|
Journal of Postgraduate Medicine, Vol. 56, No. 4, October-December, 2010, pp. 262-266 Original Article p53, carcinoembryonic antigen and carbohydrate antigen 19.9 expression in gall bladder cancer, precursor epithelial lesions and xanthogranulomatous cholecystitis V Agrawal1, A Goel1, N Krishnani1, R Pandey1, S Agrawal2, VK Kapoor3 1 Department of Pathology, Sanjay Gandhi, Postgraduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India Correspondence Address:V Agrawal, Department of Pathology, Sanjay Gandhi, Postgraduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India, vinita@sgpgi.ac.in Date of Submission: 24-Dec-2009 Code Number: jp10078 PMID: 20935395 DOI: 10.4103/0022-3859.70933 Abstract Background : Gallbladder cancer (GBC) is the commonest gastrointestinal
cancer in women of north India. Precursor epithelial lesions in GBC are
known; however, the role of xanthogranulomatous (XG) inflammation in
the pathogenesis of GBC is unknown. Keywords: Chronic cholecystitis, gallbladder neoplasm, immunohistochemistry, precursor lesions, xanthogranulomatous cholecystitis Introduction Gallbladder cancer (GBC) is the commonest gastrointestinal cancer in women in north and central India. [1],[2] Gallstone disease and chronic cholecystitis (CC) are present in about 60-90% of cases with GBC. Xanthogranulomatous cholecystitis (XGC), a variant of CC, has been reported to be associated with 3-29% of GBC. [3],[4],[5],[6] The incidence of XGC in gallstone disease is also higher in this part of the world. [3],[7] However, the role of XGC in the pathogenesis of GBC is not known. Studies have shown that abnormalities in tumor suppressor gene, p53, are involved in the pathogenesis of GBC. [12],[15],[16],[17],[18],[19],[20] However, there is paucity of studies of p53 expression in XGC. [8] Carcino-embryonic antigen (CEA) and carbohydrate antigen 19.9 (CA-19.9) have been described as serum tumor markers in gastrointestinal malignancies including GBC. Altered CEA expression has been reported in GBC. [9],[10] However, there are no reports on the expression of CA-19.9 in benign gallbladder diseases including CC and XGC. We planned this study to determine the immunohistochemical expression of p53, CEA and CA-19.9 in GBC, CC, XGC and precursor epithelial lesions including metaplasia and dysplasia. We also studied the correlation of the expression of p53 with clinico-pathological parameters. Materials and Methods The study included 171 consecutive evaluable gallbladder (GB) specimens received in the Department of Pathology at the Sanjay Gandhi Postgraduate Institute of Medical Sciences, a tertiary level referral hospital at Lucknow in north India. The study was approved by the institutional ethics committee. The GB specimens included 51 GBC, 68 CC and 42 XGC. None of the GBC was detected incidentally. Ten normal GBs removed in surgeries with no primary GB disease, viz. pancreatico-duodenectomy for periampullary cancer, choledochal cyst excision and liver resection were included as controls. The GB was fixed in 10% buffered formalin for 6 to 12 h. Two sections each from the fundus, body and neck were taken in all non-malignant GBs. Multiple sections were taken in GBC. All GBs were evaluated for the presence of epithelial lesions including metaplasia (antral and intestinal) and dysplasia (low and high grade). In GBC the adjacent mucosa was analyzed for presence of epithelial lesions. Histological criteria for metaplasia and dysplasia were as described by Albores Saavedra et al. [11] Immunohistochemistry (IHC) was performed using mouse monoclonal primary antibodies for p53 (Clone DO-7, 1:100 dilution, Dako Corporation, Denmark), CEA (Clone II-7, 1:100 dilution, Dako Corporation, Denmark) and CA-19.9 (Clone 116-NS-19-9, 1:100 dilution, Dako Corporation, Denmark) using labeled Streptavidin-Biotin Peroxidase method (Dako Corporation, Denmark); 3-4 ΅m thick sections after deparaffinisation in xylene were rehydrated in graded alcohol. Endogenous peroxidase was blocked using 3% H 2 O 2 in methanol. Antigen retrieval was done using microwave pretreatment in citrate buffer (pH 6.0). The sections were incubated with primary antibody for 2 h at room temperature. After washing, the sections were treated with biotinylated secondary antibody for 30 min. A streptavidin-HRP (horse-radish peroxidase) conjugate was applied for 30 min and the reaction was visualized by using DAB (diaminobenzidine) as chromogen (Dako, Denmark). Counterstaining was done by Mayer′s hematoxylin. A section of colon carcinoma known to be strongly positive for p53 and CEA was included as a positive control for p53 and CEA. A section from normal GB was used as control for CA-19.9. Negative controls for immunostaining were prepared by replacing the primary antibody with non-immune rabbit serum. The pattern of distribution of CEA and CA-19.9 was studied and recorded as apical and