- Email: [email protected]
Expression of highly polysialylated neural cell adhesion molecule in pancreatic cancer neural invasive lesion
Cancer Letters 137 (1999) 201±207
Expression of highly polysialylated neural cell adhesion molecule in pancreatic cancer neural invasive lesion Kunio Kameda a, Hiroshi Shimada a, Takashi Ishikawa a, Atsushi Takimoto a, Nobuyoshi Momiyama a, Satoshi Hasegawa a, Kouichiro Misuta a, Akira Nakano a, Yoji Nagashima b, Yasushi Ichikawa a,* a
b
Second Department of Surgery, Yokohama City University School of Medicine, Yokohama, Japan Department of Pathology, Yokohama City University School of Medicine, 3-9 Fukuura Kanazawa-ku, Yokohama 236, Japan Received 28 September 1998; received in revised form 12 November 1998; accepted 12 November 1998
Abstract Neurotropism of pancreatic cancer is one of the hypotheses explaining neural invasion, which is one of the characteristics of pancreatic cancer. In these studies, we immunohistochemically examined neural cell adhesion molecules (NCAM), homophilic adhesion molecules expressed on the nerve cells, as a factor of neurotropism, in 15 pancreatic cancer operatively obtained, especially in neural invasive lesions. We also investigated the role of polysialic acid (PSA), which is attached to NCAM and related to the malignant potential of cancers. NCAM was detected in 66.7% of pancreatic cancers, and in all 9 cases with massive perineural invasion. In neural invasive lesions, however, there were perineurium and endoneurium, which do not express NCAM, between the cancer and nerve cells. PSA was also detected in the pancreatic cancers expressing NCAM. Moreover, PSA expression was stronger in the perineural invasive lesions than in the main tumor and was related to the cancer cell proliferation investigated by Ki-67 staining. It is unlikely therefore, that NCAM plays an important role in neurotropism. However, the NCAM expressed on the pancreatic cancer was attached to PSA, which itself plays an important role in the malignant potential of this disease. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pancreatic cancer; Perineural invasion; Neurotropism; Neural cell adhesion molecules; Polysialic acid
1. Introduction Patients with pancreatic cancer have a poor prognosis after tumor resection due to early-onset distant metastasis and local recurrence. One of the important causes of local recurrence of pancreatic cancer is known to be neural invasive lesions, which also makes it unlikely that the removal of the tumor will be curative [1]. The mechanism of neural invasion can * Corresponding author. Tel.: 1 81-45-787-2650; fax: 1 81-45782-9161; [email protected].
be partly explained by the anatomical proximity of the pancreas and celiac artery neural plexus [1]. On the other hand, Jentzer [2] has reported pancreatic cancer `neurotropism' as another possible explanation of the neural invasion of this cancer. Kenmotsu et al., with their examination of the expression of numerous types of neuro-endocrine markers in pancreatic cancer, successfully correlated the neural invasion with expression of neural cell adhesion molecules (NCAM), suggesting a possible role of NCAM in pancreatic cancer neurotropism [3]. Recently, it has been reported that tumor-associated
0304-3835/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(98)00359-0
202
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
Table 1 The patients a No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Age and sex
68 70 64 47 65 63 65 69 66 56 72 62 66 82 46
M M F M F F M F M M M F F M M
Location
Head Tail Head Head Head Head Head Tail Head Head Head Head Head Head Head
Histology
Pap. Well Mod. Mod. Mod. Mod. Mod. Mod Mod. Mod. Por. Por. Mod. Mod Mod.
Tumour size (cm)
2.5 2.4 3.0 3.3 4.0 4.5 2.5 4.0 3.5 3.0 5.5 3.3 3.0 3.0 2.8
£ £ £ £ £ £ £ £ £ £ £ £ £ £ £
2.4 2.1 2.6 2.5 4.0 3.5 2.5 3.3 3.0 2.7 4.0 3.0 2.5 2.5 2.5
£ £ £ £ £ £ £ £ £ £ £ £ £ £ £
2.6 1.2 2.0 2.4 3.5 4.3 1.8 1.9 2.0 2.4 3.7 2.3 2.0 2.0 2.1
rp
2 2 2 2 1 1 1 2 0 2 2 2 1 3 1
Neural invasion
NCAM
PSA
ne
plx
T
N
3 2 3 3 3 3 3 2 2 3 3 3 2 3 3
1 2 1 1 2 2 2 1 2 1 1 1 1 2 1
2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
T 2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
N 2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
a rp, histological retropancreatic invasion (0; not detected, 1; slightly, 2; massively, 3; invasion to other organs); ne, intrapancreatic neural invasion (2; moderately, 3; massively); plx, extrapancreatic neural invasion; T, main tumor; N, neural invasive lesion.
NCAM is attached by polysialic acid (PSA), which was expressed on embryonic NCAM but diminished in adult tissue. PSA is known to be always with NCAM in nature. PSA is reported to act as an oncodevelopmental antigen [4]. Its role is to mask and weaken the cell±cell interactions regulated by NCAM [17] and it may thus play an important role in cancer cell metastasis and invasion. In this paper, NCAM and PSA on pancreatic cancer were immunohistochemically investigated to determine whether they contribute to the characteristic neural invasion of this disease.
2. Materials and methods 2.1. Tissue specimens Specimens were obtained by operation from 15 patients with pancreatic cancer (Table 1). Specimens were chosen which contained a main tumor lesion, neural invasive lesion, and normal-appearing portion. After routine ®xation in 10% phosphate buffered formaldehyde, specimens were embedded in paraf®n.
