World J. Surg. 16, 697-702, 1992

World Journal of Surgery © 1992 by the Soci6t~ lnlemationale de Chirurgie

Serum Chromogranin A in the Diagnosis and Follow-Up of Neuroendocrine Tumors of the Gastroenteropancreatic Tract Guido S c h f i r m a n n , M . D . , Ulrich R a e t h , M . D . , B e r t r a m W i e d e n m a n n , M . D . , H e i n z B u h r , M . D . , and Christian H e r f a r t h , M . D . , F . A . C . S . Departments of Surgery and Internal Medicine, University of Heidelberg, Heidelberg and Department of Internal Medicine, Klinikum Steglitz, Free University of Berlin, Berlin, Federal Republic of Germany Hormonally active neuroendocrine tumors may easily be diagnosed by elevated serum levels of their specific peptides and hormonal products, but there are no reliable markers for neuroendocrine tumors without hormonal activity. Chromogranin A (CgA), a secretory protein of neuroendocrine cells, has recently been characterized as a valuable tissue marker in hormonally active and non-functioning neuroendocrine tumors. This study analyzes the role of CgA as a serum marker for different neuroendocrine tumors. Thirty-three patients with neuroendocrine tumors of the stomach (n = 7), the ileum (n = 18), and the pancreas (n = 8) were investigated. Serum CgA levels were analyzed by radioimmunoassay at the time of diagnosis and during follow.up under different therapeutic regimens. Serum CgA was elevated in 30 (91%) patients. Mean CgA serum levels varied with tumor location (pancreas: 7068 -+ 3008 ng/ml, ileum: 5381 -+ 1740 ng/ml, stomach: 529 _+ 179 ng/ml, x + SEM ng/ml) but did not differ between functioning and non-functioning tumors. Eight of l0 patients treated with either somatostatin or interferOn-alpha showed changes of CgA concentrations corresponding to tumor growth. We conclude that CgA is a useful broad-spectrum tumor marker in gastroenteropancreatic neuroendocrine tumors. Its determination is especially recommended in tumors without hormonal activity.

Neuroendocrine tumors with hormonal activity may easily be diagnosed by determination of their specific secretory products. HOwever, there is limited experience with serum markers of horrnonally non-functioning tumors. Chromogranins/secretogranins are a family of secretory proteins [1]. Chromogranin A ((2gA) was the first characterized and identified [2] and has a molecular weight of 48kD according to cDNA analysis [3]. This Polypeptide is co-released from dense core granules of neuroendocrine cells with the resident peptide hormone(s) [4, 5], SUggesting that measurement of plasma CgA may be of value in the diagnosis of such tumors. In 1984, O'Connor and Bernstein [6] developed a radioimmunoassay for serum CgA based on CgA purified from chromaffin granules of a human pheochromocytoma. Since then, potential application of serum CgA as a diagnostic marker for a variety of S Presented at the International Association of Endocrine Surgeons in tOckholm, Sweden, August, 1991. uReprint requests: Dr. Guido Schfirmann, Department of Surgery, R mverslt' y of Heidelberg, Kirsc hnerstr . 1, D-6900 Heidelberg " , Federal epublic ofGermany.

neuroendocrine neoplasias has been presented [7]. However there is limited knowledge about serum CgA in non-functioning neuroendocrine tumors and about the effect of various therapeutic regimens on CgA serum concentrations. The purpose of the present study was to determine serum CgA concentrations in patients with histologically confirmed neuroendocrine tumors of the gastroenteropancreatic tract and to measure the course of CgA serum levels under different therapeutic modalities. Material and Methods

Patients Thirty-three patients undergoing surgery for neuroendocrine tumors from 1986 to 1991 were included in this study (Table 1). There were 7 patients with gastrinoma, all of them without metastases. Fifteen of 18 patients with ileal neuroendocrine tumors showed metastatic growth at the time of diagnosis, liver metastases with or without intestinal lymph node metastases (n = 14) and lymph node metastases and peritoneal metastases without liver metastases (n = 1). A carcinoid syndrome was present in half of the patients with midgut neuroendocrine tumors. In 8 patients with pancreatic neuroendocrine tumors, 2 patients had a glucagonoma, both presenting with liver metastases, and 6 patients had non-functioning tumors. Four patients had mild hypertension with systolic blood pressure from 130 mmHg to 150 mmHg.

