Clinical Research
N-cadherin Expression in Adrenal Tumors: Upregulation in Malignant Pheochromocytoma and Downregulation in Adrenocortical Carcinoma
Amir Khorram-Manesh, MD, Håkan Ahlman, MD, PHD, Svante Jansson, MD, PHD, and Ola Nilsson, MD, PHD
Abstract
Cell adhesion molecules (CAMs) are important regulators of tumor growth. The aim of the present study was to evaluate the expression pattern of CAMs in adrenal tumors regarding origin (cortex vs medulla) and biologic behavior (benign vs malignant). Eighty- seven adrenal tumors were investigated by immunocytochemistry (ICC) using monoclonal antibodies against N-cadherin (NCAD), E-cadherin (ECAD), neural cell adhesion molecule (NCAM), and CD44. Western blotting was performed on 30 tumors using the same anti- bodies. Markers for proliferation (Ki-67) and catecholamine synthesis (tyrosine hydroxy- lase) were also analyzed in tumors by ICC. NCAD was expressed in 12/27 benign pheochromocytomas (BPCs) (12 familial cases), 8/8 malignant pheochromocytomas (MPCs), 28/30 adrenocortical adenomas, and 9/22 adrenocortical carcinomas. ECAD was expressed in 0/27 BPCs, 0/8 MPCs, 0/30 adrenocortical adenomas, and 2/22 adrenocortical carcino- mas. NCAM was expressed in 26/27 BPCs, 7/8 MPCs, 21/30 adrenocortical adenomas, and 17/22 adrenocortical carcinomas. CD44 was expressed in 23/27 BPCs, 6/8 MPCs, 7/30 adreno- cortical adenomas, and 4/22 adrenocortical carcinomas. Both cortical and medullary adrenal tumors expressed NCAD, NCAM, and CD44 but were devoid of ECAD. The expres- sion of CD44 and NCAM did not correlate with the malignant potential of tumors. NCAD was upregulated in MPCs, but downregulated in adrenocortical carcinoma. Thus, NCAD appears to be involved in the development of both cortical and medullary adrenal tumors. Key Words: Adhesion molecules; adrenal tumors; N-cadherin; E-cadherin; pheochromo- cytomas.
Lundberg Laboratory for Cancer Research, Departments of Surgery and Pathology, Sahlgrenska University Hospital, Göteborg, Sweden.
Address correspondence to Dr. Amir Khorram-Manesh, Department of Surgery, Sahlgrenska University Hospital, S-413 45 Göteborg, Sweden. E-mail: amir.khorram- manesh@surgery.gu.se
Endocrine Pathology, vol. 13, no. 2, 99-110, Summer 2002 @ Copyright 2002 by Humana Press Inc. All rights of any nature whatsoever reserved. 1046-3976/02/13:99-110/ $13.00
Introduction
Specific surface proteins, cell adhesion molecules (CAMs), mediate cell adhesion. There are CAMs that require Ca+2 for adhesion (cadherins) and those that do not (e.g., neural cell adhesion molecules [NCAM]). Cadherins (E-, P-, and N) are important for tissue differentiation and structure. During differentiation, the expression of cadherins changes, which affects many aspects of cell-cell adhesion and cell migration. E-cadherin (ECAD)
holds most epithelial cells together. As the ectodermal cells invaginate, they lose their expression of ECAD and start to synthe- size N-cadherin (NCAD). In the differ- entiation of the central nervous system from the neural tube, NCAD gradually becomes the major adhesion molecule. As the neural crest cells begin to migrate, they lose their expression of all cadherins and are thus readily distinguished from the overlying ectoderm (ECAD) and the neu- ral tube (NCAD). Once the neural crest
cells reach their target organs, they differ- entiate into neurons and will again express NCAD. In the adult organism, expression of ECAD and NCAD is important for the maintenance of tissue structure and integ- rity; loss of ECAD can be seen in the tran- sition from adenoma to carcinoma and has been suggested to induce invasiveness and metastasis formation [1].
NCAMs appear during morphogenesis as neural epithelial cells fold inward to form the neural crest/tube and maintain neuron- neuron adhesion. NCAMs disappear from the neural crest cells during migration to be reinduced during formation of ganglia. In adult animals, loss of NCAM-mediated cell adhesion has been shown to play an important role in the tumorigenesis and metastatic spread of certain tumors [1,2].
