Gen-You Yao, Ji-Lin Zhou, Mao-De Lai, Xiao-Qing Chen, Pei-Hui Chen, Department of Pathology, Zhejiang University Medical School, Hangzhou 310031, China
Supported by Fundation of Health Bureau of Zhejiang Province
Correspondence to: Gen-You Yao, Research fellow of Pathology, Department of Pathology, Zhejiang University Medical School, Hangzhou, 310031 China. yaogy@zjuem.zju.edu.cn
Telephone: +86-0571-87217167
Received: 2002-10-25 Accepted: 2002-11-18

Abstract
AIM: To investigate the incidence of neuroendocrine (NE) cells and their hormone products in adenocarcinomas and evaluate their significance in clinical pathology and prognosis.

METHODS: By using tissue sectioning and immunocyto-chemistry, 356 cases of adenocarcinomas were studied to examine the presence of chromorgranin and polypeptide hormones in adenocarcinoma samples from our hospital.

RESULTS: The positive rate of NE cells and hormone products was 41.5 % (54/130) and 59.3 % (32/54), respectively in large intestinal adenocarcinoma cases; 39.6 % (38/96) and 36.8 % (14/38), respectively in gastric cancer cases; 38.1 % (8/21) and 50.0 % (4/8), respectively in prostatic cancer cases; 21.0 % (17/81) and 17.6 % (3/17), respectively in breasr cancer cases; 17.9 % (5/28) and 60.0 % (3/5), respectively in pancreatic cancer cases. Among carcinomas of large intestine, pancreas and breast, the highly differentiated NE cell numbers were higher than the poorly differentiated NE cell numbers; while the gastric carcinoma cases had more poorly differentiated NE cells than highly differentiated NE cells. The higher detection rate of NE cells and their hormone products, the higher 5-year survival rate among the large intestine cancer cases.

CONCLUSION: Close correlation was observed between NE cells and their hormone products with the cancer differentiations. For colorectal carcinomas, there is a close correlation of the presence of NE cells and their hormone products with the tumor staging and prognosis.

Yao GY, Zhou JL, Lai MD, Chen XQ, Chen PH. Neuroendocrine markers in adenocarcinomas: an investigation of 356 cases. World J Gastroenterol 2003; 9(4): 858-861
http://www.wjgnet.com/1007-9327/9/858.htm
　

INTRODUCTION
Compared with neuroendocrine cancers, little investigation is carried out on the relationship of neuroendocrine cells and their hormone products in non- neuroendocrine cancers, especially in the common adenocarcinoma cases. By using nine different antibodies and immunocytochemistry, NE cells and their hormone products in 356 adenocarcinamas was observed with the aim of revealing the incidence and distribution of NE cells and the correlation between the cancer differentiation with the biological behaviors was evaluated.

MATERIALS AND METHODS
Materials
All the 365 adenocarcima samples were got from the first affiliated hospital of Zhejiang University Medical College from 1975 to 1994. There were 96 cases of gastric cancer samples (31 samples were got from the clinical biopsy; 65 samples were got from radical operation and 22 samples had lymph nodes metastasis); there were 130 case of large intestine cancer samples got from radical operations (110 cases had the follow-up data); there were 81 and 28 case of breast and panereatic cancer, respectively. The remaining 21 samples were got from prostatic cancer biopsy.

Methods
All the samples were fixed with 10 % formaldehyde with paraffin embedding and continuous sectioning at 4 mM in thickness. Gross pathological observation was made on the HE stain slides followed by immunocytochemistry. All the samples were treated with anti-chromogranin serum fot the primary screening positive cases. Further immunocytochemistry was carried out for those positive NE samples by using peptide hormone antibodies such as ST (diluted at 1:10 000, provided by the 4th Military Medical Academy) and other Dako抯 antibodies (somatostatin diluted at 1:300; glucagon diluted at 1:800; pancreatic polypeptide diluted at 1:800; gastrin diluted at 1:350; insulin diluted at 1:150; ACTH diluted at 1:800 and calcitonin diluted at 1:150). The immunostains were done by ABC method and coloured with AEC. The antiserum of serotonin and gastrin was used in the gastric mucosa; the pancreas tissue was used to detect the chromogranin, somatostatin, glucagon, insulin and pancreatic polypeptide; calcitonin antiserum was used in the medullary carcinoma of the thyroid gland while ACTH in the pituitary was used as the positive control. The negative control was carried out by using normal sheep serum to replace the 1^st antibody. Based on the chromoganin positive NE cell numbers, all the samples were divided into three grades as the following. Negative: there was no NE cells; Positive(+): the number of NE cells was fewer than 5/mm2; Super positive (++): the number of NE cells was over 5/mm².