Tumor Therapy Surgery. In 5 patients with hormonally active ileal neuroendocrine tumors serum CgA was determined pre-operatively and 8 to 10 days postoperatively. All 5 patients presented with multiple liver metastases. Palliative resection of the primary tumor was necessary due to local intestinal stenosis. Interferon-Alpha. In 5 patients serum CgA concentrations were determined before and 6 months after starting treatment with interferon-alpha (3 MU/m 2 recombinant interferon-alpha-2a, 3

698

World J. Surg. Voi. 16, No. 4, July/Aug. 1992

Table 1. Clinical data of 33 patients with neuroendocrine tumors of the gastroenteropancreatic tract. Tumor location

No. of pts. (male:female)

Mean age (yrs) (range)

Hormonal activity (%)

Metastases at diagnosis (%)

Foregut Midgut Pancreas

7 (4:3) 18 00:8) 8 (4:4)

53.8 (37-75) 52.4 (29-71) 57.8 (37-76)

7 (100) 9 (50) 2 (25)

0 (0) 15 (83) 6 (75)

30-

CgA~no/mll x 10

CgA 108[ng/ml] 7"~ n-7

n-18

n-8 6

25

5

20

4 15 3

I

10

21

' O"

°°.%°

JL.

*~.

STOMACH (529"_179}

ILEUM (5381*_1740}

PANCREAS (7068-+3008)

,f

0 preop, [,r(",SEM]

Fig. 1. Chromogranin A serum levels of 33 patients with neuroendocrine tumors of the gastroenteropancreatic tract. *: upper reference limit of chromogranin A serum concentrations. times/week subcutaneously; Roferon-A, Hoffmann-La Roche, Switzerland). The primary tumors had been resected before beginning interferon therapy in all patients. Treatment with interferon-alpha was instituted due to widespread liver metastases and hormonal activity of the tumors. After 6 to 9 months of treatment, restaging of liver metastases was performed by computed tomography and ultrasound. Tumor progression was found in 4 patients and stable disease was found in 1 patient.

Somatostatin. In 5 patients serum CgA concentrations were determined before and 6 months after starting treatment with the somatostatin analogue ocreotide (3 times 0.2 mg/day; Sandostatin, Sandoz, Switzerland). Liver metastases were present in all cases. In 4 patients, primary tumors had been resected before somatostatin therapy and in 1 patient an intestinal bypass procedure had been performed. After 6 to 9 months of treatment, tumor progression was found only in the latter patient, whereas there was stable disease in the remaining 4 patients. Laboratory Analysis Samples of 21 mi serum probes were deep frozen and stored at minus 20°C until the cumulative analysis. Serum concentration of CgA was determined by a radioimmunoassay with CgA antiserum [8, 9]~ The upper reference level was 100 ng/ml.

Results In 30 (91%) of 33 patients with neuroendocrine tumors of the gastroenteropancreatic tract serum CgA levels were above the upper reference limit (Fig. 1). CgA concentrations varied with the site of the primary tumor with pancreas > ileum > stomach.

i poslop.

i

Fig. 2. Chromogranin A serum levels of 5 patients before and after palliative resection of neuroendocrine tumors, preop: pre-operatively; postop: postoperatively.