cytoplasmic (focal or diffuse). The p53 immunostaining was analyzed by a semi-quantitative scoring as described by Wistuba et al. [12] The intensity of staining was recorded as absent (0), mild (1), moderate (2) and intense (3) while the percentage of nuclei stained were scored as absent/<5% (0), >5 <10% (1), 10-50% (2) and >50% (3). The total score was calculated by adding the intensity score and the percentage score and ranged from 0-6. A score of >3 was considered as positive. Results The mean age was 54 years (range 26-78) in GBC (n=51), 46 years (range 16-81) in CC (n=68), and 45 years (range 21-75) in XGC (n=42). The females formed the larger proportion with M: F ratio of 1:2 in GBC, 1:1.4 in CC and 1:1.6 in XGC. Cholelithiasis was present in 67% cases of GBC and in 97% cases and 83% cases of CC and XGC respectively. Associated XGC was present in eight (16%) cases of GBC. Evaluation of precursor epithelial lesions in the mucosa adjacent to carcinoma showed antral metaplasia in eight (16%), intestinal metaplasia in 15 (29%), low-grade dysplasia in 12 (24%) and high-grade dysplasia in 25 cases (49%). In CC, assessment for precursor epithelial lesions showed antral metaplasia in 30 (44%), intestinal metaplasia in 27 (40%), low-grade dysplasia in 17 (25%) and high-grade dysplasia in four (6%) cases. In XGC antral metaplasia was seen in 28 (67%), intestinal metaplasia in two (5%) and low-grade dysplasia in 10 (24%) cases; high-grade dysplasia was not present in any case. All 10 cases of normal GB were p53-negative and showed apical staining for CEA. CA-19.9 was diffusely positive in the epithelium in all cases. Significant p53 over-expression was observed in 28 of 51 GBC (55%) [Table - 1] [Figure - 1]. Evidence of p53 over-expression was found in two of 12 (17%) low-grade dysplasia and 10 of 25 (40%) high-grade dysplasia with an overall positivity of 32% in dysplasia [Figure - 2]. No p53 expression was seen in dysplasia associated with p53-negative tumors. No p53 expression was seen in antral and intestinal metaplasia found in the mucosa adjacent to the tumor. No significant correlation was observed between p53 over-expression and clinico-pathological parameters including age, sex and cholelithiasis, histological type, tumor grade, depth of invasion and tumor stage. In CC p53 over-expression was found only in three of 21 cases (14%) of dysplasia, all of which were cases with high-grade dysplasia. p53 over-expression was not seen in low-grade dysplasia associated with CC and XGC. No p53 expression was seen in antral and intestinal metaplasia associated with CC and XGC. CEA immunostaining was observed in 42 of 51 GBC cases (82%). The CEA staining pattern was apical in 13 and cytoplasmic focal (15/51) to diffuse (14/51) in 29 cases [Figure - 3]. CEA immunostaining was observed in 18 of 68 CC cases (27%). The CEA pattern was apical in all cases along with focal cytoplasmic in nine cases. In XGC CEA immunostaining was observed in four cases (10%) with an apical staining pattern. Diffuse cytoplasmic staining was found only in GBC. Seventy-five percent (38) cases of GBC were CA-19.9-positive and all showed cytoplasmic staining. The adjoining GB epithelium showed diffuse CA-19.9 apical and cytoplasmic positivity in all cases; however, foci of antral metaplasia seen in eight cases showed no staining. All the CA-19.9-negative tumors (13/51) were high grade on histology and this was statistically significant (P<0.05). All cases of CC and XGC showed apical and diffuse cytoplasmic positivity for CA-19.9; however, foci of antral metaplasia were negative [Figure - 4]. Discussion GBC is the fifth commonest gastrointestinal malignancy and the most common biliary tract malignancy worldwide. It is two to four times more common in females than in males and has a peak age of presentation in the sixth to eighth decades. The pathogenesis of GBC is multi-factorial. Gallstones, female sex, genetic, racial and environmental factors have been implicated. Studies suggest that epithelial changes like metaplasia and dysplasia of GB epithelium and associated conditions like CC and XGC are the possible precursor lesions for the development of GBC. [13],[14] We found an associated focal to diffuse XGC in 16% of GBC. An association of XGC with GBC has been variably reported to be 3-29% in different studies. [3],[4],[5],[6] XGC may mimic GBC at ultrasonography and grossly at surgery and often extends to adjacent structures. Epithelial dysplasia and metaplasia have been described both in inflammatory and malignant GB; however, the incidence of dysplasia in GBC is reported to be higher. [11] We found dysplasia in the adjoining GB epithelium in 74% of GBC and in only 28% of non-malignant GB disease. Antral metaplasia and intestinal metaplasia were seen in 16% and 29% GBC, respectively. The prevalence of antral metaplasia was higher (53%) in CC including