic half-life, and correlation of a posi- tive marker assay result with the presence of tumor. The assay result should reflect the tumor-bearing status and prognosis of the patients. It also is desirable to reflect the amount of tumor burden and to correlate with the effectiveness of anticancer therapeutic modalities. Among the available markers, only a limited number, such as AP, AFP, HCG, calci- tonin, and possibly SP,, may fulfill these criteria with an acceptable level of false positive and false negative results.

Problem of false positive and false negative

Clinical evaluation of the efficacy of diag- nostic tests often has produced misleading results, so that tests that were initially regarded as valuable were later rejected as worthless. The initial optimism and subsequent disillusion- ment of carcinoembryonic antigens as a marker for different cancers serve as a classic example of such disappointment. The sensitivity and specificity of a diagnostic test may be detected by a simple algebraic formula (Table I). Sensi- tivity (a/[a+b]) refers to the question: is this test sensitive enough to detect the minimal amount of disease, since failure to do so produces false

TABLEI. Efficacy of tumw markers’

Test ----Cancer-- Results Present Absent Total

Positive a b a+b Negative C d c+d TOTAL a+c b+d a+b+c+d

(Patients)

UROLOGY / AUGUST 1980 / VOLUME XVI. NUMBER 2 127

Page 2: Tumor markers in urologic cancer

negative results. The specificity (d/[ b +d]) refers to the question: if the test is negative, what is the likelihood of the patient being free of dis- ease, and failure of the test to predict this which would produce a false positive result.“’

To have a low false negative initially one should examine a large number of patients with minimal disease. To avoid false positive results also requires a large number of patients with an inflammatory disease and a number of normal controls without disease of the organ under investigation. These comparative studies should be performed in a blind fashion to avoid any bias.

Concept of receptors

Studies from a number of laboratories have indicated that the initial step in hormone action is the binding of the hormone to a highly specialized receptor protein. It has been in- ferred that the presence of this receptor in target tissue is necessary for response of the target tissue to the hormone. Since the pioneer- ing work of Jensen, DeSombre, and Jungblut3 demonstrating specific binding of radioactive estrogen to some human breast cancer speci- mens, a number of investigators have con- firmed and amplified this finding. There is evi- dence that kidney4 and prostate5 have receptors, and several investigators have shown steroid receptors in cytoplasm and nucleus of prostatic cancer cells utilizing newer techniques such as microassays. Although in breast cancer re- sponsive to hormone therapy most of the estrogen receptors are cytoplasmic, in men most of the androgen steroid receptors are located in the nucleus. It has been assumed that the hormone responsive cells are mainly re- ceptor positive cells; therefore, measurement of receptor levels may be useful in selecting the patients with disseminated prostatic cancer and predicting the response in a given patient.5

Immunocytochemistry and Radioimmunoassay (RIA) of Tumor Markers

The diagnosis of cancer relies on morphologic features of tumor cells. Despite the apparent precision of the morphologic criteria, the diag- nosis of some neoplasms, such as testicular and ovarian tumor, is, at best, unsatisfactory based only on histologic features. The development of sensitive and specific immunocytochemical and RIA techniques makes it possible to detect minute amounts of tumor product in the serum or tissue sections of the neoplasm. Because of

the importance of these two techniques, they deserve a brief discussion.

lmmunoperoxiduse tissue staining

The 5-micron thick serial sections of formalde- hyde-fixed tumor are deparafhnized in xylene and cleared in the usual fashion. These sections are incubated in a humid chamber for twenty- four hours with rabbit anti-HCG (1: 200) or anti- AFP (1:200) in goat serum (3%). After washing, the sections are incubated for sixty minutes with goat anti-rabbit gamma globulin (1:200) conju- gated with horseradish peroxidase in 3 per cent goat serum. The sections were washed again and exposed to diaminobenzidine containing 0.05% H,O, for ten minutes. All the appropriate controls including sections exposed to normal rabbit serum also were included.6

RZA of tumor marker

RIAs are based on the ability of serum samples containing unknown amounts of antigen to displace known amounts of radiolabeled anti- gen from a specific antibody. Such RIAs have been developed for a number of markers includ- ing AFP, HCG, SP,, and acid phosphatase.

Radioimmunodetection (RID) of cancer

RID of cancer is based on utilization of a specific and sensitive tumor marker antiserum labeled with a gamma-emitting radionuclide such as Iodine-131. The deposition of radio- active antibody in tumors synthesizing a given marker such as AFP or HCG then can permit the localization of the marker-producing tumor by external scintigraphy.

Markers in Adrenal Tumor

The normal range of urinary cortisol, 17-OH corticosteroids, I7-ketosteroids, catecholamines, and 3-methoxy-4-hydroxymandelic acid (VMA) are indicated in Table II. These urinary end

TABLE II. Normal range of urinary end products of adrenal hormones utilized as tumor markers

Urinary Metabolites

Cortisol 17-OH

corticosteroids 17-Ketosteroids Catecholamines VMA

Normal Range --(Per 24 Hrs.)---

Women Men

78-365 pg. 108-409 /Lg.

2-6 mg. 3-10 mg. 6-15 mg. 9-22 mg. > 135 /Lg. > 135 /Lg.