In each subgroup there was 1 patient with a false-negative result. The corresponding clinical findings were 1 patient with a small gastrinoma, 1 patient with a non-functioning ileal tumor with peritoneal metastasis, and 1 patient with a small neuroendocrine tumor of the pancreas, incidentally found in a patient with colorectal carcinoma. Mean CgA concentration in 15 patients with non-functioning tumors was 3324 - 1090 ng/ml. This is close to the corresponding concentration of CgA in patients with hormonaily active tumors, 3013 -+ 986 ng/ml (n = 18). Eight to 10 days after palliative resection of the primary tumors CgA serum levels remained grossly unchanged in patients with widespread liver metastases (Fig. 2). Six months after starting interferon-alpha therapy, serum CgA increased in 3 of 5 patients, corresponding to tumor progression (Table 2). In another patient serum CgA decreased to 2% of its pretherapeutic level. At that time, computed tomography revealed stable disease. One remaining patient treated with interferon-alpha showed tumor progression under therapy with a decrease in CgA serum concentrations. Somatostatin tumor treatment was associated with changes of serum CgA levels in accordance to tumor response to therapy in 4 patients (Table 3). Another patient presented with an extremely high serum CgA level after an intestinal bypass procedure. Under somatostatin treatment, serum CgA decreased dramatically corresponding to depressed hormonal activity, however, the patient died from tumor progression after 6 months. Additionally, long-term follow-up of serum CgA levels was undertaken in 2 selected patients. Patient A: A 62 year old male patient presented with hormonally active midgut carcinoid with liver metastases (Fig. 3). After resection of the primary tumor he received interferon-alpha for 12 months. CgA serum levels

G, Schiirmann et al.: Neuroendocrine Tumors of the Gastroenteropancreatic Tract

699

Table 2. Chromogranin A serum levels before and after 6 months of treatment with interferon-alpha.

Patient age (yrs)

Sex

Site of primary tumor

CgA (ng/ml) Before treatment

After treatment

Tumor response

Hormonal activity

62 70 52 61 60

Male Male Male Female Female

Ileum Ileum, colon Pancreas Ileum Colon

958 10,000 9,378 6,710 6,980

2,782 203 7,860 I0,950 9,074

Yes Yes No Yes Yes

Yes Yes Non-functioning Yes Non-functioning

Tumor response

Hormonal activity

Yes Yes Yes Yes No

Yes Yes Non-functioning Non-functioning Yes

-....._

CgA: Serum chromogranin A. Table 3. Chromogranin A serum levels before and after 6 months of treatment with somatostatin. Patient age (yrs)

Sex

Site of primary tumor

CgA (ng/ml) Before treatment

After treatment

50 62 52 60 54

Male Male Male Female Male

Ileum Ileum Pancreas Colon Ileum

1,710 2,270 142 1,610 24,820

760 48 3,770 4,227

521

CgA: Serum chromogranin A.

II CgA 7 X10a Ing/rnll

[]

5HIAA

Img/24h]

Livermetaetaae= (CT. 8ono)

CgA

120

2O

lO3[ng/ml]

ivermelast . . . .

nochange

]

635241 "~ i ~ n ° _ _ _ _ _ _ c h a n g e

60804020100

(CT,

So---n-O-)-r increase

15 105

07 12 3 1989 OP r'~r"l

i

i

i

6

i

i

i

9

L

,

i

~

,

,

i

i

J

,

F

,

i

i

~

12 3 6 9 12 9 1990 1991 Interferon- ~ -- Somatoatatln

=

O

l~ig. 3. Chromogranin A serum levels, urinary concentrations of 5-hydroxyindoleacetic-acid and growth of liver metastases during interferOn-alpha therapy and during somatostatin therapy in a 62 year old male Patient with an ileal carcin0id.

Started at 1000 ng/ml but increased after 6 months of treatment to about 2700 ng/ml thus anticipating tumor growth subsequently confirmed by computed tomography. Medical therapy Was changed from interferon-alpha to somatostatin. CgA deCreased to 700 ng/ml in accordance to reduction of the carcinoid Syndrome and to stable disease of liver metastasis. In March 1991, however, serum CgA again increased to more than 6000 ag/ml, corresponding to tumor growth. A t this time, there was also an increased frequency of flush episodes and 5-hydroxyindoleacetic-acid concentrations in the patient's urine. Patient B: A 51 year old male patient with non-functioning metastatic Pancreatic neuroendocrine tumor was treated with interferonalpha after resection of the primary tumor (Fig. 4). CgA serum levels remained elevated over the months corresponding to a large metastases in liver Segment 7/8 and other metastases in the left lobe of the liver. Chemoembolization of the right-sided liver metastases was performed due to painful capsular tension,

0

.

