An Immunostaining Panel for Diagnosis of Malignancy in Mucinous Tumors of the Pancreas.

The detection of malignant potential in pancreatic neoplasms, especially in mucinous cystic tumors (MTs), is of clinical importance, since the surgical therapy is different for benign and malignant neoplasms. Currently, the distinction between benign and malignant MTs of the pancreas depends on the demonstration of tumor invasion in hematoxylin-eosin--stained histologic sections. Invasion may be focal and demonstrated only by extensive sectioning, which may be difficult to achieve in large specimens. It would therefore be of practical importance to have other markers available to indicate which cystic neoplasms might be malignant.

Carcinogenesis in the pancreas is a multihit process that involves various factors in DNA, RNA, and protein synthesis. In recent years, mutation of p53, the tumor suppressor gene, has been reported as one of the frequent genetic alterations in pancreatic carcinoma.[1-5] There are also reports of increased expression of HER-2/neu,[6-8] epithelial growth factor receptor (EGFR),[9-12] and transforming growth factor [Alpha] (TGF-a)[13,14] in malignant pancreatic neoplasms. However, the patterns of expression or coexpression of these markers and their relationship to proliferative index in pancreatic adenocarcinomas, MTs, and nonneoplastic pancreata have not been studied.

In the current study, we compared the expression of p53, HER-2/neu, EGFR, TGF-[Alpha], and cell proliferation marker Ki-67 in invasive ductal adenocarcinomas (CAs), MTs, chronic pancreatitis tissue (CPs), and normal pancreatic tissue (NLs). The aim was to detect whether patterns of expression of these markers might suggest malignant potential in MTs.

MATERIALS AND METHODS

Cases Studied

Partial or total pancreatic resection specimens were identified in the surgical pathology files at UMASS Memorial Health Care, Worcester, Mass, from 1993 to 1999. Sixty-five cases were obtained, including 30 CAs, 9 MTs, 14 CPs, and 12 NLs. The routine hematoxylin-eosin-stained sections were reviewed to confirm the diagnosis and select blocks for immunostaining. All cases in the CA group were invasive ductal adenocarcinomas. Of 9 cases of MTs, 8 were benign and 1 was of borderline malignant potential.

Immunohistochemistry Staining

The immunohistochemical staining was performed on 4 [micro]m of formalin-fixed and paraffin-embedded tissue sections. After deparaffinization and rehydration, the slides were subjected to antigen retrieval by microwave treatment for 5 minutes in 0.01 mol/ L citrate buffer at pH 6.0. The slides were then stained with antibodies on a TechMate 1000 (Ventana Medical Systems, Tucson, Ariz) automated immunostainer using a standard avidin-biotin complex staining procedure. The antibodies used were p53; a cocktail of 2 monoclonals (Oncogene Science, Cambridge, Mass), AB2 at the dilution of 1:300 and AB6 at the dilution of 1:1000; HER-2/neu, polyclonal (Dako, Carpinteria, Calif) at a dilution of 1:1200; Ki-67, monoclonal (Immunotech, Westbrook, Me) at a dilution of 1:100; EGFR, monoclonal (Novocastra, Newcastle Upon Tyne, England) at a dilution of 1:20; and TGF-[Alpha], monoclonal (Oncogene Science) at a dilution of 1:500. The proper positive controls for each antibody and negative controls were processed at the same time. All the sections were counterstained with hematoxylin.

Evaluation of Immunostaining

Both the quantity (percentage of positive cells) and quality (staining intensity) of immunostaining were assessed. Nuclear staining for p53 and Ki-67, cytoplasmic membrane staining for HER-2/neu and EGFR, and cytoplasmic staining for TGF-[Alpha] were considered positive. In nonneoplastic pancreatic tissue, only ducts were evaluated. The percentage of positive cells was scored on a semiquantitative scale as follows: 0 equals less than 5% of cells positive, 1 equals 5% to 33% positive, 2 equals 34% to 66% positive, and 3 equals 67% or more positive. Staining intensity was graded as 1+ equals weak, 2+ equals moderate, and 3+ equals strong.

Statistics

Percentage scores 1, 2, and 3 ([is greater than or equal to] 5% cells stained) were considered positive. The scores 2 and 3 ([is greater than or equal to] 34% cells positive) were considered strong positive staining for purposes of comparisons among the groups of tumors. Positive immunostaining for each antibody was compared among the groups using 2-tailed [chi square] and Fisher exact tests. A P value less than .05 was considered significant. The correlations of expression of p53, HER-2/neu, EGFR, and Ki-67 in carcinomas were tested by the Pearson correlation coefficient test.