0.7-6.8 mg. 0.7-6.8 mg.

128 UROLOGY / AUGUST 1980 / VOLUMEXVI, NUMBER2

Page 3: Tumor markers in urologic cancer

products of adrenal hormones may be utilized as tumor markers for detection and monitoring of the therapy of adrenal tumors. Although over 50 different steroids have been isolated from the adrenal cortex, only a few are secreted into the bloodstream, and the rest are intracellular intermediates. Recent studies on the bio- synthetic pathway of the adrenal steroids have shown that the main steroids synthesized by the adrenal cortex are hydrocortisone, cortico- sterone, aldosterone, and Il-hydroxyandro- stenedione that can be used as tumor markers.

The adrenocortical hormones are derivatives of cholesterol and are basically two structural types, those with a 2-carbon chain at the 17 position of the D ring and containing 21 carbon atoms, and those with a keto or hydroxyl group at position I7 and containing 19 carbon atoms. The 21-chain steroids are mineralocorticoid and glucocorticoid in function while the 19- carbon chains have androgenic activity. For practical purposes the only steroids secreted in sufficient quantities to be physiologically active are aldosterone, cortisol, corticosterone, and dehydroepiandrosterone. The major excretory end products of the androgens are the 17-ketosteroids, and for practical purposes approximately two thirds of the ketosteroids in the male urine are derived from the adrenal or as a consequence of breakdown of cortisol in the liver, and about one-third is from the testes. About 10 per cent of cortisol is converted in the liver to the 17-ketosteroid derivatives, but most is converted to cortisone and then tetrahydro- cortisone which is soluble and rapidly excreted in the urine. Therefore, the major cortisol derivatives in the urine are 20-hydroxy deriva- tives of tetrahydroglucuronides (30 per cent), tetrahydrocortisol glucuronides (25 per cent), and tetrahydrocortisone glucuronides (15 per cent). 7 Orthopara-DDD blocks the secretion of all steroids.

The adrenal medulla excretes both norepi- nephrine and epinephrine with approximately 80 per cent of the adrenal vein catecholamine output being epinephrine. Enzymes responsi- ble for the formation of epinephrine from norepinephrine are present in the adrenal medulla, and the secretion of the catecholamines is initiated by acetylcholine which is released from the neurons that embrace the secretory cell. The mechanism whereby acetylcholine acts to increase permeability of the cell and thereby allow calcium to induce secretion of the catecholamines is an area of active research. The

catecholamines in turn act for only a short time in the circulation, being reduced following oxidation to 3-methoxy-4-hydromandelic acid. These substances can be detected in the urine with about 35 per cent of the total secreted catecholamine appearing as VMA and 50 per cent secreted as free or conjugated metaneph- rines (30 pg. norepinephrine, 6 pg. epinephrine, 700 /.~g. of VMA).

For practical purposes, the epinephrine and norepinephrine stimulate effects that are similar to adrenergic discharge, although each has its specific actions. The effects are wide- spread, involving actions on the myocardium, vascular muscle, as well as multiple inter- mediary metabolic effects such as the mobiliza- tion of glycogen from the liver, induction of lipolysis, increase in metabolic rate, stimula- tion of glucagon, and inhibition of insulin.

Markers in Renal Carcinoma

Some renal cell carcinomas have been re- ported to be associated with elevated serum renin, and these patients have usually been cured of their hypertension after nephrectomy. However, it is not clear whether the primary or recurrent tumor can actually synthesize this marker. The other marker occasionally found in patients with renal cell carcinoma is erythro- poietin assumed to be responsible for erythro- cytosis.

Another group of markers found to be ele- vated in renal cell carcinoma and certain other urologic and nonurologic cancers are the polyamines.’ Putrescine, spermine, and spermi- dine are involved in ribonucleic acid (RNA) synthesis. The serum level may be elevated due to overproduction of these organic cations by any rapid growth of the cells such as embryonic development, regenerating tissues, and neo- plasms. It has been shown that urinary poly- amine excretion has increased in the urine of some cancer patients. Quantitative determina- tion of urinary polyamines appears to be useful in the diagnosis and monitoring of patients with renal cell carcinoma, according to certain in- vestigators. However, the exact role of these nonspecific markers remains to be clarified by more extensive investigation. Noronha, Rao, and Gregory4 have reported the presence of androgen receptors in renal cell carcinoma and also stress the absence or presence of insignifi- cant estrogen and progesterone receptors in these tumors, thus challenging the conventional use of progestational hormone. If this holds

UROLOGY I AUGUST 1980 / VOLUME XVI, NUMBER 2 129

Page 4: Tumor markers in urologic cancer

TABLE III. Tissue with ABO presence ABO isoantigens in normal tissue

SRCA Organs Positive Negative

Adrenal Cortex and medulla Blood Erythrocytes Leukocytes Blood vessels Endothelial cells Bronchus Epithelium Nervous system . Cell and glia Endocervix Squamous epithelium Basal layer Esophagus Squamous epithelium Fallopian tube Columnar epithelium Gallbladder Columnar epithelium Kidney Glomeruli Convoluted tubule Ureter Transitional cells Submucosa of

muscularis Urinary bladder Transitional cells Submucosa of

muscularis

TABLE IV. Correlation of grades, stage, and presence or absence of cancer surface antigens in

76 cases of bladder cancer; positivity and negativity of SRCA

SRCA-- Positive Negative

Pathologic grade I 7 3 II 15 25 III 0 26

Clinical stage A 12 16 R, 0 3 R, 0 6 C 0 0 D, or D, 0 39

_~_ _____

true, administration of anti-androgens will be of benefit in disseminated renal cancer rather than the conventional progesterone therapy in disseminated renal cancer.