CD44, also known as lymphocyte hom- ing receptor, is a hyaluronic acid-binding CAM, present on hematopoietic cells. Expression of CD44 and CD44 variants has been shown to be important for tumor cell migration and metastases [3].
In the present study, we have examined the expression of CAM in adrenal tumors. Three main types of primary adrenal tumors can be identified in adult humans: adrenocortical adenomas (ACAs), adreno- cortical carcinomas (ACCs), and pheochro- mocytomas (PCs) [4,5].
Microscopically, ACAs exhibit clear, lipid-laden cells arranged in sheets or nests. In published series, the majority of surgi- cally treated ACAs were hormonally active (glucocorticoids and mineral corticoids). Nonfunctional ACAs were observed in as many as 5% of adult autopsies [6]. ACCs frequently secrete well-known steroid hor- mones, their intermediates, or precursors, leading to clinical manifestations in about half of cases. ACCs are microscopically built up by both clear and compact cells and show varying degrees of nuclear pleo-
morphism. Elevated excretion of 17-keto- steroids, large tumor size (>6 cm), and evi- dence of locally invasive or distant metastatic disease are features of malig- nancy [7-10].
PCs arise from the chromaffin cells of the adrenal medulla and are the most com- mon type of adrenal medullary tumor in the adult, usually benign with synthesis of catecholamines (reflected by expression of the rate-limiting enzyme tyrosine hydroxy- lase [TH]) and chromogranins as carrier. The histologic appearance is highly vari- able, sometimes with pronounced cellular pleomorphism. Demonstration of distant metastases is the only indisputable crite- rion of malignancy [11-13]. All PCs are hormonally active and secrete catechola- mines and chromogranins as carrier mol- ecules. Adrenal tumors frequently present problems in the assessment of their bio- logic behavior. Attempts to identify histo- pathologic markers indicating malignancy in adrenal cortical tumors have been made [5,8-11,14,15], but the clinical outcome of an individual tumor is still difficult to predict.
New markers that can distinguish between benign and malignant adrenal tumors are clearly needed [16-22]. CAMs have been shown to play an important role in tumorigenesis and may serve as a prog- nostic marker for many tumor types [1,2,23,24], but the role of CAM in adrenal tumors has not been studied in detail [25].
The aim of the present study was to evaluate the expression pattern of CAM in adrenal tumors regarding origin (cortex vs medulla) and biologic behavior (benign vs malignant). We chose to study the expres- sion of NCAD, ECAD, NCAM, and CD44 in several tumors because these CAMs have previously been shown to be of great importance in human cancer development.
| Table 1. Characterizations of Primary Antibodies Used for Immunocytochemistry | |||||||
|---|---|---|---|---|---|---|---|
| Antigen | Epitope | Typeª | Clone | Catalogue | Dilution | Antigen retrieval | Source |
| TH | Unknown | MC | 2/40/15 | 1017 381 | 1/150 | Microwave treated, EDTA-NaOH, pH 6.0 | Boehringer Mannheim GmbH (Germany) |
| CD44 | Unknown | MC | A3D8 | C7923 | 1/100 | Microwave treated, citrate buffer, pH 6.0 | Sigma (St. Louis, MO) |
| ECAD | Unknown | MC | HECD-1 | 13-1700 | 1/1000 | Microwave treated, EDTA-NaOH, pH 6.0 | Zymed (San Francisco, CA) |
| NCAD | Unknown | MC | 3B9 | 18-0224 | 1/50 | Microwave treated, EDTA-NaOH, pH 6.0 | Zymed |
| NCAM | Unknown [26] | MC | ERIC-1 | SC-106 | 1/500 | Microwave treated, EDTA-NaOH, pH 6.0 | Santa Cruz Biotechnology (Santa Cruz, CA) |
| Ki-67 | 66-bp encoded repetitive element [29] | MC | MIB-1 | 0607 | 1/50 | Microwave treated, EDTA-NaOH, pH 8.0 | Immunotech (Marseille, France) |
“MC, mouse monoclonal.