RESULTS

Frequency of Positive Immunostaining

The results of the immunochemical staining are summarized in Figure 1, and the staining patterns for p53 and HER-2/neu in the NLs, MTs, and CAs are shown in Figure 2, A through C.

Figure 1. Graph showing the percentage of cases in
each diagnostic group with more than 5% of cells positive for
each antibody. EGFR indicates epithelial growth factor receptor;
TGF-[Alpha], transforming growth factor [Alpha];
NLs, normal pancreatic tissue specimens; MTs, mucinous cystic tumors;
CPs, chronic pancreatitis specimens; and CAs, primary pancreatic
ductal adenocarcinomas.

Cases, %
                p53    HER-2/neu    Ki-67    EGFR     TGF-[Alpha]

NLS (n=12)       0       8.3         25       0            100
MTs (n=9)        0      42.8         50       7.1           78.6
CPs (n=14)      33.3    88.9        100      11.1          100
CAs (n=30)      73.3    80          100      36.6           83.3

Note: Table made from a bar graph.

p53 overexpression was detected only in neoplasms and was more frequent in the CAs (22 of 30 cases, 73.3%) than in the MTs (3 of 9 cases, 33.3%) (P = .000028 to .05). Strong expression (score 2 or 3, [is greater than or equal to] 34% cells) of p53 was seen only in the CAs (46.7%) (P = .00126 to .00964) (Figure 3). The expression of HER-2/neu was significantly more frequent in neoplasms (CAs = 80% and MTs = 88.9%) than in CPs (42.8%) or NLs (8.3%) (P = .000028 to .03308) (Figure 3). However, unlike p53, there was no significant difference in the frequency of HER-2/neu expression for CAs compared with the MTs (P = .3744). Ki-67 positivity in more than 5% of cell nuclei was found in all (100%) of the neoplasms, 50% of CPs, and in ducts of 25% of the NLs (P = .000049 to .014 for CAs and MTs compared with CPs and NLs). Epithelial growth factor receptor was expressed in 36.6% of the CAs, 11.1% of the MTs, 7.1% of the CPs, and none of the NLs (P = .0128 and .0363 for CAs compared with NLs and CPs). As with p53, strong expression ([is greater than or equal to] 34% positive cells) of EGFR (13.3%) was seen only in the CAs. Transforming growth factor [Alpha] was present in all cases and did not discriminate among the groups.

A
        3+     2+     1+      0

NLs                         100

CPs                         100

MTs                  33.3  66.7

CAs    36.7    10    26.7  26.6

B
       3+     2+     1+      0

NLs                   8.3  91.7

CPs           22.2   35.7  57.2

MTs    22.2   55.6   11.1  11.1

Cas    33.3   30     26.7  20

Figure 3. Graph showing the percentage of cases in each diagnostic
group with staining score of 0, 1, 2, or 3 for p53 (A) and HER-2/neu
(B) NLs indicates normal pancreatic tissue specimens; MTs, mucinous
cystic tumors; CPs, chronic pancreatitis specimens; and Cas, primary
pancreatic ductal adenocarcinomas.

Note: Table made from a bar graph.

The frequencies of coexpression of the markers are shown in Figure 4. Coexpression of p53 and HER-2/neu was present in 17 (56.7%) of 30 CAs and 2 (22.2%) of 9 MTs and was not detected in the CPs and the NLs. Similarly, coexpression of EGFR and HER-2/neu was more frequent in the CAs (26.7%) than in the MTs (11.1%) and not present in the CPs and the NLs. Coexpression of p53 and EGFR was seen only in the CAs. No conclusive correlation was found among p53, EGFR, HER-2/neu, and Ki-67 expressions in the CAs.

Figure 4. Graph showing the percentage of
cases in each diagnostic group with coexpression
of p53/HER-2/neu, epithelial growth
factor receptor (EGFR)/HER-2/neu, and p53/
EGFR. NLs indicates normal pancreatic tissue
CPs, chronic pancreatitis specimens; and
CAs, primary pancreatic ductal adenocarcinomas.

                  NLs        CPs       MTs      CAs
               (n = 12    (n = 14)   (n = 9)  (n = 30)

p53/HER-2/neu     0          0         22.2     56.7
EGFR/HER-2/neu    0          0         11.1     26.7
P53/EGFR          0          0            0     23.3

Note: Table made from a bar graph.