2.2. Immunohistochemistry Paraf®n sections (5 mm) mounted on PLL coated glass slides (Superfrostw: Matsunami, Japan) were deparaf®nized, rehydrated and subsequently immersed in absolute methanol containing 0.3% H2O2 to block endogenous peroxidase activity. Following microwaving twice in 0.01 M citrate buffer (pH 6) at 908C for 5 min, the sections were incubated with normal goat or rabbit serum at 1:30 dilution for 30 min. Monoclonal antibody against human NCAM (diluted 1:50, Chemicon International Inc., Tomecula, CA), polyclonal antibody against human PSA (diluted 1:100, kindly provided by Dr. Seki [5,6], 2nd Department of Anatomy, Juntendou University School of Medicine), or anti-MIB-1 antibody [7] (Immunotech, Marseille, France) to detect Ki-67 was added at 48C overnight (Ki-67 is a useful indicator of cell proliferation). The sections were treated with biotinylated goat antimouse IgG (Dakopatts, Denmark) or with biotinylated rabbit anti-sheep IgG (Dakopatts, Denmark) for 30 min. The peroxidase labeled streptavidin (Dakopatts, Denmark) was then added to sections for 30 min at room temperature. Reaction products were developed by immersing the sections in 3,3 0 -diaminobenzidine tetrahydrochloride solution containing 0.1%
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
203
Fig. 1. NCAM staining of normal pancreas and pancreatic cancer. In normal pancreas, acinar cells or ductal epithelium did not show NCAM (A,B) and only islet cells expressed it (B). Pancreatic cancer showed NCAM clearly in its cytoplasm (C). D, ductal epithelium; Ac, acinar cells; I, islet; Ca, primary lesion of pancreatic carcinoma.
H2O2. Slides were couterstained lightly with hematoxylin.
P , 0:05 was considered to be statistically signi®cant.
2.3. Evaluation of immunostaining NCAM or PSA immunostaining was de®ned positive when more than 60% positive cells were detected under a light microscope at 50 £ magni®cation. Immunostaining of Ki-67 was quanti®ed by the ratio of positive cells per 1000 cancer cells as a labeling index. 2.4. Statistics Differences in the expression of NCAM or PSA versus Ki-67 were evaluated by Fisher's exact probability test. Correlation between factors was evaluated by rank test using Spearman's correlation coef®cient.
3. Results 3.1. NCAM expression in pancreatic cancer and normal pancreatic tissue In normal pancreatic tissue, NCAM was found only in islet cells, but not in ductal or acinar cells (Fig. 1). NCAM was also detected in peripheral nerve cells. NCAM expression appeared in the cytoplasm of cancer cells in 10 (66.7%) out of 15 pancreatic cancer cases. Neural invasion was observed in all 15 cases (Table 1). Six cases showed minimal invasion within the
204
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
Fig. 2. NCAM staining of neural invasive lesions. (A) Scheme of a cross section of a peripheral nerve. The outer layer of the nerve is the epineurium. Each of the smaller fascicles of the nerve are enclosed in concentric layers of connective tissue forming the perineurium. There is a ®ne membrane, the endoneurium, around each individual nerve ®ber. Neural invasion was de®ned as in®ltration to inside the perineurium. There are three spaces, that is, between the layers of the perineurium (*, perineural invasion), between the perineurium and endoneurium (**, intraperineural invasion), and inside the endoneurium (***, intraendoneural invasion). (B) Perineural invasion. The cancer-cell-forming gland invades perineural space, which was made by the piled perineurium. NCAM expression is detected in both the cancer cells and neural cells in the fascicle; however, it is not detected in the perineurium. (C) Intraperineural invasion: the cancer cells invaded to within the perineurium. There is a space between the cancer and neural cells (indicated by black triangle). Ca, cancer cells; F, fascicle.