Materials and Methods
Tumors
A total of 87 adrenal tumors were investigated by immunocytochemistry (ICC) and/or Western blotting. Benign adrenal tumors (sporadic and familial) were randomly selected among specimens reg- istered at the Department of Pathology, Sahlgrenska University Hospital between 1965 and 1999. All malignant adrenal tumors, PCs, and ACCs, registered at the Department of Pathology between 1979 and 1999 were included. The criterion for malignancy was the presence of distant metastases at follow-up (range: 3-324 mo). Fresh tumor tissues were obtained from patients operated at the Endocrine Surgi- cal Unit, Sahlgrenska University Hospital. The following tumors were investigated by ICC: 27 benign PCs (12 familial PCs, multiple endocrine neoplasia 2A), 8 spo- radic malignant PCs, 30 ACAs (24 non- functional, 1 Cushing, and 5 Conn syndromes), and 22 ACCs [26,27]. Fresh
tumor tissues available from 30 tumors (10 sporadic benign PCs, 10 ACAs, and 10 ACCs) were analyzed by Western blotting.
Immunocytochemistry
For ICC staining routinely fixed (for- malin) and paraffin-embedded material was used. Sections (5 um) were mounted on positively charged glass slides, deparaf- finized, and rehydrated. All sections were subjected to antigen retrieval by microwave treatment. Primary antibody was applied overnight at 4℃ (Table 1). Bound anti- bodies were visualized by indirect immuno- peroxidase techniques (Dako EnVision +; Dako). All slides were coded and evaluated by three observers. Immunolabeling was graded as follows: - = < 1%, + = 1-24% of tumor cells positive, ++ = 25-75% of tumor cells positive, and +++ = >75% of tumor cells positive. The number of Ki-67 posi- tive nuclei and mitotic figures was counted
| TH | CD44 | NCAD | ECAD | NCAM | Ki-67 index | |
|---|---|---|---|---|---|---|
| BPCs (n = 27) | 27/27 (0/0/0/27) | 23/27 (4/6/1/16) | 12/27 (15/4/6/2) | 0/27 (27/0/0/0) | 26/27 (0/2/8/16) | 8 ± 12 |
| MPCs (n= 8) | 8/8 (0/0/0/8) | 6/8 (2/0/1/5) | 8/8 (0/3/1/4) | 0/8 (8/0/0/0) | 7/8 (1/0/1/6) | 14 ±24 |
| ACAs (n = 30) | 0/30 (30/0/0/0) | 7/30 (23/6/1/0) | 28/30 (2/5/6/17) | 0/30 (30/0/0/0) | 21/30 (9/4/8/9) | 11 ±13 |
| ACCs (n = 22) | 0/22 (22/0/0/0) | 4/22 (18/1/2/1) | 8/22 (13/5/3/1) | 2/22 (20/2/0/0) | 17/22 (5/4/7/6) | 367 ±780 |
“The results for TH, CD44, N-cadherin, E-cadherin, and NCAM are given as the number of positive tumors/total number of tumors. The percentage of labeled cells in positive tumors is given in parentheses (0/+/++/+++). The number of positive cells was graded as follows: 0 = no labeled tumor cells, + = 1-24% positive cells, ++ = 25-75% positive cells, +++ = >75% positive cells. For Ki-67, the number of positive nuclei in 10 HPF is given as the mean ± SD.
| Table 3. Expression of CAMs in Adrenal Tumors Studied by Western Blottinga | ||||
|---|---|---|---|---|
| NCAM | ECAD | NCAD | CD44 | |
| ACCs | 8/10 | 0/10 | 5/10 | 6/10 |
| ACAs | 9/10 | 0/10 | 9/10 | 9/10 |
| BPCs | 10/10 | 0/10 | 10/10 | 10/10 |
“Values represent the number of positive tumors/total number of tumors investigated.
in 10 high-power fields (HPF) (0.16 mm2) from each tumor.
Western Blotting
Frozen tumor tissues were homogenized in 10 mM potassium-phosphate buffer (pH 6.8) containing 1 mMEDTA, 10 mM 3-(3-cholamidopropyl) dimethylammonio- 1-propane sulfate, 1 ug/mL of aprotinin, 10 µg/mL each of leupeptin and pepstatin, and 1 mg/mL of Pefablock. Homogenates were sonicated twice for 15 s followed by centrifugation for 10 min at 10,000g. The clear supernatant was withdrawn, assayed for protein content according to Bradford, and stored at -80℃. Aliquots of proteins (20-35 µg protein) were diluted in sample buffer. Reducing agents were added followed by denaturation at 70℃ for 10 min and electrophoresis on precast polyacrylamide gels (10% NuPAGE Bis- Tris-gels; Novex, San Diego, CA) using NuPAGE 3-(N-morphonolino) propane- sulfonic acid (MOPS) sodium dodecyl sul- fate as running buffer. Proteins were
transferred to polyvinyldifluoride mem- branes using a NOVEX blotting system. Membranes were incubated with primary antibodies (Table 1) at 4℃ overnight fol- lowed by alkaline phosphatase-conju- gated goat-anti-mouse antibody and CDP-Star (Tropix, Bedford, MA) as substrate. Membranes were exposed to enhanced chemiluminescence film at room temperature for 10-120 s. Molecu- lar weight markers (See-Blue or Mark 12; Novex) were used to calculate the appar- ent size of immunoreactive proteins.