XGC. The prevalence of intestinal metaplasia in CC including XGC was however comparable to GBC (26.4%). Roa et al. observed metaplasia, dysplasia, and carcinoma in situ (CIS) in the mucosa adjacent to the cancer in 66%, 81% and 69% cases, respectively. [14] Albores-Saavedra et al., found low and high-grade dysplasia in GBC in 41% and 79% cases, respectively, and in CC in 13.5% and 3.5% cases, respectively. [11] In a study of 400 consecutive GBs, Mukopadhyay et al., found antral metaplasia, intestinal metaplasia and dysplasia in 238 (59.5%), 39 (9.8%), and 20 (5%) cases, respectively. [13] p53, a tumor suppressor gene, plays an important role in the genesis of many cancers. The reported incidence of p53 over-expression in GBC is variable (20-92%). [12],[15],[16],[17],[18],[19],[20] We observed significant p53 positivity (score >3) in 55% (28/51) GBC. Roa et al. in a study of 157 primary GBC, found p53 over-expression in 45% tumors. [18] A previous Indian study has reported it to be 70%. [19] There are few reports on the expression of p53 in precursor lesions of GBC. [12],[15],[16],[17] We found p53 over-expression in dysplasia associated with both GBC and CC. In dysplasia associated with GBC, p53 over-expression was seen in 17% of low-grade dysplasia and 40% of high-grade dysplasia. The dysplasia associated with CC showed p53 over-expression in 14%. Normal GB and metaplastic and non-metaplastic epithelium in malignant and non-malignant GB including XGC was negative for p53. Takada et al., found no abnormality in XGC as compared to GBC in polymerase chain reaction amplification studies for mutation of p53. [8] They suggested that the etio-pathologic factors of XGC might have relation with cancer, but XGC itself may not be the direct cause for cancer. Wistuba et al., found p53 over-expression in 44.7% CIS and 32.4% dysplasia in GBC and observed no expression in normal and metaplastic epithelium. [12] Wee et al., observed p53 over-expression in 28% while Kamel et al., observed it in 25% dysplasia in GBC. [16],[17] Although an occasional study has reported significant correlation between p53 over-expression and presence of gallstones, tumor stage, tumor grade and liver invasion, we observed no significant correlation between p53 over-expression in GBC and clinical parameters including age, gender, histological type and tumor grade, depth of invasion and tumor stage. [19] Other studies have also shown similar results. [12],[15],[21] There is limited information on the expression patterns of CEA in GBC. [22],[23],[24] In our study, apical to focal cytoplasmic CEA expression was seen in 55% of GBC, 27% of CC and 10% of XGC. However, diffuse cytoplasmic expression of CEA was seen only in GBC (14/51, 27.5%). Albores Saavedra et al., using both rabbit and goat antibodies to analyze CEA expression in GBs, also observed diffuse cytoplasmic staining in GBC as compared to apical staining in normal GB epithelium. [22] Serum CA-19.9 is employed as a tumor marker to discriminate patients of GBC from patients with gallstones and no cancer. [25] The expression of CA-19.9 protein in GBC has not been studied. We observed apical and cytoplasmic expression of CA-19.9 in all normal GB, CC, XGC and 75% GBC. All CA-19.9-negative GBC were of high tumor grade on histology (P<0.05). CA-19.9 expression was not seen in the foci of antral metaplasia associated with both GBC and non-malignant GB. The absence of CA-19.9 expression in the foci of antral metaplasia in malignant and non-malignant GBs and also in poorly differentiated GBC may be explained by the loss of biliary differentiation in these lesions. We observed a high incidence of p53 over-expression in GBC and in dysplasia associated with GBC as well as CC, suggesting that dysplasia is an important and early event in the pathogenesis and development of GBC. Moreover, the increasing incidence of p53 over-expression from dysplasia to invasive tumor provides further support to the view that this may be the usual route for the development of invasive GBC. In an earlier publication, we have found higher genomic instability in FHIT (fragile histidine triad) gene in the form of MSI (micro-satellite instability) and LOH (loss of heterozygosity) in both XGC and CC and no instability in normal GB; we hypothesized that both XGC and CC may be steps in the etiopathogenesis of GBC. [26] Our current findings suggest that though epithelial metaplasia and XG inflammation are often associated with GBC they do not appear to play a role in its pathogenesis through the p53 pathway. Acknowledgment We acknowledge the technical support of Lab technician Mr. Vishwakarma for performing immunohistochemistry. References
Copyright 2010 - Journal of Postgraduate Medicine The following images related to this document are available:Photo images[jp10078f3.jpg] [jp10078f2.jpg] [jp10078f1.jpg] [jp10078t1.jpg] [jp10078f4.jpg] |
|