intrapancreatic peripheral nerve (as `ne' in Table 1), while 9 cases showed excessive invasion involved in the celiac plexus, i.e. `extrapancreatic neural invasion' (as `plx' in Table 1). NCAM was found in 8 out of 9 cases with extrapancreatic neural invasion: plx (1), but in only 1 of 6 cases with plx (2) (see No. 14 in Table 1). The correlation between NCAM expression and degree of neural invasion was found to be signi®cant (P , 0:01). Generally speaking, extrapancreatic neural invasion (plx) is considered to be related with histological retroperitoneal invasion (rp) or tumor
size. In this study, 80% of NCAM (1) cancer showed massive retroperitoneal invasion (rp 2 or 3), and 80% of NCAM (2) cases were slight invasion (rp 0 or 1). There was statistical difference between the two groups in rp factor (P , 0:05). On the other hand, tumor size did not show signi®cant difference (NCAM (1), 3:3 ^ 0:87 cm, (2): 3:4 ^ 0:87 cm, mean ^ SD). We further examined the neural pathway of the cancer invasion in 10 cases expressing NCAM. One hundred sites of neural invasion, randomly selected from 10 neural invasive lesions of each case, were
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
205
microscopically investigated. Neural spaces are made by three membranes including endoneurium, perineurium and epineurium (Fig. 2). Therefore, three pathways around the nerve are: (a) between the layers of the perineurium, perineural invasion; (b) between the perineurium and endoneurium, intraperineural invasion; (c) inside the endoneurium, intraendoneural invasion. In our investigation, 72 lesions showed intraperineural invasions, with the other 28 showing perineural invasions (Fig. 2). Intraendoneural invasive lesions were not indicated. It is interesting to note that, in neural invasive lesions, there are perineurium and endoneurium, neither of which express NCAM, between neural cells and pancreatic cancer cells, therefore NCAM is not available as an adhesion molecule between neural cells and pancreatic cancer cells. 3.2. PSA expression in normal pancreas, main tumors, and neural invasive lesions PSA was not detected in normal islet cells, which expressed NCAM. On the other hand, PSA was found in all NCAM-expressing cancerous lesions. Finally, PSA was stronger in cancer cells of neural invasive lesions than it was in the main tumors (Fig. 3). 3.3. Ki-67 staining in normal pancreas, main tumors, and neural invasive lesions The labeling index (LI) of Ki-67 staining was signi®cantly higher in the main tumors (16:4 ^ 2:2 mean ^ standard error) than in normal pancreas (2:1 ^ 0:4). Moreover, LI was signi®cantly higher in perineural invasive lesions (42:5 ^ 5:5) than in main tumors (Fig. 4). When perineural invasive lesions were examined, LI was signi®cantly higher in the PSA positive (51:6 ^ 6:6) than in the PSA negative (20:4 ^ 4:2) tumors. 4. Discussion Perineural space is known to be an important route of pancreatic cancer invasion. Pancreatic cancer cells invade the neural plexus distributed in pancreatic parenchyma and spread through perineural space to reach extrapancreatic nerves, such as the celiac plexus or the supra mesenteric artery plexus. Cancer cell invasion into the neural plexus often results in non-
Fig. 3. PSA staining. (A) PSA expression of neural invasive lesion. PSA is presented in the cytoplasm of the perineural invasive cancer cells, but not in the nerve cells in the fascicle. Ca, cancer cells; F, fascicle. (B) PSA expression of main tumor. PSA is also detected in the main tumor, but to a signi®cantly lesser degree than in the perineural invasive lesion.
curative resection and, therefore, retroperitoneal recurrence [11]. Perineural invasion is thus explained as a type of lymphogenic metastasis since there are lymph vessels in the perineural spaces [12]. There is another theory, however, to explain the neural invasion of pancreatic cancer. The concept of pancreatic cancer `neurotropism' was developed by Kenmotsu et al., who reported that advanced pancreatic cancer with neural invasion expressed numerous types of neuroendocrine markers, including S-100, Synaptophysin, Substance-P, Enkephalin, and NCAM. These authors implicated neurotropism as a possible cause of the frequent neural invasion in pancreatic cancer, since NCAM expression was observed in both peripheral nerve and pancreatic cancer [3]. The peripheral nerve system is covered by three membranes: the endoneurium, perineurium and
206
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
Fig. 4. Labeling index (LI) of Ki-67 staining. Immunostaining of Ki-67 was performed utilizing MIB-1 antibody and was quanti®ed by the ratio of positive cells per 1000 cancer cells as a labeling index.
epineurium (Fig. 2). The route of cancer cell invasion may thus involve any or all of these. Theoretically, the neurotropism of pancreatic cancer by NCAM succeeds only when cancer cells invade to inside the endoneurium by breaking through all three membranes. In the present study, NCAM expression was found in 66.7% of pancreatic cancer cases. The expression was detected more frequently in lesions where neural invasion was excessive. A closer observation of cancer cell invasion, however, reveals that there are perineurium and endoneurium, which do not express NCAM, between the invasive cancer cells and neural cells. This ®nding leads us to the conclusion that NCAM does not directly contribute to neurotropism of pancreatic cancer cells. Moreover, NCAM expression was correlated not only with extrapancreatic neural invasion but also retroperitoneal invasion, which is one of the important factors of pancreatic cancer progression. The case of No. 14, which showed NCAM expression but did not have extrapancreatic neural invasion, showed massive retroperitoneal invasion. In this study, we did not have a case of early stage pancreatic cancer. However, from Kenmotsu's report saying that none of the three
cases of stage I pancreatic cancer expressed NCAM [3], we can guess that such early stage pancreatic cancer may not express NCAM and NCAM expression is according to pancreatic cancer progression. So, NCAM expression of pancreatic cancer with extrapancreatic neural invasion may be a result of cancer progression. The overexpression of homophilic adhesion molecules, such as cadherin and NCAM, is believed to inhibit invasion and metastasis of cancer cells, since increased intercellular adhesion prevents the detachment of cancer cells from the tumor. In reality, the invasiveness of gastric and esophageal cancer is related to the loss of cadherin or cathenin expression [8,9]. NCAM expression is reported to be stronger in the non-metastatic melanoma cell line than in the metastatic one [10]. The DCC gene, which is a homologue of NCAM, is also reported to be reduced in the metastatic phenotype of colon cancer [13]. Thus far, however, NCAM has been also reported to be expressed on poor prognostic carcinomas. All small-cell lung carcinomas are positive for NCAM. NCAM-positive non-small cell lung carcinomas (NSCLC), which account for approximately 20% of all NSCLC, showed a signi®cantly shorter postoperative overall and disease-free survival time than NCAM-negative NSCLC [14]. For lung cancer patients, NCAM is becoming an important predictor for prognosis [15]. Recently, tumor-associated NCAM was reported to be attached by polysialic acid (PSA), which is expressed on embryonic NCAM but is diminished in adult tissue. PSA is known to be always with NCAM in nature. Roth et al. identi®ed re-expression of PSA (which is suppressed in normal adult kidneys), in Wilm's tumor and thus considered it an oncodevelopmental antigen [4]. NCAM expressed on small cell lung carcinomas also combined with PSA [16]. The most important role of PSA, however, is the masking and weakening of cell±cell interactions regulated by NCAM [17]. Therefore, PSA weakens NCAM's capability for adhesion and suppression of tumor metastasis, and increases cancer cell detachment from the tumor. In this manner, PSA may play an important role in cancer cell metastasis and invasion. PSA is also known to modulate the body's immune response to tumor-speci®c cell antigens through its masking effect
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
[18]. Therefore, cancer cells expressing PSA may be able to escape from the immuno-system and successfully metastasize to other organs. Moreover, PSA is related to neural cell development and proliferation [5]. In our study, the Ki-67 index was signi®cantly higher in NCAM- and PSApositive lesions than in negative lesions. The index was particularly high in neural invasive lesions where PSA was upregulated (Fig. 4). Therefore, PSA may also be associated with the malignant potential of pancreatic cancer through the mechanism of cell proliferation. These results clari®ed that the presence of NCAM cannot suf®ciently explain the mechanism of neural invasion in pancreatic cancer, but that PSA, which is combined with NCAM, may play a role in such invasion. Acknowledgements
[6] [7]
[8] [9]
[10]
[11] [12]
We specially thank Dr. Seki who kindly provided PSA antibody and gave us important advice.