Results
The results of ICC and Western blot- ting are summarized in Tables 2 and 3.
N-cadherin
ICC demonstrated NCAD in the normal adrenal cortex and medulla. The labeling was confined to cell membranes of cortical parenchymal cells and chroma- ffin cells. Strong labeling was observed in all layers of the adrenal cortex, whereas labeling of the medulla was much weaker (Fig. 1). Twelve of 27 (45%) benign PCs (BPCs), 8 of 8 (100%) malignant PCs (MPCs), 28 of 30 (94%) ACAs, and 9 of 22 (41%) ACCs were positive for NCAD. The labeling was confined to the cell mem- branes of tumor cells with no staining over
NCAD
NCAM
CD44
Cortex
Medulla
stroma, blood vessels, or lymphocytes (Fig. 2). In only one adenoma, NCAD staining was predominant in the cyto- plasm. Two ACAs with negative staining were both nonfunctional. Western blotting demonstrated a single immunoreactive band of approx 100 kDa in 10 of 10 (100%) investigated BPCs, 9 of 10 (90%) ACAs, and 5 of 10 (50%) ACCs (Fig. 3). The results of ICC and Western blotting were identical except for three ACCs (two were positive by Western blotting but nega- tive by ICC, and one was positive by ICC but negative by Western blotting).
E-cadherin
ICC could not demonstrate ECAD in the normal adrenal cortex or medulla. None of 27 BPCs (0%), 8 MPCs (0%), and 30 ACAs (0%) were positive for ECAD. The vast majority of the ACCs was also negative; two of twenty-two ACCs were positive for ECAD with mainly cyto- plasmic staining (Fig. 2). Western blotting did not detect ECAD in any of the inves- tigated tumors (0 of 10 PCs, 0 of 10 ACAs,
0 of 10 ACCs). The results of ICC and Western blotting were identical in all tumors except for three highly malignant ACCs (stage IV), which all were positive only by ICC.
Neural Cell Adhesion Molecule
ICC demonstrated NCAM in the normal adrenal cortex and medulla. The labeling was confined to cell membranes of both cortical parenchymal cells and chromaffin cells. Strong labeling was observed in the medulla, whereas labeling in all layers of the adrenal cortex was weaker (Fig. 1). Immunocytochemically 26 of 27 (96%) BPCs, 7 of 8 MPCs (88%), 21 of 30 (70%) ACAs, and 17 of 22 (78%) ACCs were positive for NCAM. The label- ing was mainly distributed over the cell membranes of tumor cells (Fig. 2). How- ever, a few tumor cells also had a cytoplasmic staining. Western blotting demonstrated double immunoreactive bands of approx 180 and 150 kDa, respectively, in 10 of 10 (100%) investigated BPCs, 9 of 10 (90%) ACAs, and 8 of 10 (80%) ACCs (Fig. 3).
PC -NCAD
MPC - CD44
ACC - NCAD
ACC - ECAD
PC -NCAM
ACA - NCAM
The results of ICC were almost identical to those of Western blotting.
CD44
ICC demonstrated focal and weak CD44 labeling in cell membranes of cor- tical parenchymal cells (zona glomerulosa) and strong labeling in cell membranes of chromaffin cells of the medulla (Fig. 1).