Surfin et al.’ have determined plasma renin, erythropoietin, and HCG in 57 patients with renal cell carcinoma. In this series, renal eleva- tions were found in 37 per cent of patients with high-grade, high-stage lesions of mixed histo- logic cell type, and predicted a poor prognosis. Serum erythropoietin was elevated in 63 per cent of patients and was more sensitive than renin as an indicator of the presence of renal carcinoma. These investigators have concluded that serum erythropoietin was less specific and did not correlate with tumor grade, stage, histologic type, prognosis of renal cancer. None of the 57 patients had elevated levels of serum HCG. They have concluded that serum renin

and erythropoietin determinations may be help- ful as tumor markers in renal cell carcinoma. Sherwood and Goldwasser” have developed a specific and sensitive RIA for measuring erythropoietin. These authors have detected erythropoietin production by human renal carci- noma cells in culture.” Whether or not the application of this RIA will make the determina- tions of serum erythropoietin more valuable in renal cell carcinoma is not clear. Weber” has reported that in the human renal cell cancer the purine synthetic enzyme, aminotransferase, in- creases although xanthine oxidase decreases. The value of this finding in diagnosis and management of patients with renal cancer is not clear.

Markers in Bladder Cancer

The conventional histopathologic examination of the primary or metastatic tumor rarely pre- dicts the potentiality for recurrence or dis- semination. There is some evidence that cellu- lar differentiation in bladder cancer is reflected in the presence or absence of certain cell surface antigens. Included in the cell surface antigens are the A, B, and O(H) blood grouping antigens. These cell surface antigens can be detected easily by an immunologic technique known as specific red cell adherence test (SRCA).13 A number of investigators have demonstrated the reliability of SRCA in detecting the anaplastic potentiality of cancer of the cervix, stomach, pancreas, and lungs (Table III). Our laboratory also has applied this principle to bladder cancer and found a correlation between presence of this cell surface antigen and prognosis in Stage 0 A bladder cancer.14-” The technique of this test is simple and could be mastered in a uro- logic laboratory as has been discussed pre- viously. l3

It appears that there is a correlation between the absence of cell surface antigen A, B, or O(H) and invasion of bladder cancer. However, due to the weaker nature of the O(H) antigen, occasional falsely negative tests may be encountered. We recently have minimized the occasional false negative result in patients with group 0 blood by prolonging the incubation of tissue slide with the specific antisera and utilizing buffer solution with pH of 7.4 and fresh red cells.

Correlation of antigens with stage and grade

Seventy-six bladder tumors of various stages and grades were examined for the presence of the A, B, O(H) cell surface antigen employing

130 UROLOGY / AUGUST 1980 i VOLUME XVI, NUMBER 2

Page 5: Tumor markers in urologic cancer

the SRCA technique. Seventy per cent of the grade I lesions studied were positive for the cell surface antigen, and none of the 26 grade III tumors retained their antigens (Table IV). When correlated with clinical stage, the tumors that showed no antigens were those in Stages B to D, while 12 of 18 Stage A lesions were positive for the antigen (Table IV). The findings of this study show that the SRCA is a valuable technique for predicting malignant potential in low-grade, low-stage cancer of the bladder. If supported by further investigation, this tech- nique may offer the capability of selecting low- grade, low-stage bladder tumors that are destined to invade or metastasize while they are at curable stages. Several laboratories, including ours, have been interested in the cell surface antigen in urinary cytology specimens from patients with bladder cancer. la The value of SRCA on cytology is not clear at the present time. Even if it proves to be valuable it should not replace the SRCA on tissue from a bladder biopsy. In 10 consecutive patients with carci- noma in situ we have observed the loss of cell surface antigens. lg This finding confirms the natural history of bladder carcinoma in situ as has been reported by Melicow2’ and Utz et ~1.~’

The major role of SRCA on urinary cytology in our opinion is on carcinoma in situ when the cystoscopic examination of the bladder may show a normal bladder in spite of persistently positive urinary cytology. In this situation if SRCA shows loss of surface antigens from the apparently malignant transitional cells, it not only will confirm the malignant nature of the cell, but also will predict the potential invasive- ness of the flat carcinoma in situ of the bladder.

Other markers

A number of other markers, including CEA- like substance and p-glucuronidase have been reported in the urine and tissue of patients with bladder cancer. Recently several investigators have reported the presence of p-glucuronidase in normal urine and bladder epithelium of rat and human.22 This enzyme also has been found in urine and tumor tissue of patients with bladder cancer. However, this enzyme has not been used as a tumor marker because it is present also in urinary tract infection.23 Ele- vated levels of p-glucuronidase activity have been reported in the urine of bilharzial and bladder cancer patients.24 The value of this enzyme as a tumor marker is not determined as yet.