[13]
References
[14]
[1] M. Matsuda, Y. Nimura, Perineural invasion of pancreas head carcinoma, Jpn. J. Surg. 84 (1983) 719±728. [2] A. Jentzer, Neurotropisme des cellules cancereuses clinique et therapeutique des cancers neurotropes, Am. J. Cancer 16 (1932) 37. [3] M. Kenmotsu, Relationship between peritoneal invasion and expression of neuro-endocrine markers and neural cell adhesion molecule in human pancreatic carcinoma, Nihon Shokaki Geka Gakkai Zasshi 23 (1990) 2580±2585. [4] J. Roth, C. Zumer, P. Wagner, I. Blaha, D. Bitter-Suermann, P.U. Heitz, Presence of long chain form of polysialic acid of neural cell adhesion molecule in Wilm's tumor ± identi®cation of a cell adhesion molecules as an oncodevelopmental antigen and implications for tumor histogenesis, Am. J. Pathol. 133 (1988) 227±240. [5] T. Seki, Y. Arai, Distribution possible roles of the highly
[15] [16]
[17] [18]
207
polysialylated neural cell adhesion molecule (NCAM-H) in the developing and adult central nervous system, Neurosci. Res. 17 (1993) 265±290. T. Seki, Y. Arai, Age-related production of new granule cells in the adult dentate gyrus, NeuroReport 6 (1995) 2479± 2482. J. Gerdes, M.H.G. Becker, G. Key, G. Cattoretti, Immunohistological detection of tumor growth fraction (Ki-67 antigen) in formalin-®xed routinely processed tissue, J. Pathol. 168 (1992) 85±86. Y. Shimoyama, S. Hirohashi, Expression E- and P-cadherin in gastric cancer, Cancer Res. 51 (1991) 2185±2192. H. Shiozaki, K. Iihara, H. Oka, T. Kadowaki, S. Matsui, J. Gofuke, M. Inoue, A. Nagafuchi, S. Tsukita, T. Mori, Immunohistochemical detection of a-catenin expression in human cancers, Am. J. Pathol. 144 (1994) 667±674. D. Linnemann, A. Raz, E. Bock, Differential expression of cell adhesion molecules in variants of K1735 melanoma cells differing in metastatic capacity, Int. J. Cancer 43 (1989) 709±712. H. Nagai, A. Kuroda, Y. Morioka, Lymphatic and local spreading of T1 and T2 pancreatic cancer, Ann. Surg. 204 (1986) 65±71. T. Kihara, Das extravaskulaÈre Saftbahnensystem, Okajima's Folia Anat. Jpn. 28 (1956) 601±621. E.R. Faeron, W.E. Pierceall, The deleated in colorectal cancer (DCC) gene: a candidate tumor suppressor gene encoding a cell surface protein with similarity to neural cell adhesion molecules, Cancer Surv. 24 (1995) 3±17. R.E. Kibbelaal, C.E.C. Moolenaar, R.J.A.M. Michalides, P.C. Van Bodegom, D. Van der Scueren, O. Dalesio, N. Van Zandwijk, W.J. Mooi, Neural-cell-adhesion-molecule expression, neuro-endocrine differentiation and prognosis in lung carcinoma, Eur. J. Cancer 27 (1991) 431±435. J. Niklinsky, M. Furman, Clinical tumour marker in lung cancer, Eur. J. Cancer Prev. 4 (1995) 129±138. P. Komminoth, J. Roth, P.M. Lackie, D. Bitter-Suermann, P.U. Heitz, Polysialic acid of the neural cell adhesion molecule distinguishes small cell lung carcinoma from carcinoids, Am. J. Pathol. 139 (1991) 297±304. P. Yang, H. Yin, U. Rutishauser, Intercellular space is affected by polysialic acid content of NCAM. , J. Cell Biol. 116 (1992) 1487±1496. D. Bitter-Suermann, In¯uence bacterial polysialic capsules on host defence: masquerade and mimicry, in: J. Roth, U. Rutishauser, F. Troy II (Eds.), Polysialic Acid: From Microbes to Man, Birkhauser, Basel, 1993, pp. 11.