Positive CD44 labeling was demonstrated in 23 of 27 (85%) BPCs and 6 of 8 (75%) MPCs, while the figures for ACA and ACC were 7 of 30 (23%) and 4 of 22 (18%), respectively. All PCs had strong membrane staining, while cortical tumors had very weak cytoplasmic staining (Fig. 2). CD44+ lymphocytes were also observed in the stroma. However, the number of positive
PC
ACA
ACC
1 2 3 4 5 6 7 8 9 10 11 12
- 191 kDa
NCAD
- 64 kDa
- 51 kDa
- 39 kDa
PC
ACA
ACC
1 2 3 4
5 6 7 8
9 10 11 12
- 191 kDa
NCAM
- 64 kDa
- 51 kDa
- 39 kDa
PC
ACA
ACC
1 2 3 4 5 6 7 8 9 10 11 12
- 191 kDa
CD44
- 64 kDa
- 51 kDa
- 39 kDa
lymphocytes was low in all tumors exam- ined. Western blotting demonstrated a single immunoreactive band of approx 65 kDa in all tumors studied. Ten of 10 BPCs (100%) and 9 of 10 (90%) ACAs were positive, while only 6 of 10 (60%) ACCs turned out to be positive. The results con- firm those seen for BPCs and ACAs using
ICC but differ for ACCs (all stage IV). These four ACCs were positive using West- ern blotting, but were negative using ICC.
Ki-67 and Mitotic Index
The number of Ki-67-positive tumor cells in adrenal tumors was as follows (mean
+ SEM immunopositive cells per 10 HPF, n = number of tumors): BPCs (8 ± 12/ 10 HPF; n = 27), MPCs (14±24/10 HPF; n = 8), ACAs (11 ± 13/10 HPF; n = 30), ACCs (367 ± 780/10 HPF; n = 22). The corresponding figures for mitotic index were as follows: BPCs (0.17 ±0.38/10 HPF; n = 27), MPCs (0.14 + 0.37/10 HPF; n = 8), ACAs (0.07 ± 0.25/10 HPF; n = 30), ACCs (3.25 ±2.75/10 HPF; n = 22) (Fig. 2).
Tyrosine Hydroxylase
All adrenal tumors were stained with antibody against TH in order to confirm or exclude the medullary origin of the tumors. All BPCs and MPCs, demon- strated strong cytoplasmic staining for TH. No staining was observed in any adreno- cortical adenomas (0/30) or adrenocorti- cal carcinomas (0/22). Delicate TH-positive nerve fibers and varicosities were observed in the adrenal cortex. These findings con- firm the adrenomedullary origin of all benign and malignant PCs [28].
Discussion
CAMs have been shown to be impor- tant for tumorigenesis and invasiveness of many tumors [1,2,24]. Assessment of the malignant potential of adrenal tumors is difficult [11-14]. In the present study, we evaluated the expression of CAM in adre- nal tumors and their possible role in tumor development.
NCAD expression in normal adrenal cortex and medulla has not been thor- oughly evaluated [29,30]. In endocrine tumors emanating from pancreatic B-cells, NCAD is not downregulated with the transition from adenoma to carcinoma [1]. In the present study, NCAD was dem- onstrated in both normal adrenal medulla and cortex. NCAD was immunocyto- chemically expressed in all types of adre-
nal tumors studied. The results were con- firmed by Western blotting, showing only slightly different expression. Thus, NCAD is expressed in both cortical and medul- lary tumors regardless of their malignant potential. During cell transformation and tumor progression, NCAD expression seemed to be upregulated in PCs but was downregulated in cortical tumors (Fig. 4). NCAD therefore appears to be involved in the development of malignant tumors of both the adrenal medulla and the adre- nal cortex. However, it is not clear why NCAD is upregulated in malignant med- ullary tumors, but downregulated in adre- nal cortical tumors. The downregulation of NCAD in ACCs is consistent with the concept of NCAD being a tumor suppres- sor and may therefore be of prognostic sig- nificance in these tumors.
ECAD is not expressed in the adrenal cortex [31]. It has been reported that the loss of ECAD expression coincides with the transition from well-differentiated adenoma to invasive carcinoma in a transgenic mouse model of pancreatic B-cell carcinogenesis [1,2] and during the development of sev- eral human epithelial cancers [24]. In the present study, ECAD was not demon- strated in the normal adrenal cortex and medulla, nor could it be detected in the adrenal tumors studied. ECAD is thus not important for normal adrenal gland func- tion or for the development of adrenal tumors. Furthermore, ECAD has no use as a prognostic marker.