Statistical analysis
The date were analysed by x² test.

RESULTS
Morphology of NE cells and their incidence
Among the five common adenocarcinomas from different tissue sources, the incidence rate was 41.5 % (54/130) for the large intestinal carcinomas; 39.6 % (38/96) for the gastric carcinoma; 38.1 % (8/21) for the prostatic carcinoma; 21.0 % (17/81) for the breast cancer and 17.9 % (5/28) for the pancreatic cancers, respectively. The highest incidence was seen in large intestinal carcinomas while the lowest in the pancreatic carcinomas. When observing the chromoganin stained slides, clear edges of NE cells and brownish granules could be seen in the cytoplasm under the microscopy. In the low differentiated carcinomas, the NE cells presented as an oval, round or irregular shape without polarizations. Abnormal structural characteristics were observed among these low differentiated NE cells, which was similar to the adjacent tumor cells; while for those highly differentiated carcinomas, the NE cells were pyramid or bar shaped with the apex pointing to the cavity of the gland. A few NE cell processes could be observed reaching the gland cavity surfaces. The distribution of NE cells were scattered or localized infiltrating all the layers with the cancer cells. NE cells could be seen in both of the primary carcinoma and the metastasis sites.

Relationship between NE cells and carcinoma differentiation
No exact correlation between NE cells and carcinoma differentiation was observed among different carcinomas. The highly differentiated NE cell incidences were 41.7 % (5/12) for the large intestinal carcinomas, 42.9 % (3/7) for the pancreas carcinomas and 32.5 % (14/43) for the breast cancers, which was much higher than that of the low differentiated carcinomas. Prostatic carcinomas had the same tendency but there was no statistical significance due to fewer case numbers. In the low differentiated gastric carcinomas, 50 % (27/54) had the positive NE cells, which was significantly higher than that of the highly differentiated carcinomas.

Distribution of hormone products of NE cells in tumors
From Table 1, the number of hormone products types was more in large intestinal and gastric carcinomas (5 types of hormone products); hormone products detected in breast cancers were the fewest (only three in 17 cases). Most of them were the tumor origin tissue hormones, but few of them were ecotopic hormones.

Relationship between positive cell of hormone products and tumor differentiation
In the large intestinal carcinomas, 9 cases were low differentiated carcinomas whose positive cell percentage of hormone products against the total NE cells was 27.0 %, which was obviously lower than that in high differentiated carcinomas (15 cases with the percentage of 43.9 %) (x²=115.9, P<0.01); It was also the same in the highly differentiated large intestinal carcinomas whose percentage was lower than that in the normal mucus membrane (10 cases with the percentage of 83.1 %)(x²=212.3, P<0.01) and the mucus membranes adjacent to the tumors (25 cases with the percentage of 88.7 %) (x²=168.8, P<0.01). The gastric carcinoma had the similar results: the positive cell percentage of hormone products against the total NE cells was 17.5 %. But in the positive cells of hormone products from 5 gastric sinus mucus membranes, the positive cell percentage of hormone products against the total NE cells was 78.6 % (x²=1611.8, P<0.01); the samples adjacent to the gastric sinus areas had the obviously higher percentage (46.6 %, x²=266.4, P<0.01). Significant difference was also observed between the percentage of the adjacent mucus membrane tissues of the tumors and the normal mucus membranes (x²=242.0, P<0.01).