Staining Intensity

For p53 and EGFR, moderate-to-strong intensity of staining (2+ and 3+) was seen only in the CAs (56.6% and 20%). A 2+ staining of HER-2/neu was more frequent in the CAs (46.7%) than in the benign MTs (22.2%). There was a good correlation between the quantity and intensity of positivity for p53, HER-2/neu, and EGFR (r = 0.814 to 0.955). For Ki-67 and TGR-[Alpha], there was no difference in staining intensity among the groups.

COMMENT

The multiple-hit model has been accepted for pancreatic carcinogenesis. Pancreatic tumors are now being evaluated for accumulation of damage to critical regulatory genes and growth factors, p53 is a tumor suppressor gene that encodes a 53-kd nuclear phosphoprotein that is intimately involved in the regulation of transcription, DNA repair, and apoptosis.[15,16] Mutation of p53 leads to accumulation of nuclear p53 protein, overexpression of which has been used as an immunohistochemical marker for p53 mutation.[2] There are numerous reports in recent years describing alterations in the p53 gene or p53 protein in various pancreatic cancers.[1-5] Likewise, our study demonstrated that p53 was overexpressed in 73.3% of the CAs and 33.3% of the benign MTs and was virtually absent in NLs and CPs. Strong expression (with a cutoff [is greater than or equal to] 34% cells) and moderate and strong (2+ and 3 +) intensity of staining of p53 were seen only in CAs. Our results were consistent with other studies[17-18] and indicate that p53 can be a very useful marker to aid in detection of malignant potential in MTs of the pancreas and in differentiating reactive atypia in CPs from a well-differentiated adenocarcinoma.

The HER-2/neu (c-erb-b2) proto-oncogene encodes a 185-kd transmembrane glycoprotein that is closely related in structure to the EGFR and has been found to be amplified and overexpressed in pancreatic tumors.[19,20] Hall et al[21] reported that HER-2/neu was detected in the cell cytoplasm of both CPs and CAs. The study by Yamanaka et al[6] also showed HER-2/neu expressed in normal acinar and ductal cells. Our study also showed HER-2/neu immunoreactivity in CPs, MTs, and CAs. This suggests that the c-erb-b2 oncoprotein may possibly be involved in proliferative responses of pancreatic epithelium. Dugan et al[TM] reported that HER-2/neu gene expression is not common in tumors that lack glandular differentiation and suggested that the pattern of the HER-2/neu expression is related to glandular differentiation and early oncogenesis. Our study showing HER-2/neu expression in MTs and in CAs further supports Dugan's concept.

Epithelial growth factor receptor belongs to a family of closely related transmembrane proteins, including HER-2/neu.[23] Its expression has been reported in 30% to 50% of pancreatic carcinomas.[10,11,24] Our results are consistent with these reports. There are various data concerning the possibility of synergy or other interactions between the products of the c-erb-b2 gene and other growth factor receptors, including EGFR.[25] It has been said that coexpression of HER-2/neu and EGFR at levels insufficient individually to transform cells will convert fibroblasts to a fully malignant state.[26] Whether this is the case in pancreatic epithelium is not known. In our study, coexpression of HER-2/neu and EGFR was seen only in neoplasms, both benign and malignant. It is possible that HER-2/neu and EGFR have a synergic effect in the carcinogenesis of pancreatic tumors. However, the correlation coefficient analysis of our results did not confirm this relationship statistically.

Transforming growth factor [Alpha] overexpression has been reported in pancreatic tumors.[13,27] However, we were unable to confirm these results, since TGF-[Alpha] expression was found in all the groups and was not discriminative.

The impetus for this study was to identify 1 or more markers that might be helpful predictors of malignant potential in MTs of the pancreas and indicate which one might require further workup to search for invasive foci. When the group of MTs was compared with the 2 nonneoplastic groups (CPs and NLs), p53 (33% vs 0) and HER-2/neu (89% vs 43% and 8.3%) expression was more frequent and Ki-67+ (proliferating) nuclei were increased. In addition, coexpression of p53/HER-2/neu and EGFR/ HER-2/neu also discriminated between NLs/CPs and MTs. Furthermore, coexpression of p53/EGFR discriminated between benign and malignant tumors. Thus, in a group of MTs appearing histologically benign, some had increased markers indicative of malignancy. The follow-up on our group of MTs ranges from 3 months to 7 years, with no recurrences documented. Unfortunately, the follow-up on these patients is short. In addition, since this was a retrospective study, we cannot identify a group of MTs in which further sectioning revealed outright carcinoma or in which the patients subsequently developed carcinoma. However, the limitations of this study do allow the conclusion that MTs are intermediate between outright invasive carcinomas and nonneoplastic tissues in their expression of p53, EGFR, and Ki-67. Our 1 MT that was of borderline histology in routine sections was negative for p53 but positive for HER-2/neu and had an intermediate proliferative rate by Ki-67 staining.