Expression of highly polysialylated neural cell adhesion molecule in pancreatic cancer neural invasive lesion Kunio Kameda a, Hiroshi Shimada a, Takashi Ishikawa a, Atsushi Takimoto a, Nobuyoshi Momiyama a, Satoshi Hasegawa a, Kouichiro Misuta a, Akira Nakano a, Yoji Nagashima b, Yasushi Ichikawa a,* a
b
Second Department of Surgery, Yokohama City University School of Medicine, Yokohama, Japan Department of Pathology, Yokohama City University School of Medicine, 3-9 Fukuura Kanazawa-ku, Yokohama 236, Japan Received 28 September 1998; received in revised form 12 November 1998; accepted 12 November 1998
Abstract Neurotropism of pancreatic cancer is one of the hypotheses explaining neural invasion, which is one of the characteristics of pancreatic cancer. In these studies, we immunohistochemically examined neural cell adhesion molecules (NCAM), homophilic adhesion molecules expressed on the nerve cells, as a factor of neurotropism, in 15 pancreatic cancer operatively obtained, especially in neural invasive lesions. We also investigated the role of polysialic acid (PSA), which is attached to NCAM and related to the malignant potential of cancers. NCAM was detected in 66.7% of pancreatic cancers, and in all 9 cases with massive perineural invasion. In neural invasive lesions, however, there were perineurium and endoneurium, which do not express NCAM, between the cancer and nerve cells. PSA was also detected in the pancreatic cancers expressing NCAM. Moreover, PSA expression was stronger in the perineural invasive lesions than in the main tumor and was related to the cancer cell proliferation investigated by Ki-67 staining. It is unlikely therefore, that NCAM plays an important role in neurotropism. However, the NCAM expressed on the pancreatic cancer was attached to PSA, which itself plays an important role in the malignant potential of this disease. q 1999 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pancreatic cancer; Perineural invasion; Neurotropism; Neural cell adhesion molecules; Polysialic acid
1. Introduction Patients with pancreatic cancer have a poor prognosis after tumor resection due to early-onset distant metastasis and local recurrence. One of the important causes of local recurrence of pancreatic cancer is known to be neural invasive lesions, which also makes it unlikely that the removal of the tumor will be curative [1]. The mechanism of neural invasion can * Corresponding author. Tel.: 1 81-45-787-2650; fax: 1 81-45782-9161; [email protected].
be partly explained by the anatomical proximity of the pancreas and celiac artery neural plexus [1]. On the other hand, Jentzer [2] has reported pancreatic cancer `neurotropism' as another possible explanation of the neural invasion of this cancer. Kenmotsu et al., with their examination of the expression of numerous types of neuro-endocrine markers in pancreatic cancer, successfully correlated the neural invasion with expression of neural cell adhesion molecules (NCAM), suggesting a possible role of NCAM in pancreatic cancer neurotropism [3]. Recently, it has been reported that tumor-associated
0304-3835/99/$ - see front matter q 1999 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(98)00359-0
202
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
Table 1 The patients a No.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Age and sex
68 70 64 47 65 63 65 69 66 56 72 62 66 82 46
M M F M F F M F M M M F F M M
Location
Head Tail Head Head Head Head Head Tail Head Head Head Head Head Head Head
Histology
Pap. Well Mod. Mod. Mod. Mod. Mod. Mod Mod. Mod. Por. Por. Mod. Mod Mod.
Tumour size (cm)
2.5 2.4 3.0 3.3 4.0 4.5 2.5 4.0 3.5 3.0 5.5 3.3 3.0 3.0 2.8
£ £ £ £ £ £ £ £ £ £ £ £ £ £ £
2.4 2.1 2.6 2.5 4.0 3.5 2.5 3.3 3.0 2.7 4.0 3.0 2.5 2.5 2.5
£ £ £ £ £ £ £ £ £ £ £ £ £ £ £
2.6 1.2 2.0 2.4 3.5 4.3 1.8 1.9 2.0 2.4 3.7 2.3 2.0 2.0 2.1
rp
2 2 2 2 1 1 1 2 0 2 2 2 1 3 1
Neural invasion
NCAM
PSA
ne
plx
T
N
3 2 3 3 3 3 3 2 2 3 3 3 2 3 3
1 2 1 1 2 2 2 1 2 1 1 1 1 2 1
2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
T 2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
N 2 1 1 1 2 2 2 1 2 1 1 1 1 1 1
a rp, histological retropancreatic invasion (0; not detected, 1; slightly, 2; massively, 3; invasion to other organs); ne, intrapancreatic neural invasion (2; moderately, 3; massively); plx, extrapancreatic neural invasion; T, main tumor; N, neural invasive lesion.
NCAM is attached by polysialic acid (PSA), which was expressed on embryonic NCAM but diminished in adult tissue. PSA is known to be always with NCAM in nature. PSA is reported to act as an oncodevelopmental antigen [4]. Its role is to mask and weaken the cell±cell interactions regulated by NCAM [17] and it may thus play an important role in cancer cell metastasis and invasion. In this paper, NCAM and PSA on pancreatic cancer were immunohistochemically investigated to determine whether they contribute to the characteristic neural invasion of this disease.
2. Materials and methods 2.1. Tissue specimens Specimens were obtained by operation from 15 patients with pancreatic cancer (Table 1). Specimens were chosen which contained a main tumor lesion, neural invasive lesion, and normal-appearing portion. After routine ®xation in 10% phosphate buffered formaldehyde, specimens were embedded in paraf®n.