Ecotopic hormones and tumor differentiation
Except for the pancreatic carcinomas, ecotopic hormones were revealed in other four types of the adenocarcinomas. One of the large intestinal carcinomas cases showed gastrin positive; Six gastric carcinoma cases showed ACTH positive; Two prostatic cancer cases were glucagons positive; One breast cancer case was somatostain positive while another breast cancer case was calcitonin positive. Except for the large intestinal and gastric carcinomas, all the other nine cases were low differentiated carcinomas among the ecotopic hormone carcinomas.

NE cells and tumor differentiation
Observed in large intestinal carcinomas, Dukes A stage accounted for 41.7 % of the NE cell (++) cases (12 cases), which was much higher than that in NE(-) group (19.7 %,76 cases). (x²=4.668, P<0.05). Among the 110 cases with following-up, the 5-year survival rate was 81.8 % in NE cell (++) group, which was obviously higher than that in the NE (+) group (45.7 %, 35 cases) (x²=4.000, P<0.05) and in NE cell (-) group (42.2 %, 64 cases) (x²=4.397, P<0.05).
   Among the 32 hormone products positive cases with polypeptide hormones (PH), Dukes A stage cases accounted for 44.1 %, which was higher than that of NE cell positive cases with hormone products negative (36.1 %), yet no statistical difference was found between the two groups (x²=0.351, P>0.05). In hormone products (+) group (17 cases), the 5-year survival rate was 70.6 %, which was higher than that of hormone products (-) group (37.9 %, 29 cases) (x²=4.148, P<0.05).

Table 1 Distribution of hormone productions of NE cells in tumors

DISCUSSION
The commonly used staining methods for revealing NE cells include silver staining, neuron-specific enolase (NSE), synaptophysin (SY) and chromorgranin (CG) immunocytochemistry. Silver staining is the traditional staining method with less specificity and sensitivity. Although NSE, CG and SY are all the common markers, NSE has poor specificity with distributions in different tissues but is localized in the cytoplasm. CG is distributed in neuroendocrine granules. Both CG and SY are good markers and corresponding to respective subcellular structures. CG is a specific matrix component of endocrine granules^[1-3]. While SY is localized within small capsule membranes related to the secretion granules, whose specificity and sensitivity are less than those of CG. That's why CG is considered as a realistic marker for NE cells^[4-8]. Studies have confirmed that CG could be served as a new way of revealing NE cells and for the diagnosis of NE tumors^[9].
   Our study demonstrated that NE cell numbers were closely correlated to the tumor differentiation in large intestinal, pancreatic, breast and prostatic carcinomas^[10]. The higher differentiated tumors had the higher incidence of the NE cells^[8]. That was contradicted to our gastric carcinoma observations, but corresponding to the publication reports^[11]. Further studies are needed to be conducted to reveal these differences to see if they are related to the embryology, etiology and tissue development of the tumors.
   No serial report were seen about the hormone products of NE cells from the common adenocarcinomas. We observed 5 types of adenocarcinomas and found out that large intestinal and gastric carcinomas had the higher hormone products in their NE cells; but in the gastric and highly differentiated carcinomas, they had lower hormone products in NE cell than those of poor differentiated carcinomas. The hormone products were more in the large intestinal and gastric carcinomas than those in normal mucus membranes and tissues adjacent to the carcinomas. Neoplastic NE cells had lower hormone products and they were decreased with anaplasia which may be due to the fact that these cells were in immature state with lower hormone synthesis. Thus, the amount of hormone products in NE cells of the carcinomas can serve as an index for the determination of tumor differentiation and the diagnosis of benign and malignant tumors^[12-15].
   Based on the study of the large intestinal carcinomas, we found that hormone products and distribution of NE cells were closely correlated to the tumor grade, clinical pathological stage of the tumor and prognosis of the patients^[16-23]. Carcinomas with NE (++) releasing PH were the early stage carcinomas. The higher 5-year survival rate may be due to the somatostatin's inhibition of the tumors^[24-29]. Zollinger-Ellison syndrome was reported in some gastric carcinoid cases^[30-32], but our study only revealed there were only different hormone products but without sign and symptoms of Zollinger-Ellison syndrome as well as other hormone signs and symptoms, which may be the fact that the hormone products produced by NE cells were not enough in the inactivated form or inactivated by the liver enzymes. Further study is required to examine whether these hormone products participate in the immune regulations of the tumor or the hormone products exert the direct effects on the tumor development and growth.