C hromosomal marker

The abnormalities of cellular chromosomes in terms of number and/or structure may be an early bladder tumor marker.25 It appears to be of prognostic value in low-grade, low-stage bladder cancer to predict the rate of recurrence and invasiveness of an apparently superficial tumor, since tumors with normal chromosomes are noninvasive. In our experience combining the SRCA and chromosome study has been reli- able in predicting the natural history of a super- ficial bladder tumor. The following case illus- trates the use of these combined modalities.

Case abstract. A fifty-six-year-old man was referred to NIH with hematuria. Physical exam- ination, excretory urogram, and urinary cytol- ogy were unremarkable. Cystoscopic ex- amination revealed a papillary tumor confined to the left lateral wall of the bladder. Resection of this mass revealed a grade II transitional cell carcinoma. Chromosome study showed several abnormal chromosome markers. SRCA showed the loss of cell surface antigen. Although the tumor appeared to be superficial and low grade, it was considered potentially invasive since it carried abnormal chromosome and had lost its antigen. A cystoscopic examination twelve weeks later revealed recurrence over the left lateral wall, and resection of this lesion revealed it to be Stage B. The patient received radiation therapy in preparation for cystectomy. Both SRCA and chromosomal abnormalities indicated tumor invasion, but this was not apparent from either endoscopy or biopsy.

Markers in Prostatic Cancer

Immunoussay of prostatic acid phosphatase

The human prostatic acid phosphatase is syn- thesized by the prostatic epithelial cells. It is a specific phosphatase isoenzyme, one of a large molecular family of phosphatases. Significant concentrations of nonprostatic acid phosphatase are normally present in red blood cells, leuko- cytes, platelets, osteoclasts, and the reticuloen- dothelial cells.

The prostatic acid phosphatase has been demonstrated to be antigenic. Therefore, specific antibody can be raised against the pro- static acid phosphatase in preparation for the specific and accurate measurement of the pro- static acid phosphatase by RIA. Such radio- immunoassay techniques have been developed by several investigators to detect the presence of prostatic acid phosphatase in the serum and

UROLOGY / AUGUST 1980 / VOLUME XVI. NUMBER 2 131

Page 6: Tumor markers in urologic cancer

bone marrow of the patients with prostatic cancer. Although there are controversial reports in reliability and specificity of such immunoas- say, reports from Cooper and Fotiz6 Choe et a1.,27 and Chu et al. 28 claim specificity and accu- racy of their immunoassay technique.2s

Recently RIA and counterimmunoelectro- phoresis (CIEP) techniques have been de- veloped through the National Prostatic Cancer Project (NPCP) of the National Cancer In- stitute. RIA was a conventional double anti- body technique utilizing acid phosphatase from the ejaculate to raise antiserum in rabbits. The second antibody was goat antirabbit immuno- globulin G serum. The CIEP method for acid phosphatase (AP) was utilization of a specific antiprostatic acid phosphatase antiserum and serum specimens from patients.’ Chu and asso- ciates28 isolated AP from prostatic cancer, and the purified enzyme was used to immunize female rabbits utilizing this antisera and serum from a patient in an electric field. A positive result indicated that the serum specimen con- tained a detectable amount of at least 20 ng./ml. of prostatic AP. 2g An intergroup study con- ducted by the NPCP has confirmed the relative specificity, sensitivity, and reproducibility of CIEP and recommended at this time to be used as the preferred method in the initial clinical staging of prostatic cancer. Comparing RIA and CIEP, the NPCP has stated that al- though RIA is highly sensitive, its proper per- formance requires highly qualified technical experts as well as sufficient funds to maintain such a costly, complex program. However, the value of CIEP as a screening method in early detection of prostatic cancer is not clear as yet. At the present time the superiority of RIA over enzymatic methods is questioned.30 Also due to relatively frequent false positive results, RIA of PAP is not practical in early detection of prostatic cancer.

As far as androgen receptors are concerned, the preliminary results from Walsh et aZs5 suggest that the duration of hormonal response in men with symptomatic Stage D prostatic cancer is related to androgen receptor content of prostatic cancer cells. If this can be con- firmed in a longer follow-up, it will be helpful in selecting patients for cytotoxic agents or endocrine therapy.

Urinary hydroxyproline excretion has been reported to be elevated in prostatic cancer with bony metastases. 31 It has been suggested that this marker should be utilized in staging

and monitoring of prostatic cancer. Another non- specific marker reported to be elevated in the serum of patients with prostatic cancer is creatinine kinase BB (CK-BB).32 The serum levels of CK-BB have been found also in prostatic fluid and decrease with effective therapy. Whether or not multiple markers such as PAP, CK-BB, and hydroxyproline will be more accurate than a single marker remains to be determined.