In summary, the findings of this study were as follows. First, the CAs showed an increased expression of p53, HER-2/neu, Ki-67, and EGFR. Second, the MTs were intermediate between invasive carcinomas and nonneoplastic tissues in their expression of p53, EGFR, and Ki-67. Third, p53 was the most discriminative marker in the diagnosis of pancreatic malignancy. Its overexpression was seen only in the neoplasms and was significantly more frequent in the CAs than in the MTs. Fourth, coexpression of p53/HER-2/neu and EGFR/HER-2/neu discriminated between the neoplasms and the nonneoplastic tissues. Coexpression of p53/EGFR discriminated between the MTs and the CAs. Fifth, strong expression and strong intensity of staining (2+ and 3+) of p53 and EGFR were seen only in the CAs. These findings indicate that the patterns of expression of p53, HER-2/neu, EGFR, and Ki-67 detected by immunohistochemical staining can be a useful tool in the diagnosis of pancreatic neoplasms in routine pathology practice. In addition, it may also be helpful in the diagnosis of fine needle aspirates. Immunohistochemical staining is currently widely available in pathology laboratories in the United States, whereas the molecular study of tumor specimens often requires microdissection tissue samples to obtain highly enriched neoplastic components. To this point of review, using immunochemical methods to study genetic alteration of pancreatic tumors is faster, easier, and more practical and may be useful in both community hospitals and large laboratories.

This work was supported by the Department of Pathology, UMASS Memorial Health Care.

References

[1.] Boschman CR, Stryker S, Reddy JK, Rao MS. Expression of p53 protein in precursor lesions and adenocarcinoma of human pancreas. Am J Pathol. 1994; 145:1291-1295.

[2.] Barton CM, Staddon SL, Hughes CM, et al. Abnormalities of the p53 tumor suppressor gene in human pancreatic cancer. Br J Cancer. 1991;64:1076-1082.

[3.] Scarpa A, Capelli P, Mukai K, et al. Pancreatic adenocarcinoma frequently show p53 gene mutation. Am J Pathol. 1993;142:1534-1543.

[4.] Ruggeri B, Zhang S-Y, Caamano J, Di Rado M, Flynn SD, Klein-Szanto AJ. Human pancreatic carcinomas and cell lines reveal frequent and multiple alterations in the p53 and Rb-1 tumor-suppressor gene. Oncogene. 1992;7:1503-1511.

[5.] Redston MS, Caldas C, Seymour AB, et al. p53 mutations in pancreatic carcinoma and evidence of common involvement of homopolymer tracts in DNA microdeletions. Cancer Res. 1994;54:3025-3033.

[6.] Yamanaka Y, Friess H, Kobrin MS, et al. Overexpression of HER2/neu oncogene in human pancreatic carcinoma. Hum Pathol. 1993;24:1127-1134.

[7.] Satoh K, Sasano H, Shimosegawa T, et al. An immunohistochemical study of the c-erbB-2 oncogene product in intraductal mucin-hypersecreting neoplasms and in ductal cell carcinomas of the pancreas. Cancer. 1993;72:51-56.

[8.] Jaskiewicz K, Krige JEJ, Thomsom J. Expression of p53 tumor suppressor gene, oncoprotein c-erbB-2, cellular proliferation and differentiation in malignant and benign pancreatic lesions. Anticancer Res. 1994;14:1919-1922.

[9.] Friess H, Berberat P, Schilling M, Kunz J, Korc M, Buchler MW. Pancreatic cancer: the potential clinical relevance of alterations in growth factors and their receptors. J Mol Med. 1996;74:35-42.

[10.] Lemoine NR, Hughes CM, Barton CM. The epidermal growth factor receptor in human pancreatic cancer. J Pathol. 1992;166:7-12.