2.2. Immunohistochemistry Paraf®n sections (5 mm) mounted on PLL coated glass slides (Superfrostw: Matsunami, Japan) were deparaf®nized, rehydrated and subsequently immersed in absolute methanol containing 0.3% H2O2 to block endogenous peroxidase activity. Following microwaving twice in 0.01 M citrate buffer (pH 6) at 908C for 5 min, the sections were incubated with normal goat or rabbit serum at 1:30 dilution for 30 min. Monoclonal antibody against human NCAM (diluted 1:50, Chemicon International Inc., Tomecula, CA), polyclonal antibody against human PSA (diluted 1:100, kindly provided by Dr. Seki [5,6], 2nd Department of Anatomy, Juntendou University School of Medicine), or anti-MIB-1 antibody [7] (Immunotech, Marseille, France) to detect Ki-67 was added at 48C overnight (Ki-67 is a useful indicator of cell proliferation). The sections were treated with biotinylated goat antimouse IgG (Dakopatts, Denmark) or with biotinylated rabbit anti-sheep IgG (Dakopatts, Denmark) for 30 min. The peroxidase labeled streptavidin (Dakopatts, Denmark) was then added to sections for 30 min at room temperature. Reaction products were developed by immersing the sections in 3,3 0 -diaminobenzidine tetrahydrochloride solution containing 0.1%
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
203
Fig. 1. NCAM staining of normal pancreas and pancreatic cancer. In normal pancreas, acinar cells or ductal epithelium did not show NCAM (A,B) and only islet cells expressed it (B). Pancreatic cancer showed NCAM clearly in its cytoplasm (C). D, ductal epithelium; Ac, acinar cells; I, islet; Ca, primary lesion of pancreatic carcinoma.
H2O2. Slides were couterstained lightly with hematoxylin.
P , 0:05 was considered to be statistically signi®cant.
2.3. Evaluation of immunostaining NCAM or PSA immunostaining was de®ned positive when more than 60% positive cells were detected under a light microscope at 50 £ magni®cation. Immunostaining of Ki-67 was quanti®ed by the ratio of positive cells per 1000 cancer cells as a labeling index. 2.4. Statistics Differences in the expression of NCAM or PSA versus Ki-67 were evaluated by Fisher's exact probability test. Correlation between factors was evaluated by rank test using Spearman's correlation coef®cient.
3. Results 3.1. NCAM expression in pancreatic cancer and normal pancreatic tissue In normal pancreatic tissue, NCAM was found only in islet cells, but not in ductal or acinar cells (Fig. 1). NCAM was also detected in peripheral nerve cells. NCAM expression appeared in the cytoplasm of cancer cells in 10 (66.7%) out of 15 pancreatic cancer cases. Neural invasion was observed in all 15 cases (Table 1). Six cases showed minimal invasion within the
204
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
Fig. 2. NCAM staining of neural invasive lesions. (A) Scheme of a cross section of a peripheral nerve. The outer layer of the nerve is the epineurium. Each of the smaller fascicles of the nerve are enclosed in concentric layers of connective tissue forming the perineurium. There is a ®ne membrane, the endoneurium, around each individual nerve ®ber. Neural invasion was de®ned as in®ltration to inside the perineurium. There are three spaces, that is, between the layers of the perineurium (*, perineural invasion), between the perineurium and endoneurium (**, intraperineural invasion), and inside the endoneurium (***, intraendoneural invasion). (B) Perineural invasion. The cancer-cell-forming gland invades perineural space, which was made by the piled perineurium. NCAM expression is detected in both the cancer cells and neural cells in the fascicle; however, it is not detected in the perineurium. (C) Intraperineural invasion: the cancer cells invaded to within the perineurium. There is a space between the cancer and neural cells (indicated by black triangle). Ca, cancer cells; F, fascicle.
intrapancreatic peripheral nerve (as `ne' in Table 1), while 9 cases showed excessive invasion involved in the celiac plexus, i.e. `extrapancreatic neural invasion' (as `plx' in Table 1). NCAM was found in 8 out of 9 cases with extrapancreatic neural invasion: plx (1), but in only 1 of 6 cases with plx (2) (see No. 14 in Table 1). The correlation between NCAM expression and degree of neural invasion was found to be signi®cant (P , 0:01). Generally speaking, extrapancreatic neural invasion (plx) is considered to be related with histological retroperitoneal invasion (rp) or tumor
size. In this study, 80% of NCAM (1) cancer showed massive retroperitoneal invasion (rp 2 or 3), and 80% of NCAM (2) cases were slight invasion (rp 0 or 1). There was statistical difference between the two groups in rp factor (P , 0:05). On the other hand, tumor size did not show signi®cant difference (NCAM (1), 3:3 ^ 0:87 cm, (2): 3:4 ^ 0:87 cm, mean ^ SD). We further examined the neural pathway of the cancer invasion in 10 cases expressing NCAM. One hundred sites of neural invasion, randomly selected from 10 neural invasive lesions of each case, were
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
205
microscopically investigated. Neural spaces are made by three membranes including endoneurium, perineurium and epineurium (Fig. 2). Therefore, three pathways around the nerve are: (a) between the layers of the perineurium, perineural invasion; (b) between the perineurium and endoneurium, intraperineural invasion; (c) inside the endoneurium, intraendoneural invasion. In our investigation, 72 lesions showed intraperineural invasions, with the other 28 showing perineural invasions (Fig. 2). Intraendoneural invasive lesions were not indicated. It is interesting to note that, in neural invasive lesions, there are perineurium and endoneurium, neither of which express NCAM, between neural cells and pancreatic cancer cells, therefore NCAM is not available as an adhesion molecule between neural cells and pancreatic cancer cells. 3.2. PSA expression in normal pancreas, main tumors, and neural invasive lesions PSA was not detected in normal islet cells, which expressed NCAM. On the other hand, PSA was found in all NCAM-expressing cancerous lesions. Finally, PSA was stronger in cancer cells of neural invasive lesions than it was in the main tumors (Fig. 3). 3.3. Ki-67 staining in normal pancreas, main tumors, and neural invasive lesions The labeling index (LI) of Ki-67 staining was signi®cantly higher in the main tumors (16:4 ^ 2:2 mean ^ standard error) than in normal pancreas (2:1 ^ 0:4). Moreover, LI was signi®cantly higher in perineural invasive lesions (42:5 ^ 5:5) than in main tumors (Fig. 4). When perineural invasive lesions were examined, LI was signi®cantly higher in the PSA positive (51:6 ^ 6:6) than in the PSA negative (20:4 ^ 4:2) tumors. 4. Discussion Perineural space is known to be an important route of pancreatic cancer invasion. Pancreatic cancer cells invade the neural plexus distributed in pancreatic parenchyma and spread through perineural space to reach extrapancreatic nerves, such as the celiac plexus or the supra mesenteric artery plexus. Cancer cell invasion into the neural plexus often results in non-