Testicular Tumor Markers

Over the past few years considerable progress has been made in the field of fetoplacental pro- teins as markers for cancer. Most studies have been on alpha fetoprotein (AFP), human chorionic gonadotropin (HCG), and human placental lactogen (HPL). There is also evidence that placenta secretes a chorionic thyrotropin (HCT), a chorionic follicle-stimulating hormone (HCFSH), ACTH, melanocyte stimulating hor- mone, relaxin, oxytocin, vasopressin, and renin. Thus it appears that the placenta would perform a function analogous to the hypothalamus and anterior pituitary gland, though the feedback control of this system is not clear as yet. In addition to these hormonally active peptides, the placenta also secretes a specific heat stable alkaline phosphatase, and a number of proteins designated as pregnancy associated proteins, including pregnancy specific pi glycoprotein (SP,). Because of the recent advances in AFP, HCG, and SP, and their importance in the diagnosis and monitoring of the therapeutic efficacy in testicular cancer, they deserve some comment.

Testicular cancers synthesize special protein moieties which circulate in the serum. Among the proteins which have been found in the serum of patients with germ cell tumors of the testis are AFP, HCG, and SP,.

AFP is the major protein in the human fetus, reaching a peak of 3 mg./ml. In initial studies AFP levels were detected by the agar gel precipitation test which is simple and specific with a sensitivity of about 3,000 ng./ml. With the subsequent development of specific RIA, the sensitivity was sufficient to measure 1 to 16 ng./ml. of this glycoprotein present in normal serum.33 Utilizing this assay 85 to 87 per cent of embryonal carcinomas with or without teratoma have an elevated level of AFP (> 20 ng./ml.) (Table V). None of the 130 pure seminomas had an elevated level of AFP. However, 102 of 145

132 UROLOGY i AUGUST1980 i VOLUMEXVI, NUMBER2

Page 7: Tumor markers in urologic cancer

TABLE V. Cellular origin of testicular tumor markers

Cancer AFP HCG _ Seminoma - - Seminoma with STGC* _ + Embryonal carcinoma + -

with STGC + + Yolk sac tumor +

with STGC + + Mature teratoma -

__- *Syncytiotrophoblastic giant cell.

determination of serum HCG and AFP appears to be able to stage testicular cancer more accurately than other clinical tests. Utilizing these markers, we presently have decreased the staging error in Stage I patients from an over-all 53 per cent, reported for embryonal carcinoma, to 9 to 14 per cent for Stage I and 5 to 10 per cent for Stage II (Table VII). In- terpretation of these markers requires con- sideration of their biologic half-lives (AFP, 3-5 days; HCG, 18-24 hours). Approximately 10 per cent of nonseminomatous and over 90 per

TABLE VI. Frequency of HCG and AFP in patients with testicular cancer ____

-- Frequency--- Cancer AFP (%) HCG (%) AFP and HCG (%j- -___

Seminoma o/130 (0) 10/130 (7.5) 10/130 (7.7:1 Teratoma 6/16 (37.5) 4/16 (25.0) 7116 (43.7‘) Embryonal

carcinoma 102/145 (70.3) 87/145 (60.0) 1271145 (87.5) Teratocarcinoma* 3;;p (y4;) 32/56 (57.0) 48156 (85.7‘) Choriocarcinoma 5/5 (100.0) 515 (100.0) Yolk sac tumor 314 (75.0) l/4 (25.0) 314 (75.0) - ~~ ____--

*Embryonal carcinoma with temtoma.

patients (70 per cent) with embryonal carci- noma, 36 of 56 (64 per cent) embryonal carci- noma with teratoma, 3 of 4 yolk sac tumors (75 per cent), and none of 5 choriocarcinomas had elevated levels of serum AFP in our series (Table VI). Th e cellular localization of this glycoprotein in yolk sac tumor, embryonal carci- noma (with or without teratoma) has been reported previously (Table V).

HCG is a glycoprotein secreted by syncytio- trophoblastic cell of normal placenta and is present in most patients with choriocarci- noma. Vaitukaitis, Braunstein, and ROSS~~ de- veloped a radioimmunoassay to measure HCG with antibodies to the p subunit. Using this RIA they showed that normal persons have serum values below 1 ng./ml. In our series 10 of 130 (7.7 per cent) seminomas, 87 of 145 (60 per cent) embryonal carcinomas, 32 of 56 (57 per cent) embryonal carcinoma with teratoma, and 5 of 5 (100 per cent) pure testicular choriocarci- noma had elevated levels of serum HCG.

AFP and HCG in staging and monitoring

Previously we have utilized a staging method that is useful in therapy and is not based on determinations of serum markers.35-3g However,

TABLE VII. Improved staging of NSTT with markers

Stage- Clinical Pathologic Error (%)

Markers I II I II I II

Without 72 4:;P) 52 28 to 32* 13 With 57 52 19 to 14* 5 to 10

6(N) 2(N)

KEY: P = positive; N = negative. *Based on 3 patients considered Stage III on follow-up.

cent of seminomatous testicular tumors do not have elevated serum markers utilizing the present RIA and have been designated clinically as nonmarker-producing tumors.

The important features which tumor markers add to the staging of testicular cancer are the following:

1. The clinical staging based on markers de- creases the staging error to a level of 5 to 14 per cent and may direct the physicians to appro- priate diagnosis and therapy.