NCAM is transiently expressed in many tissues during embryogenesis, including epithelial cells. In adult tissue, the expres- sion of NCAM is confined to the nervous system, skeletal muscle cells, and some neu- roendocrine organs [2,32]. NCAM expres- sion is observed in a variety of human tumors, including those with neural ori- gin but also those not related to the ner- vous system [2,23,31]. In the present study,
| Chromaffin cell | Pheo | Malignant Pheo |
|---|---|---|
| + CD44 | + CD44 | + CD44 |
| + NCAM | + NCAM | + NCAM |
| + N-CAD | + N-CAD | + N-CAD |
| - E-CAD | - E-CAD | - E-CAD |
| Cortical cell | ACA | ACC |
| + CD44 | + CD44 | + CD44 |
| + NCAM | + NCAM | + NCAM |
| + N-CAD | + N-CAD | ± N-CAD |
| - E-CAD | - E-CAD | + E-CAD |
Fig. 4. During cell transformation and tumor progression, NCAD expression seemed to be downregulated in ACCs.
NCAM was demonstrated in both the nor- mal adrenal cortex and medulla. Likewise, both adrenocortical and medullary tumors (benign and malignant) expressed NCAM; Western blotting confirmed this result. NCAM is thus expressed in normal adre- nal glands. We obtained no evidence for involvement of NCAM in adrenal tum- origenesis, since there was no relation between NCAM expression and malignant phenotype.
CD44 has been demonstrated with focal labeling in the cell membrane of cortical parenchymal cells and strong labeling in the cell membranes of chromaffin cells of the adrenal medulla [33]. The possibility of using CD44 as a marker to differentiate ACC from PC has been examined. This analysis revealed morphologic differences between adrenocortical tumors and PCs; greater than 80% of ACAs and almost 70% of ACCs showed weak to moderate cyto-
plasmic staining, but all PCs had strong membrane staining [33]. Our immuno- cytochemical results confirmed that CD44 was expressed in all types of adrenal tumors; all PCs had strong membrane staining, whereas cortical tumors had very weak cytoplasmic staining. Western blotting confirmed the immunocytochemical results. The role of CD44 in tumorigen- esis of adrenal tumors and its importance as a prognostic marker are still unclear.
Ki-67 is a proliferation marker, which is detectable in all phases of the cell cycle. Apoptosis appears to play an important role in the maintenance of cell turnover in nor- mal cortex, as well as in tumor cell dynam- ics. High Ki-67 activity has been reported for ACCs with significantly higher values than seen for ACAs. In other studies, Ki-67 activity was highest in ACCs with metastases. Ki-67 activity had a specificity of 100% and a sensitivity of 50% in determining
10000
1000
£4ª
100
10
1
ACA ACC BPC FPC MPC
the malignant potential of PCs [34-40]. In our study, the Ki-67 index was higher in ACC than in any other type of adrenal tumor. Half of the ACCs had a Ki-67 index higher than 100 (Fig. 5). For adrenal med- ullary tumors, Ki-67 staining was expressed in both MPCs and BPCs with no statisti- cal differences. The Ki-67 index thus has a potential to discriminate between malig- nant and benign cortical tumors.
In conclusion, adrenal tumors express specific patterns of CAM, but these pat- terns do not correlate to the malignant potential. NCAD seemed to be differen- tially expressed in adrenal cortical and medullary tumors; less than half of the benign PCs displayed NCAD vs all malig- nant PCs, while almost all ACAs expressed NCAD vs less than half of the ACCs. Downregulation of NCAD in ACC may thus aid as a prognostic marker in adrenal cortical tumors. High expression of the proliferative marker Ki-67 and high mitotic activity in ACC are helpful in distinguish- ing between malignant and benign adreno- cortical tumors.
Acknowledgments
This work was supported by grants from the Swedish MRC (5520), Swedish Cancer Research Fund (3911), Göteborg Medical Society, The A. Gabrielsson Foundation, Sahlgrenska Hospital Research Funds, B. Uhlander Memorial Fund, and Camilla Landgren Memorial Fund.
References
1. Perl AK, Wilgenbus P, Dahl U, Semb H, Christofori G. A causal role of E-cadherin in the transition from adenoma to carcinoma. Nature 392:190-193, 1998.
2. Perl AK, Dahl U, Christofori G. Reduced expression of neural cell adhesion molecule induces metastatic dissemination of pancreatic tumor cells. Nat Med 5(3):286-291, 1999.