REFERENCES
1    Papotti M, Macri L, Finzi G, Capella C, Eusebi V, Bussolati G. Neuroendocrine differentiation in carcinomas of the
      breast:a study of 51 cases. Semin Diagn Pathol 1989; 6: 174-188
2    Van Laarhoven HA, Gratama S, Wereldsma JC. Neuroendocrine carcinoid tumours of the breast: a variant of carcinoma
      with neuroendocrine differentiation. J Surg Oncol 1991; 46: 125-132
3    Giovanella L, Marelli M, Ceriani L, Giardina G, Garancini S, Colombo L. Evaluation of chromogranin A expression in serum
      and tissues of breast cancer patients. Int J Biol Markers 2001; 16:268-272
4    Kimura N, Sasano N, Yamada R, Satoh J. Immunohistochemical study of chromogranin in 100 cases of
      pheochromocytoma, carotid body tumour, medullary thyroid carcinoma and carcinoid tumour. Virchows Arch A Pathol
      Anat Histopathol 1988; 413: 33-38
5    Kimura N, Hoshi S, Takahashi M, Takeha S, Shizawa S, Nagura H. Plasma chromogranin A in prostatic carcinoma
      and neuroendocrine tumors. J Urol 1997; 157: 565-568
6    Portel-Gomes GM, Grimelius L, Johansson H, Wilander E, Stridsberg M. Chromogranin A in human neuroendocrine tumors:
      an immunohistochemical study with region-specific antibodies. Am J Surg Pathol 2001; 25: 1261-1267
7    Bernini GP, Moretti A, Ferdeghini M, Ricci S, Letizia C, D'Erasmo E, Argenio GF, Salvetti A. A new human chromogranin
      'A'immunoradiometric assay for the diagnosis of neuroendocrine tumours. Br J Cancer 2001; 84: 636-642
8    Yao GY, Chen PH, Lai MD, Hong LY. Study of 9 neuroendocrine markers in pancreatic tumors. Zhenjiang Yike Daxue
      Xuebao 1995; 24: 56-58
9    Chen FX, Corti A, Siccardi AG. DIBIT; San Raffaele H. Characterizzation of antigenic sites of Human Chromogranin A.
      Shanghai Mian Yixue Zazhi 1998; 18: 215-219
10  di Sant'Agnese PA. Neuroendocrine differentiation in prostatic carcinoma: an update. Prostate Suppl 1998; 8: 74-79
11  Chen BF, Yin H. Neuro-endocrine type of gastric carcinoma. Immunohistochemical and electron microscopic studies of
      100 cases. Chin Med J 1990; 103: 561-564
12  Yu JY, Wang LP, Meng YH, Hu M, Wang JL, Bordi C. Classification of gastric neuroendocrine tumors and its
      clinicopahologic significance. World J Gastroenterol 1998; 4: 158-161
13  Chabot V, de Keyzer Y, Gebhard S, Uske A, Bischof-Delaloye A, Rey F, Dusmet M, Gomez F. Ectopic ACTH Cushing's
      syndrome: V3 vasopressin receptor but not CRH receptor gene Expression in a pulmonary carcinoid tumor. Horm Res
      1998; 50: 226-231
14  Mao C, el Attar A, Domenico DR, Kim K, Howard JM. Carcinoid tumors of the pancreas. Status report based on two cases
      and review of the world's literature. Int J Pancreatol 1998; 23: 153-164
15  Duchesne G, Cassoni A, Pera M. Radiosensitivity related to neuroendocrine and endodermmal differentiation in lung
      carcinoma lines. Radiother Oncol 1988; 13: 153-161
16  Tezel E, Nagasaka T, Nomoto S, Sugimoto H, Nakao A. Neuroendocrine-like differentiation in patients with
      pancreatic carcinoma. Cancer 2000; 89: 2230-2236