2. Persistently elevated serum markers after orchiectomy for testicular cancer invariably indicate Stage II or III disease.

3. Persistently elevated serum markers after lymphadenectomy indicate Stage III disease.

UROLOGY / AUGUST 1980 / VOLUME XVI, NUMBER 2 133

Page 8: Tumor markers in urologic cancer

4. When lymphadenectomy is negative for tumor but postlymphadenectomy serum markers are persistently elevated, patients invariably have Stage III disease.

5. Perhaps the most important applications of these markers are in monitoring of testicular tumor when serially measured.

The discordance between AFP and HCG is not an unusual observation during intensive chemotherapy of these tumors. This observa- tion led to the utilization of immunocyto- chemical technique on the premise that the cellular origin of these two markers should be different. Utilizing an indirect immunoperoxi- dase technique, we were able to localize HCG in syncytiotrophoblastic giant cells of the placenta, choriocarcinoma, and AFP in em- bryonal and yolk sac tumor cells of the testes (Table V). At the present time, the most reli- able method of staging testicular cancer is by retroperitoneal lymphadenectomy since a negative marker postorchiectomy may have a 14 per cent false negative report.

It has been shown that by concentrating the HCG in urine specimens, HCG production can be monitored more sensitively than by measuring the serum HCG. Utilizing the kaolin-acetone precipitate from 24-hour urine specimens, the HCG was extracted, and the total volume was adjusted to 15 ml. Utilizing an H93 radioimmunoassay specific to the unique carboxyl-terminal peptide of HCG-P (residues No. 123-145 of HCG-P subunit) described pre- viously, the urinary HCG was measured in 32 specimens of 15 patients with initially HCG- producing testicular tumors. These 15 patients had simultaneous measurements of serum HCG utilizing an antisera against the HCG-P sub- unit (SB6) d an urinary HCG concentrates utiliz- ing H93 RIA. The initial serum HCG of these I5 patients was elevated and returned to normal levels after therapy. However, the urinary HCG was elevated in these 15 patients in spite of returning of serum HCG to normal levels after therapy.

The initial 5 patients with elevated levels of urinary HCG but normal serum HCG were found to have recurrence. The remaining 10 patients were treated with no recurrence, confirming the sensitivity of urinary HCG. The availability of this sensitive RIA in measuring urinary HCG coupled with effective chemo- therapeutic agents has made a dramatic im- provement in selecting the patients in whom further therapy is warranted.

Pregnancy specijic PI glycoprotein (SPl) and sera of testicular germ cell tumors

In the past few years, a new placental protein has been purified and utilized as a tumor marker by several investigators.40-43 Specific RIA and IP techniques have been developed to identify SP, in the sera and tumor cells of 97 men with testicular cancer. SP, was elevated at 11 to 440 ng./ml. in 3 of 6 with choriocarcinoma, 5 of 17 with embryonal carcinoma and teratoma, and 5 of 50 with embryonal carcinoma. None of 24 sera from men with seminoma and none of 5 men with orchitis had elevated SPle41 The highest value in a group of patients with non- malignant disease was 9.1 ng./ml. The new biologic marker was identified in the syncytio- trophoblastic giant cells (STGC) (Table VIII). STGC are seen occasionally in patients with embryonal carcinoma, teratoma, and seminoma. 42

It remains to be determined whether or not SP, concentrations correlate with body burden of tumor, prognosis, or therapy, and whether or not other tumor cell(s) also produce this marker.

TABLE VIII. Comparing cellular origin of SP, to those of HCG and AFP

Tumor SP, HCG AFP

Choriocarcinoma + + - Syncytiotrophoblastic

giant cells + + - Placenta + + -

Lactic dehydrogenease (LDH)

Electrophoretically in man there are five heterogenous isoenzymes called LDH. During the past several years it has become apparent that cancer cells have increased glycolysis lead- ing ta increased synthesis of lactate. Therefore, LDH may be utilized as a nonspecific tumor marker in several cancers including that of testicular germ cells. We have utilized LDH as a tumor marker in testicular tumor because of: (1) availability and simplicity of the assay when compared with RIA; (2) lack of frequency of other markers in seminoma; (3) bulky non- seminomatous testicular cancer with normaliza- tion of serum AFP and HCG while patients are on intensive chemotherapy the serum LDH may monitor the therapy; and (4) monitoring the therapy of bulky seminomatous tumor. The correlation of LDH with tumor bulk and serum levels of HCG and AFP is under investigation.

134 UROLOGY ! AUGUST1980 i VOLUMEXVI. NUMBER2

Page 9: Tumor markers in urologic cancer

Other markers

The significance of placental lactogen, placental alkaline phosphatase, and lactic de- hydrogenase are being investigated currently in our laboratory.

Gonadal stromal tumors

It also has been shown that the immuno- peroxidase technique can be used to identify steroid hormones in sections of fixed embedded tissues.40 This advance paves the way for prospective and retrospective studies of specific sites of steroid hormone localization in a poorly understood group of gonadal neoplasms, namely, those within the sex cord-stromal cate- gory. In the past, ultrastructural and histo- chemical studies were inferential in assigning the function of steroid hormone synthesis to cells containing organelles thought to be asso- ciated with steroid synthesis, or in demonstrat- ing the presence of intracytoplasmic lipid or certain enzymes, such as glucose-6-phosphate dehydrogenase and S-beta-01 dehydrogenase, which have been shown to be present in high concentrations in cells that produce steroid hormones.