.

.

.

9

.

.

.

.

.

.

.

12 3 1989 t 1990 1 OP Interferon-

Chernoernboltsation

"r

6

.

.

.

.

9

.

= , Ill 12 3 6 1991 ....... Somatostatin .

.

.

.

Fig. 4. Chromogranin A serum levels and growth of liver metastases under different therapeutic modalities in a'51 year old male patient with a pancreatic neuroendocrine tumor.

resulting in subtotal necrosis of the tumor. Somatostatin therapy followed. CgA serum levels decreased rapidly almost to normal values and have remained low f o r > l year. Discussion

Serum chromogranin A has been found to be elevated in small series of patients with pheochromocytoma, aortic body tumor, pancreatic islet cell tumor, medullary thyroid carcinomas, C-cell hyperplasia, and in a variety of pituary tumors [1, 8, 10], Furthermore, serum CgA correlates with serum creatinine [11] with the highest values attained in patient.s with end-stage renal disease probably due to reduced renal ~embval of CgA from circulation [7]. It is elevated in essential hypertension and may be depressed by some antihypertensive drugs [12, 13]. In the present series, serum CgA was increased in 71% of the

700

patients with confirmed gastrointestinal neuroendocrine tumors (Fig. 1). In accordance to reports by others [10, 14], CgA concentrations in gastrinoma patients were much lower as compared to other gastroenteropancreatic tumor iocalizations (Fig. 1). This may be due to low tumor burden in these neoplasias with most patients presenting without metastases. A correlation between CgA levels and tumor mass has been shown in pheochromocytomas [15]. Serum CgA concentrations did not differ between patients with functioning and non-functioning neuroendocrine tumors, providing evidence that in addition to CgA co-release with other peptides in functioning tumors there may be a second mode of independent secretion of CgA. In humans, stimulation of normal enteroendocrine secretory cells (and other endoendocrine cells except adrenal medulla) with cell-type selective secretagogues failed to induce significant increments in serum CgA despite inducing measurable increments in the concentration of the residual peptide hormones [9]. Immunohistochemical demonstration of CgA expression in tissues of non-functioning neuroendocrine tumors [14, 16-18] support the concept of a secretory autonomy of CgA. For its immunoreactivity in hormonally active and quiescent tumors CgA has been referred to as a "panendocrine" or "broadspectrum" marker [17]. There is still controversy about parallels between serum CgA concentrations and the amount of peptide production in hormonally active tumors. In accordance to our experience (Fig. 3), Moattari and coworkers [10] found a weak association between CgA levels and either flushing or diarrhea in patients with carcinoid tumors and hormone levels in patients with other hormonally active neuroendocrine tumors. In untreated patients with gastrinoma, serum CgA levels did not correlate with serum gastrin values [19]. Thus in tumors with hormonal activity, CgA levels seem to be a less sensitive marker than the appropriate basal hormone concentration specific for the endocrine syndrome [10]. During interferon-alpha therapy, serum CgA changes reflected tumor response to therapy in 4 of 5 patients including 3 patients who presented with progressive disease (Table 2). The benefit of treating malignant carcinoid tumors with interferonalpha has been demonstrated repeatedly [20-22], However, the mechanism of interferon action on neuroendocrine tumor cells in general and on the production and release of secretory products remains to be elucidated. Somatostatin decreased CgA levels dramatically in 4 of 5 patients reflecting tumor response in 3 of them. In another patient CgA decreased despite tumor progression but parallelled clinical and biochemical reduction of hormonal activity (Table 3) indicating the suppressive effect of somatostatin on both hormone secretion and CgA release [23]. To our knowledge, the only clinical report on CgA serum levels and somatostatin therapy [10] shows that changes in CgA parallelled the corresponding changes in plasma and urinary hormones (vasoactive intestinal peptide in VIPoma, gastrin in gastrinoma, glucagon in glucagonoma and 24 hour urine 5-hydroxyindoleacetic-acid in patients with carcinoid syndrome). Serum CgA did not correlate with blood serotonin. Hence, CgA may be important in neuroendocrine tumors with negative 5-hydroxyindoleacetic-acid excretion, especially since the only abnormal