[11.] Kioppel G, Maillet B, Schwere K, et al. Immunocytochemical detection of epidermal growth factor receptor (EGFr) and transferrin receptor (TR) on normal, inflamed and neoplastic pancreatic tissue. Pancreas. 1989;4:623.

[12.] Yamanaka Y, Onda M, Uchida E. Immunochemical study on epidermal growth factor and its receptor in human pancreatic carcinoma. Pancreas. 1989; 4:649.

[13.] Barton CM, Hall PA, Hughes CM, Gullick WJ, Lemoine NR. Transforming growth factor alpha and epidermal growth factor in human pancreatic cancer. J Pathol. 1991;163:111-116.

[14.] Yamanaka Y, Friess H, Kobrin MS, Buchler M, Beger HG, Korc M. Coexpression of epidermal growth factor receptor and ligands in human pancreatic cancer is associated with enhanced tumor aggressiveness. Anticancer Res. 1993; 13:565-569.

[15.] Levine AJ, Momand J, Finley CA. The p53 tumor suppressor gene. Nature. 1991;351:453-456.

[16.] Steele RJ, Thompson AM, Hall PA, Lane DP. The p53 tumor suppressor gene. Br J Surg. 1998;85:1460-1467.

[17.] Islam HK, Fujioka Y, Tomidokoro T, et al. Immunohistochemical analysis of expression of molecular biologic factors in intraductal papillary-mucinous tumors of pancreas-diagnostic and biologic significance. Hepatogastroenterology. 1999;46:2599-2605.

[18.] Apple SK, Hecht JR, Lewin DN, Jahromi SA, Grody WW, Nieberg RK. Immunohistochemical evaluation of K-ras, p53, and HER-2/neu expression in hyperplastic, dysplastic, and carcinomatous lesions of the pancreas: evidence for multistep carcinogenesis. Hum Pathol. 1999;30:123-129.

[19.] Williams TM, Weiner DB, Greene MI, Maguire HC Jr. Expression of c-erbB-2 in human pancreatic adenocarcinoma. Pathobiology. 1991;59:46-52.

[20.] Day JD, Di Giuseppe JA, Yeo C, et al. Immunohistochemical evaluation of HER-2/neu expression in pancreatic adenocarcinoma and pancreatic intraepithelial neoplasms. Hum Pathol. 1996;27:119-124.

[21.] Hall PA, Hughes CM, Staddon SL, Richman PI, Gullick WJ, Lemoine NR. The c-erbB-2 proto-oncogene in human pancreatic cancer. J Pathol. 1990;161: 195-200.

[22.] Dugan MC, Dergham ST, Kucway R, et al. HER-2/neu expression in pancreatic adenocarcinoma: relation to tumor differentiation and survival. Pancreas. 1997;14:229-236.

[23.] Sirivatanauksorn V, Sirivatanauksorn Y, Lemoine NR. Molecular pattern of ductal pancreatic cancer. Langenbecks Arch Surg. 1998;383:105-115.

[24.] Uegaki K, Nio Y, Inoue Y, et al. Clinicopathological significance of epidermal growth factor and its receptor in human pancreatic cancer. Anticancer Res. 1997;17:3841-3848.

[25.] Sterm DF, Kamps MP. EGF-stimulated tyrosine phosphorylation of [p185.sup.neu] a potential model of receptor interactions. EMBO J. 1988;7:995-1001.

[26.] Kokai Y, Myers JN, Wada T. Synergistic interaction of [p185.sup.c-neu] and EGF receptor leads to transformation of rodent fibroblasts. Cell. 1989;58:287-292.

[27.] Lemoine NR, Hall PA. Growth factors and oncogenes in pancreatic cancer. Baillieres Clin Gastroenterol. 1990;4:815-832.

Accepted for publication January 19, 2001.

From the Department of Pathology, UMASS Memorial Health Care, Worcester, Mass. Dr Zhao is currently with the Department of Pathology and Laboratory Medicine, MCP Hahnemann University, Philadelphia, Pa.

Presented in part as a poster at the Annual Meeting of the American Society of Clinical Pathologists and College of American Pathologists, San Diego, Calif, October 2000.

Reprints: Barbara Banner, MD, Department of Pathology, UMASS University Hospital Center, 55 Lake Ave N, Worcester, MA 01655 (e-mail: [e-mail]).

Jin Zhao, MD; Sharon X. Liang, MD, PhD; Lou Savas, BS; Barbara F. Banner, MD