The ability to localize specifically testosterone, estrogen, and progesterone has challenged many of the time-honored concepts of steroid biosynthesis. In the past, specific hormone syn- thesis was attributed more or less to specific types of cells. Theta cells were thought to be responsible for estrogen synthesis and Leydig cells for testosterone production; granulosa and Sertoli cells were regarded as inactive generally. By using highly specific antibodies for testos- terone, estradiol, and progesterone, it now has been shown that all these cells are functionally active and, furthermore, that most have the capacity to synthesize both estrogens and androgens.

Testosterone most frequently is localized in Leydig cells, but it may also be present in Sertoli cells and occasionally in granulosa cells (Table IX). Estradiol is found not only in theta cells but also frequently in granulosa, Sertoli, and Leydig cells, whereas progesterone appears to be localized mainly in luteinized theta cells and less commonly in granulosa and Leydig cells. 4o

Present and Future Perspectives

Among the most important contributions of immunology to the care of cancer patients is

TABLE IX. Correlation of cell type and steroid localization*32

Cell Type Testosterone Estradiol Progesterone

Granulosa + ++ + Sertoli + + - Theta - + Leydig ++ + -:z, _ -

*Cell marking patterns derived principally by study of corre- sponding neoplasms (parenthesis denotes exceptional finding).

perhaps the development of radioimmunoassay and immunocytochemical technique to detect the presence of a small amount of tumor marker in the sera and cancer cells of these patients. Although several urologic cancers including testicular, prostatic, and bladder cancers have relatively reliable markers, correlation of these markers with therapy and survival in controlled studies are needed. The development of placental proteins has been encouraging in testicular cancer. Although it appears that SP, is another useful additional marker for testicular cancer, more controlled study is needed to clarify further the value of this marker. Also, reliable new markers are needed for renal cell carcinoma, rhabdomyosarcoma, and other uro- logic cancers. It is the author’s belief that mul- tiple markers will be of more value in a cancer patient.

Bethesda, Maryland 20205

References

1. Ransohoff DE, and Feinstein AR: Problems of spectrum and bias in evaluating the efficacy of diagnostic tests, N. Engl. J. Med. 299: 926 (1978).

2. Feinstein AR: On the sensitivity, specificity, and dis- crimination of diagnostic tests, in Clinical Biostatistics, St. Louis, C. V. Mosby, 1977, p. 214.

3. Jensen EV, DeSombre ER, and Jungblut PW: Estrogen receptors in hormone-responsive tissues and tumors, in Wissler RW, DAo TL, and Wood S, Jr. (Eds): Endogenous Factors In- fluencing Host-Tumor Balance, Chicago, University of Chicago Press, 1967, pp. 15-30.

4. Noronha RF, Rao BR, and Gregory JG: Hormone receptors in human renal and bladder carcinomas, abstract, 74th Annual Meeting AUA, New York, 1979.

5. Walsh PC, Hicks LL, Reiner WG, and Trachtenberg J: The use of androgen receptors to predict the duration of hormonal response in prostatic cancer, abstract, 74th Annual Meeting AUA, New York, 1979.

6. Kurman RJ, et al: Cellular localization of alpha-fetoprotein and human chorionic gonadotropin in germ cell tumors of the testis using an indirect immunoperoxidase technique: a new approach to classification utilizing tumor markers, Cancer 40: 2136 (1977).

7. Javadpour N, Brennan MF, and Weltering EA: Recent advances in adrenal neoplasm, Curr. Probl. Surg. 17: 1 (1980).

8. Sanford EM, et al: Preliminary evaluation of urinary polyamines in the diagnosis of gcnitourinary tract malignancy, J. Urol. 113: 218 (1975).

UROLOGY / AUGUST 1980 / VOLUME XVI, NUMBER 2 135

Page 10: Tumor markers in urologic cancer

9. Surfin G, et a/: Hormones in renal cancer, ibid. 117: 433 (1977).

10. Sherwood JB, and Goldwasser E: A radioimmunoassay for erythropoietin, Blood 54: 885 (1979).

11. IDEM: Erythropoietin production by human renal carci- noma cells in culture, Endocrinology 99: 504 (1976).

12. Weher G: Enzymology of cancer cells, N. Engl. J. Med. 296: 486 (1977).

13. Davidsohn I: Early immunologic diagnosis and prog- nosis of carcinoma, Am. J. Clin. Pathol. 57: 715 (1972).

14. DeCenzo JM, Howard P, and Irish CE: Antigenic deletion and prognosis of patients with stage A transitional cell bladder carcinoma, J. Urol. 114: 874 (1975).

15. Bergman S, and Javadpour N: The cell surface antigen A, B or O(H) as an indicator of malignant potential in Stage A bladder carcinoma: preliminary report, ibid. 119: 49 (1978).

16. Lange PH, Limas C, and Fraley EE: Tissue blood-group antigens and prognosis in low-stage transitional cell carcinoma of the bladder, ibid. 119: 52 (1978).