17  Yao G, Zhou J, Zhao Z. Studies on the DNA content of breast carcinoma cells with neuroendocrine differentiation. Chin Med
      J 2002; 115: 296-298
18  Maluf HM, Zukerberg LR, Dickersin GR, Koerner FC. Spindle-cell argyrophilic mucin-producing carcinoma of the
      breast. Histological, ultrastructural, and immunohistochemical studies of two cases. Am J Surg Pathol 1991; 15: 677-686
19  Scopsi L, Andreola S, Pilotti S, Testori A, Baldini MT, Leoni F, Lombardi L, Hutton JC, Shimizu F, Rosa P. Argyrophilia
      and granin (chromogranin/secretogranin) expression in Female breast carcinomas. Their relationship to survival and
      other disease parameters. Am J Surg Pathol 1992; 16: 561-576
20  Grabowski P, Schindler I, Anagnostopoulos I, Foss HD, Riecken EO, Mansmann U, Stein H, Berger G, Buhr HJ, Scherubl
      H. Neuroendocrine differentiation is a relevant progostic factor in stage III-IV colorectal cancer. Eur J Gastroenterol
      Hepatol 2001; 13: 405-411
21  Sapino A, Papotti M, Righi L, Cassoni P, Chiusa L, Bussolati G. Clinical significance of neuroendocrine carcinoma of the
      breast. Ann Oncol 2001; 12(Suppl 2): S115-117
22  Matsui T, Kataoka M, Sugita Y, Itoh T, Ichihara T, Horisawa M, Koide A, Ichihara S, Nakao A. A case of small cell carcinoma
      of the stomach. Hepatogastroenterology 1997; 44: 156-160
23  Yang GC, Rotterdam H. Mixed (composite) glandular-endocrine cell carcinoma of the stomach. Report of a case and review
      of liteerature. Am J Surg Pathol 1991; 15: 592-598
24  Upp JR Jr, Olson D, Poston GJ, Alexander RW, Townsend CM Jr, Thompson JC. Inhibition of growth of two human
      pancreatic adenocarcinomas in vivo by somatostatin analog SMS 201-995. Am J Surg 1988; 155: 29-35
25  Anderson JV, Bloom SR. Neuroendocrine tumours of the gut: long-term therapy with the somatostatin analogue SMS
      201-995. Scand J Gastroenterol Suppl 1986; 119: 115-128
26  Oda Y, Tanaka Y, Naruse T, Sasanabe R, Tsubamoto M, Funahashi H. Expression of somatostatin receptor and effects
      of somatostatin analog on pancreatic endocrine tumors. Surg Today 2002; 32: 690-694
27  O'Byrne KJ, Carney DN. Somatostatin and the lung. Lung Cancer 1993; 10: 151-172
28  Taylor JE, Moreau JP, Baptiste L, Moody TW. Octapeptide analogues of somatostatin inhibit the clonal growth and
      vasoactive intestinal peptide-stimulated cyclic AMP formation in human small cell lung cancer cells. Peptides
      1991; 12: 839-843
29  Taylor JE, Bogden AE, Moreau JP, Coy DH. In vitro and in vivo inhibition of human small cell lung carcinoma (NCI-H69)
      growth by a somatostatin analogue. Biochem Biophys Res Commun 1988; 153: 81-86
30  Diaz-Sanchez CL, Molano Romero RA, Martinez Gonzalez M, Marquez Rivera ML, Halabe-Cherem J. Gastric
      neuroendocrine tumor. Rev Gastroenterol Mex 1998; 63: 97-100
31  Tomassetti P, Migliori M, Lalli S, Campana D, Tomassetti V, Corinaldesi R. Epidemiology, clinical features and diagnosis
      of gastroenteropancreatic endocrine tumours. Ann Oncol 2001; 12(Suppl 2): S95-99
32  Eriksson B, Oberg K, Stridsberg M. Tumor markers in neurodocrine tumors. Digestion 2000; 621: 33-38

Edited by Xu XQ