Fig. 3. PSA staining. (A) PSA expression of neural invasive lesion. PSA is presented in the cytoplasm of the perineural invasive cancer cells, but not in the nerve cells in the fascicle. Ca, cancer cells; F, fascicle. (B) PSA expression of main tumor. PSA is also detected in the main tumor, but to a signi®cantly lesser degree than in the perineural invasive lesion.
curative resection and, therefore, retroperitoneal recurrence [11]. Perineural invasion is thus explained as a type of lymphogenic metastasis since there are lymph vessels in the perineural spaces [12]. There is another theory, however, to explain the neural invasion of pancreatic cancer. The concept of pancreatic cancer `neurotropism' was developed by Kenmotsu et al., who reported that advanced pancreatic cancer with neural invasion expressed numerous types of neuroendocrine markers, including S-100, Synaptophysin, Substance-P, Enkephalin, and NCAM. These authors implicated neurotropism as a possible cause of the frequent neural invasion in pancreatic cancer, since NCAM expression was observed in both peripheral nerve and pancreatic cancer [3]. The peripheral nerve system is covered by three membranes: the endoneurium, perineurium and
206
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
Fig. 4. Labeling index (LI) of Ki-67 staining. Immunostaining of Ki-67 was performed utilizing MIB-1 antibody and was quanti®ed by the ratio of positive cells per 1000 cancer cells as a labeling index.
epineurium (Fig. 2). The route of cancer cell invasion may thus involve any or all of these. Theoretically, the neurotropism of pancreatic cancer by NCAM succeeds only when cancer cells invade to inside the endoneurium by breaking through all three membranes. In the present study, NCAM expression was found in 66.7% of pancreatic cancer cases. The expression was detected more frequently in lesions where neural invasion was excessive. A closer observation of cancer cell invasion, however, reveals that there are perineurium and endoneurium, which do not express NCAM, between the invasive cancer cells and neural cells. This ®nding leads us to the conclusion that NCAM does not directly contribute to neurotropism of pancreatic cancer cells. Moreover, NCAM expression was correlated not only with extrapancreatic neural invasion but also retroperitoneal invasion, which is one of the important factors of pancreatic cancer progression. The case of No. 14, which showed NCAM expression but did not have extrapancreatic neural invasion, showed massive retroperitoneal invasion. In this study, we did not have a case of early stage pancreatic cancer. However, from Kenmotsu's report saying that none of the three
cases of stage I pancreatic cancer expressed NCAM [3], we can guess that such early stage pancreatic cancer may not express NCAM and NCAM expression is according to pancreatic cancer progression. So, NCAM expression of pancreatic cancer with extrapancreatic neural invasion may be a result of cancer progression. The overexpression of homophilic adhesion molecules, such as cadherin and NCAM, is believed to inhibit invasion and metastasis of cancer cells, since increased intercellular adhesion prevents the detachment of cancer cells from the tumor. In reality, the invasiveness of gastric and esophageal cancer is related to the loss of cadherin or cathenin expression [8,9]. NCAM expression is reported to be stronger in the non-metastatic melanoma cell line than in the metastatic one [10]. The DCC gene, which is a homologue of NCAM, is also reported to be reduced in the metastatic phenotype of colon cancer [13]. Thus far, however, NCAM has been also reported to be expressed on poor prognostic carcinomas. All small-cell lung carcinomas are positive for NCAM. NCAM-positive non-small cell lung carcinomas (NSCLC), which account for approximately 20% of all NSCLC, showed a signi®cantly shorter postoperative overall and disease-free survival time than NCAM-negative NSCLC [14]. For lung cancer patients, NCAM is becoming an important predictor for prognosis [15]. Recently, tumor-associated NCAM was reported to be attached by polysialic acid (PSA), which is expressed on embryonic NCAM but is diminished in adult tissue. PSA is known to be always with NCAM in nature. Roth et al. identi®ed re-expression of PSA (which is suppressed in normal adult kidneys), in Wilm's tumor and thus considered it an oncodevelopmental antigen [4]. NCAM expressed on small cell lung carcinomas also combined with PSA [16]. The most important role of PSA, however, is the masking and weakening of cell±cell interactions regulated by NCAM [17]. Therefore, PSA weakens NCAM's capability for adhesion and suppression of tumor metastasis, and increases cancer cell detachment from the tumor. In this manner, PSA may play an important role in cancer cell metastasis and invasion. PSA is also known to modulate the body's immune response to tumor-speci®c cell antigens through its masking effect
K. Kameda et al. / Cancer Letters 137 (1998) 201±207
[18]. Therefore, cancer cells expressing PSA may be able to escape from the immuno-system and successfully metastasize to other organs. Moreover, PSA is related to neural cell development and proliferation [5]. In our study, the Ki-67 index was signi®cantly higher in NCAM- and PSApositive lesions than in negative lesions. The index was particularly high in neural invasive lesions where PSA was upregulated (Fig. 4). Therefore, PSA may also be associated with the malignant potential of pancreatic cancer through the mechanism of cell proliferation. These results clari®ed that the presence of NCAM cannot suf®ciently explain the mechanism of neural invasion in pancreatic cancer, but that PSA, which is combined with NCAM, may play a role in such invasion. Acknowledgements
[6] [7]
[8] [9]
[10]
[11] [12]
We specially thank Dr. Seki who kindly provided PSA antibody and gave us important advice.