17. Emmott RC, Javadpour N, Bergman SM, and Soares T: Correlation of the cell surface antigens with stage and grade in cancer of the bladder, ibid. 121: 37 (1979).

18. Jakse G, and Hofstadter F: Further experience with the specific red cell adherence test (SRCA) in bladder cancer, a histological and cytological study, Eur. Ural. 4: 356 (1978).

19. Emmott RC, Droller MJ, and Javadpour N: Tbe A, B, O(H) cell surface antigens in carcinoma in situ and nonmalignant lesions of the bladder, J. Urol. (in press).

20. Melicow MM: Histological study of vesical urothelium intervening between gross neoplasms in total cystectomy, ibid. 68: 261 (1952).

21. Utz DC, et al: Carcinoma in situ of the bladder, Cancer 45: 1842 (1980).

22. Connolly JG, et al: Enzyme hydrolysis in the bladder, its possible applications in the chemotherapy of bladder cancer, Can. J. Surg. 16: 379 (1973).

23. Bryant CT: Urinary bladder cancer: potentials of and prob- lems associated with early intervention strategies, Semin. Oncol. 6: 161 (1979).

24. El-Aaser AA: A study on the etiological factors of bilharzial bladder cancer in Egypt - /%glucuronidase, Eur. J. Cancer 15: 573 (1979).

25. Lamb D: Correlation of chromosome counts with histo- logical appearance and prognosis in transitional cell carcinoma of bladder, Br. Med. J. 1: 273 (1967).

26. Cooper JF, and Foti G: A radioimmunoassay for prostatic acid phosphatase, J. Natl. Cancer Inst. 49: 235 (1978).

27. Choe BK, Pontes EJ, Morrison MD, and Rose NR: Human prostatic acid phosphatase: II. A double antibody radioimmunoassay, Arch. Androl. 1: 227 (1978).

28. Chu TM, et al: Immunochemical detection of serum

prostatic acid phosphatase. Methodology and clinical evaluation, Invest. Ural. 15: 319 (1978).

29. Wajsman Z, et al: Two new, direct, and specific methods of acid phosphatase determination, national field trial, Urology 8: 9 (1979).

30. Quinones G. Rohner TJ, Demers LM, and Drago JR: Will prostatic acid phosphatase determination by radioimmunoassay increase the diagnosis of early prostatic cancer, abstract no. 229, 75th annual meeting AUA, San Francisco, California, 1980.

31. Mooppan MMU, et al: Urinary hydroxyproline excretion as a marker of osseous metastases in carcinoma of the prostate, J. Urol. 123: 694 (1980).

32. Hindsley JP, et al: Creatine kinase BB in prostatic carci- noma, abstract no. 232, 75th Annual Meeting AUA, San Fran- cisco, California, 1980.

33. Waldmann TA, and McIntire KR: The use of a radio- immunoassay for alpha-fetoprotein in the diagnosis of malignancy, Cancer 34: 1510 (1974).

34. Vaitukaitis JL, Braunstein GD, and Ross GT: A radio- immunoassay which specifically measures human chorionic gonadotropin in the presence of human luteinizing hormone, Am. J, Obstet. Gynecol. 113: 751 (1972).

35. Javadpour N: The role of biologic markers in testicular cancer, Cancer 45: 1755 (1980).

36. Javadpour N, McIntire KR, and Waldmann TA: Immuno- chemical determination of human chorionic gonadotropin (HCG) and alpha-fetoprotein (AFP) in sera and tumors of patients with testicular cancer, Natl. Cancer Inst. Monogr. 49: 209 (1978).

37. Javadpour N: The value of biologic markers in diagnosis and treatment of testicular cancer, Semin. Oncol. 6: 37 (1979).

38. Lange PH: Serum alpha-fetoprotein and human chorionic gonadotropin in the management of testicular tumor, Natl. Cancer Inst. Monogr. 49: 215 (1978).

39. Javadpour N, et al: The role of radioimmunoassay of serum alpha-fetoprotein and human chorionic gonadotropin in the intensive chemotherapy and surgery of metastatic testicular tumor, J. Urol. 119: 739 (1978).

40. Searle F, Leake BA, Bagshawe KD, and Dent J: Serum- SP,-pregnancy-specific-3 glycoprotein in choriocarcinoma and other neoplastic disease, Lancet 1: 579 (1978).

41. Rosen WS, Javadpour N, Calvert I, and Kaminska J: Pregnancy specific beta, glycoprotein (SP,) is increased in certain nonseminomatous germ cell tumors, J. Natl. Cancer Inst. 62: 1439 (1979).

42. Javadpour N: Radioimmunoassay (RIA) and immuno- peroxidase (IP) of pregnancy specific beta, glycoprotein in sera and tumor cells of patients with certain testicular germ cell tumors, J. Ural. 123: 514 (1980).

43. Taylor CR, Kurman RJ, and Warner NE: The potential value of immunohistologic technique in classification of ovarian and testicular tumors, Hum. Pathol. 9: 417 (1978).

136 UROLOGY i AUGUST 1980 / VOLUME XVI, NUMBER 2