World J. Surg. Vol. 16, No. 4, July/Aug. 1992

marker, blood serotonin, does not change with treatment or remission of symptoms [24]. In conclusion, CgA is a useful broad-spectrum marker in the diagnosis ofgastroenteropancreatic neuroendocrine tumors. Its determination is especially recommended in tumors without hormonal activity, since other reliably measurable secretory tumor products are not available. Measurements of CgA under medical therapy may provide additional means for monitoring tumor response. R~sum~

Le diagnostic d'une tumeur neuro-endocrine s6cr6tante du point de vue hormonale est facile lorsqu'existent des taux s6riques 61ev6s des peptides et des hormones s6cr6t6es sp6cifiquement par ta tumeur, alors qu'il n'existe aucun moyen de faire le diagnostic de tumeurs neuro-endocrines qui ne sont pas s6cr6tantes. La chromogranine A (CgA), une pro t6ine s6cr6toire des cellules neuro-endocrines, a r6cemment 6t~ raise en 6vidence comme un marqueur aussi bien dans les tumeurs s6cr6tantes que non s6cr6tantes, Cet article a comme but d'6tudier le r61e de CgA comme marqueur de diff6rentes tumeurs neuro-endocrines. Trentetrois patientg ayant une tumeur neuro-endocrine de l'estomac (n = 7), de l'intestin gr61e (n = 18) ou du pancr6as (n = 8) ont eu ce dosage. Les taux de CgA ont 6t6 dos6s par des techniques radioimmunologiques au moment du diagnostic et dans la p6riode postop6ratoire avec des r6gimes th6rapeutiques diff6rentes. Le taux de CgA 6tait 61ev6 chez 30 patients (91%). Le taux s6rique moyen de CgA variait scion la localisation anatomique de la tumeur (pancr6as: 7068 --- 3008; intestin gr~le: 5381 ± 1740; estomac: 529 +-- 179 (moyenne - erreur standard de la moyenne), mais il ne diff6rait pas dans les tumeurs fonctionnelles ou non-fonctionnelles. Huit des dix patients trait6s soit par la somatostatine, soit pas l'interf6ron alpha avaient des taux 61ev6s de CgA en rapport avec une croissance tumorale. On conclue que la CgA est un marqueur tumoral utile dans une grande vari6t6 de tumeurs neuro-endocrines. On recommande de la doser surtout dans les tumeurs non s6cr~tantes du point de rue hormonal, Resumen

Los tumores neuroendocrinos hormonalmente activos pueden ser f~icilmente diagnosticados pot la elevaci6n de los niveleS s6ricos de sus productos peptfdicos y hormonales especfficos, pero no existen marcadores confiables de los tumores neuroen" docrinos desprovistos de actividad hormonal. La cromogranina A (CgA), que es una protefna secretora de las c61ulas neuroen" docrinas, ha sido recientemente identificada como un marcador tisular de valor tanto en los tumores neuroendocrinos hor" monalmente activos como en los tumores no funcionantes. Ea el presente trabajo se estudia el papel de la CgA como marcador s6rico para la identificaci6n de diferentes tumores neuroendo" crinos. Se investigaron 33 pacientes con tumores neuroendocri" nos dei est6mago (n = 7), el ileon (n = 18) y el pfincreas (n = 8). Los niveles de CgA fueron analizados por radioinmunoan,51isiS en el momento del diagn6stico y en el curso del suguimientO bajo diferentes regfmencs terap6uticos. Se encontr6 elevaci6O de la CgA en 30 pacientes (91%), con variaci6n del nivel promedio segtin la ubicaci6n del tumor: 7068 - 3008 para el