[13]
References
[14]
[1] M. Matsuda, Y. Nimura, Perineural invasion of pancreas head carcinoma, Jpn. J. Surg. 84 (1983) 719±728. [2] A. Jentzer, Neurotropisme des cellules cancereuses clinique et therapeutique des cancers neurotropes, Am. J. Cancer 16 (1932) 37. [3] M. Kenmotsu, Relationship between peritoneal invasion and expression of neuro-endocrine markers and neural cell adhesion molecule in human pancreatic carcinoma, Nihon Shokaki Geka Gakkai Zasshi 23 (1990) 2580±2585. [4] J. Roth, C. Zumer, P. Wagner, I. Blaha, D. Bitter-Suermann, P.U. Heitz, Presence of long chain form of polysialic acid of neural cell adhesion molecule in Wilm's tumor ± identi®cation of a cell adhesion molecules as an oncodevelopmental antigen and implications for tumor histogenesis, Am. J. Pathol. 133 (1988) 227±240. [5] T. Seki, Y. Arai, Distribution possible roles of the highly
[15] [16]
[17] [18]
207
polysialylated neural cell adhesion molecule (NCAM-H) in the developing and adult central nervous system, Neurosci. Res. 17 (1993) 265±290. T. Seki, Y. Arai, Age-related production of new granule cells in the adult dentate gyrus, NeuroReport 6 (1995) 2479± 2482. J. Gerdes, M.H.G. Becker, G. Key, G. Cattoretti, Immunohistological detection of tumor growth fraction (Ki-67 antigen) in formalin-®xed routinely processed tissue, J. Pathol. 168 (1992) 85±86. Y. Shimoyama, S. Hirohashi, Expression E- and P-cadherin in gastric cancer, Cancer Res. 51 (1991) 2185±2192. H. Shiozaki, K. Iihara, H. Oka, T. Kadowaki, S. Matsui, J. Gofuke, M. Inoue, A. Nagafuchi, S. Tsukita, T. Mori, Immunohistochemical detection of a-catenin expression in human cancers, Am. J. Pathol. 144 (1994) 667±674. D. Linnemann, A. Raz, E. Bock, Differential expression of cell adhesion molecules in variants of K1735 melanoma cells differing in metastatic capacity, Int. J. Cancer 43 (1989) 709±712. H. Nagai, A. Kuroda, Y. Morioka, Lymphatic and local spreading of T1 and T2 pancreatic cancer, Ann. Surg. 204 (1986) 65±71. T. Kihara, Das extravaskulaÈre Saftbahnensystem, Okajima's Folia Anat. Jpn. 28 (1956) 601±621. E.R. Faeron, W.E. Pierceall, The deleated in colorectal cancer (DCC) gene: a candidate tumor suppressor gene encoding a cell surface protein with similarity to neural cell adhesion molecules, Cancer Surv. 24 (1995) 3±17. R.E. Kibbelaal, C.E.C. Moolenaar, R.J.A.M. Michalides, P.C. Van Bodegom, D. Van der Scueren, O. Dalesio, N. Van Zandwijk, W.J. Mooi, Neural-cell-adhesion-molecule expression, neuro-endocrine differentiation and prognosis in lung carcinoma, Eur. J. Cancer 27 (1991) 431±435. J. Niklinsky, M. Furman, Clinical tumour marker in lung cancer, Eur. J. Cancer Prev. 4 (1995) 129±138. P. Komminoth, J. Roth, P.M. Lackie, D. Bitter-Suermann, P.U. Heitz, Polysialic acid of the neural cell adhesion molecule distinguishes small cell lung carcinoma from carcinoids, Am. J. Pathol. 139 (1991) 297±304. P. Yang, H. Yin, U. Rutishauser, Intercellular space is affected by polysialic acid content of NCAM. , J. Cell Biol. 116 (1992) 1487±1496. D. Bitter-Suermann, In¯uence bacterial polysialic capsules on host defence: masquerade and mimicry, in: J. Roth, U. Rutishauser, F. Troy II (Eds.), Polysialic Acid: From Microbes to Man, Birkhauser, Basel, 1993, pp. 11.