3. Wallach-Dayan Shulamit B, Grabovsky V, Moll J, Sleeman J, Herrlich P, Alon R, Naor D. CD44-dependent lymphoma cell dissemi- nation: a cell surface CD44, supports in vitro lymphoma cell rolling on hyaluronic acid sub- strate and its in vitro accumulation in the peripheral lymph nodes. J Cell Sci 114:3463- 3477, 2001.
4. Langman J. Central nervous system. Medical embryology. 4th ed. Chapter 20. Baltimore: Williams & Wilkins, 1980; 351-353.
5. Dahia PLM, Grossman AB. The molecular pathogenesis of adrenal tumors. Endocr Relat Cancer 2(4):267-279, 1995.
6. Schröder S, Padberg BC, Achilles E, Klöppel G. Immunohistochemistry in adrenocortical tumors: a clinicomorphological study of 72 neoplasms. Virchows Arkiv A, Pathol Anat Histopathol 420:65-70, 1992.
7. Saeger W. Pathology of adrenal neoplasm. Minerva Endocrinol 20:1-7, 1995.
8. Hough AJ, Hollifield JW, Page DL, Hartmann WH. Prognostic factors in adrenal cortical tumors. Am J Clin Pathol 72:390-399, 1979.
9. Weiss LM. Comparative histologic study of 43 metastasizing and nonmetastasizing adrenocortical tumors. Am J Surg Pathol 8:163-169, 1984.
10. Medeiros LJ, Weiss LM. New Developments in the pathologic diagnosis of adrenal cortical neoplasms. Am J Clin Pathol 97:73-83, 1992.
11. Grant CS. Pheochromocytoma. In: Clark OH, Duh Q-Y, eds. Textbook of endocrine surgery. Chapter 65. Philadelphia: Saunders, 1997; 513-522.
12. Montresor E, Iacono C, Nifosi F, Serio G. Retroperitoneal Paragangliomas: role of immunohistochemistry in the diagnosis of malignancy and in assessment of prognosis. Eur J Surg 160:547-552, 1994.
13. John H, Ziegler WH, Hayri D, Jaeger P. Pheo- chromocytomas: can malignant potential be predicted? Urology 53(4):679-683, 1999.
14. Komminoth P, Roth J, Schröder S, Sare- maslani P, Heitz PU. Overlapping expression of immunohistochemical markers and synapto- physin mRNA in pheochromocytomas and adrenocortical carcinomas. Lab Invest 72(4):424-431, 1995.
15. McCluggage WG, Burton J, Maxwell P, Sloan JM. Immunohistochemical staining of nor- mal, hyperplastic, and neoplastic adrenal cor- tex with a monoclonal antibody against alpha inhibin. Clin Pathol 51:114-116, 1998.
16. Miettinen M. Neuroendocrine differentiation in adrenocortical carcinoma: new immuno- histochemical findings supported by electron microscopy. Lab Invest 66:169-174, 1992.
17. Alsabeh R, Medeiros LJ, Weiss LM. Adrenal cortical tumors clinically mimicking pheo- chromocytoma. Am J Clin Pathol 104(4): 382-390, 1995.
18. Haak HR, Fleuren GJ. Neuroendocrine dif- ferentiation of adrenocortical tumors. Cancer 75(3):860-864, 1995.
19. Clarke MR, Weyant RJ, Watson CG, Carty SE. Prognostic markers in pheochromo- cytoma. Hum Pathol 28:522-526, 1998.
20. Pelky TJ, Frierson HF Jr, Mills SE, Stoler MH. The alpha subunit of inhibin in adrenal cortical neoplasia. Mod Pathol 11(6):516-524, 1998.
21. De Krijger RR, Van Der Harst E, Van Der Ham F, Stijnen T, Dinjens WN, Koper JW, Bruining HA, Lamberts SWJ, Bosman FT. Prognostic value of p53, bcl-2, and c-erbB-2 protein expression in phaeochromocytomas. J Pathol 188:51-55, 1999.
22. Tartour E, Caillou B, Tenenbaum F, Schröder S, Luciani S, Talbot M, Schlumberger M. Immunohistochemical study of adrenocorti- cal carcinoma. Cancer 73:3296-3303, 1993.
23. Hage R, Elbers HRJ, de la Riviéra AB, van den Bosch JMM. Neural cell adhesion molecules expression. Clin Invest 114(5):1316-1320, 1998.