G. Schiirmann et al.: Neuroendocrine Tumors of the Gastroenteropancreatic Tract

PAncreas, 5381 -+ 1740 para el fleo y 529 ---+179 para el estrmago, x ± SEM ng/ml, pero sin diferencias entre los tumores funcionantes y los no funcionantes. Ocho de diez pacientes tratados con somatostatina o con interferrn alfa exhibieron cambios en las concentraciones sdricas de CgA, correspondientes al erecimiento del tumor. La c o n c l u s i r n es que la CgA es un marcador tumoral de utilidad en los tumores neuroendocrinos gastroenteropancr~iticos y que se recomienda su determinacirn especialmente en los tumores desprovistos de actividad hormonal. Acknowledgment

The authors are in debt to Prof. Dr. Schmidt-Gayk (Laboratories Im Breitspiel, D-6900 Heidelberg, Germany) for performing serum analysis and t h e authors thank Dr. Babro Eriksson, Ludwig-Institute for Cancer Research, Uppsala, Sweden for additional serum analysis. CgA antibody was kindly provided by Daniel T. O ' C o n n o r (Department of Medicine, University of California, San Diego, California, U.S.A.). References

1. Wiedenmann, B., Huttner, W.B.: Synaptophysin and chromogranins/secretogranins: Widespread constituents of distinct types of neuroendocrine vesicles and new tools in tumor diagnosis. Vichows Arch. [B] 58:95, 1989 2. Smith, A.D., Winkler, H.: Purification and properties of an acidic protein from chromaffin granules of bovine adrenal medulla. Biochem. J. 103:483, 1967 3. Benedum, U.M., Baeuerle, P.A., Konecki, D.S., Frank, R., Powell, J., Mallet, J., Huttner, W.B.: The primary structure of bovine chromogranin A: A representative of a class of acidic secretory proteins common to a variety of peptidergic cells. EMBO J. 5:1495, 1986 4. O'Connor, D.T., Frigon, R.P.: Chromogranin A: The major catecholamin storage vesicle soluble protein. J. Biol. Chem. 259:3237, 1984 5. Eiden, L.E., Huttner, W.B., Mallet, J., O'Connor, D.T., Winkler, H., Zanini, A.: A nomenclature proposal for the chromogranin/ secretogranin proteins. Neuroscience 21:1019, 1987 6. O'Connor, D.T., Bernstein, K.N.: Radioimmunoassay of chromogranin A in plasma as a measure of exocytotic sympathoadrenal activity in.normal subjects and patients with pheochromocytoma. N. Engl. J. Med. 311:764, 1984 7. Takiyyudin, M.A., Barbosa, J.A., Hsiao, R.J., Parmer R.P., O'Connor, D.T.: Diagnostic value of chromogranin A measured in circulation. In Markers for Neural and Endocrine Cells, M. Gratzl, K. Langley, editors, Weinheim-New York-Basel-Cambridge, VCH-Verlag, 1991, pp. 191-201 8. O'Connor, D.T., Deftos, L.J.: Secretion of chromogranin A by peptide-producing endocrine neoplasms. N. Engl. J. Med. 314: 1145, 1986

Invited Commentary Stuart D. Wilson, M.D. epartment of Surgery, Medical College of Wisconsin, Milwaukee, isconsin, U.S.A.