24. Christofori G, Semb H. The role of cell- adhesion molecule E-cadherin as a tumor-sup- pressor gene. TIBS 24:73-76, 1999.
25. Gupta D, Shidham V, Holden J, Layfield L. Prognostic value of immunohistochemical expression of topoisomerase alpha II, MIB-1, p53, E-cadherin, retinoblastogene protein product, and HER-2/neu in adrenal and extra- adrenal pheochromocytoma. Appl Immuno- histochem Mol Morphol 8(4):267-274, 2000.
26. Khorram-Manesh A, Ahlman H, Jansson S, Wängberg B, Nilsson O, Jakobsson CE, Eliasson B, Lindstedt S, Tisell LE. Adreno- cortical carcinoma: surgery and mitotane for treatment and steroid profiles for follow-up. World J Surg 22:605-612, 1998.
27. Ahlman H, Khorram-Manesh A, Jansson S, Wängberg B, Nilsson O, Jacobsson CE, Lindstedt S. Cytotoxic treatment of adreno- cortical carcinoma. World J Surg 25:927-933, 2001.
28. Tischler AS, Sheldon W, Gray R. Immunohisto- chemical and morphological characterization of spontaneously occurring pheochromo- cytoma in the aging mouse. Vet Pathol 33:512-520, 1996.
29. Kim JB, Islam S, Kim Y, Prudoff RS, Sass KM, Wheelock MJ, Johnson KR. N-cadherin extracellular repeat 4 mediates epithelial to mesenchymal transition and increased motil- ity. J Cell Biol 151:1193-1204, 2000.
30. Lambert M, Padilla F, Mége RM. Immobi- lized dimers of N-cadherin-Fc chimera mimic cadherin-mediated cell contact formation: contribution of both outside-in and inside- out signals. J Cell Sci 113:2207-2219, 2000.
31. Zeromski J, Lawniczak M, Golusinski P. Expression of CD56/N-CAM antigen and some other adhesion molecules in various human endocrine glands. Folia Histochem Cytobiol 36(3):119-125, 1998.
32. Meyer MB, Bastholm L, Nielsen MH, Elling F, Rygaard J, Chen W, Öbrink B, Bock E, Edvardsen K. Localization of NCAM on NCAM-B-express- ing cells with inhibited migration in collagen. APMIS 103:197-208, 1995.
33. Barshack I, Goldberg I, Nass D, Olchovsky D, Kopolovic J. CD44 expression in normal adrenal tissue and adrenal tumors. J Clin Pathol 51:52-54, 1998.
34. Iino K, Sasano H, Yabuki N, Oki Y, Kikuchi A, Yoshimi T, Nagura H. DNA topoisomerase II alpha and Ki-67 in human adrenocortical
neoplasms: a possible marker of differentia- tion between adenomas and carcinomas. Mod Pathol 10(9):901-907, 1997.
35. Nakazumi H, Sasano H, Iino K, Ohashi Y, Orikasa S. Expression of cell cycle inhibitor p27 and Ki-67 in human adrenocortical neoplasms. Mod Pathol 11(12):1165-1170, 1998.
36. Sasano H, Imatani A, Shizawa S, Suzuki T, Nagura H. Cell proliferation and apoptosis in normal and pathologic human adrenal. Mod Pathol 8(1):11-17, 1995.
37. Key G, Becker MHG, Baron B, Duchrow M, Schlüter C, Flad HD, Gerdes J. New Ki-67- equivalent murine monoclonal antibodies (MIB1-3) generated against bacterially expressed parts of the Ki-67 cDNA containing three
62 base pair repetitive elements encoding for the Ki-67 epitope. Lab Invest 68(6):629-636, 1993.
38. Nagura S, Katoh R, Kawaoi A, Kobayashi M, Obara T, Omata K. Immunohistochemical estimations of growth activity to predict biological behavior of pheochromocytoma. Mod Pathol 12(12):1107-1111, 1999.
39. Vargas MP, Vargas HI, Kleiner DE, Merino MJ. Adrenocortical neoplasm: role of prog- nostic markers MIB-1, P53, and RB. Am J Surg Pathol 21(5):556-562, 1997.
40. Edgren M, Eriksson B, Oberg K. Biological characteristics of adrenocortical carcinomas: a study of p53, IGF, EGF-r, Ki-67 and PCNA in 17 adrenocortical carcinomas. Anticancer Res 17(2B):1303-1309, 1997.