With the development of radioimmunoassay techniques, one Can now measure minute quantities of hormones circulating

701

9. O'Connor, D.T., Pandian, M.R., Cervenka, J., Mezger, M., Panner, R.J.: What is the source and disposition of chromogranin A in normal human plasma? Clin. Res. 35:605, 1987 10. Moattari, A.R., Deftos, L.J., Vivnik, A.I.: Effects of sandostatin on plasma chromogranin A levels in neuroendocrine tumors. J. Clin. Endocrinol. Metab. 69:902, 1989 11. Hsiao, R.J., Mezger, M.S., O'Connor, D.T.: Chromogranin A in uremia: Progressive retention of immunoreactive fragments. Kidney Int. 37:955, 1990 12. O'Connor, D.T.: Plasma chromogranin A: Initial studies in human hypertension. Hypertension 7:176, 1985 13. Takiyyudin, M.A., Cervenka, J.H., Hsiao, R.J., Barbosa, J.A., Parmer, R.J., O'Connor, D.T.: Chromogranin A: Storage and release in hypertension. Hypertension 15:237, 1990 14. Eriksson, B., Arnberg, H., 0berg, K., Hellmann, U., Lundqvist, G., Wernstedt, C., Wilander, E.: Chromogranins: New sensitive markers for neuroendocrine tumors. Acta Oncol. 28:325, 1989 15. Hsiao, R.J., Parmer, R.J., Takiyyudin, M.A., O'Connor, D.T.: Chromogranin A storage and secretion: Sensitivity and specificity for the diagnosis of pheochromocytoma. Medicine (Baltimore) 70:33, 1991 16. Wiedenmann, B., Waldherr, R., Buhr, H., Hille, A., Rosa, P., Huttner, W.B.: Identification of gastroenteropancreatic neuroendocrine cells in normal and neoplastic human tissue with antibodies against synaptophysin, chromogranin A, secretogranin I (chromogranin B), and secretogranin II. Gastroenterology 95:1364, 1988 17. Sch~irmann, G., Betzler, M., Buhr, H.: Chromogranin A, neuronspecific enolase and synaptophysin as neuroendocrine cell markers in the diagnosis of tumours of the gastro-entero-pancreatic system. Eur. J. Surg. Oncol. •6:298, 1990 18. Sobol, R.E., Memoli, V., Deftos, L.J.: Hormone-negative, chromogranin A positive endocrine tumors. N. Engl. J. Med. 7:444, 1989 19. Stabile, B.E., Howard, T.J., Passaro, E., O'Connor, D.T.: Source of plasma chromogranin A elevation in gastrinoma patients. Arch. Surg. 125:451, 1988 20. 0berg, K., Funa, K., Aim G.: Effects of leucocyte interferon on clinical symptoms and hormone levels in patients with mid-gut carcinoid tumors and carcinoid syndrome. N. Engl. J. Med. 309: 129, 1983 21. 0berg, K., Norheim, I., Lind, E., Aim, G., Lundqvist, G., Wide, L., Jonsdottir, B., Magnusson, A., Wilander, E.: Treatment of malignant carcinoid tumors with human leucocyte interferon: Long-term results. Cancer Treat. Pep. 70:1297, 1986 22. Nobin, A., Lindblom, A., Mansson, B., Sundberg, M.: Interferon treatment in patients with malignant carcinoids. Acta Oncol. 28: 445, 1989 23. Takiyyudin, M.A., Baron, A.D., Cervenka, J.H., Barbosa, J.A., Neumann, H.P., Parmer, R.J., Sullivan, P.A., O'Connor, D.T.: Suppression of chromogranin A release from neuroendocrine sources in man: Pharmacological studies. J. Clin. Endocrinol. Metab. 72:616, 1991 24. Vinik, A.I., Tsai, S., Moattari, A.R., Cheung P.: Somatostatin alalogue (SMS 201995) in patients with gastrinoma. Surgery 104: 834, 1988

in the patient's serum that are elaborated by "functioning" endocrine neoplasms, i.e., gastrinoma, insulinoma, etc. Many endocrine tumors, however, are " n o n - f u n c t i o n i n g " and do not elaborate a specific hormone marker that can be measured, at least by currently available methods. A secretory product or " m a r k e r " that can be measured in the serum and tissues of patients that harbor non-functioning neuroendocrine neoplasms would have great utility. There are obvious diagnostic and therapeutic applications for such a marker. This study evaluates the role of serum chromogranin A (CgA)