APMIS 123: 502–508

© 2015 APMIS. Published by John Wiley & Sons Ltd. DOI 10.1111/apm.12379

Colorectal adenocarcinoma with mucinous component: relation of MMP-13, EGFR, and E-cadherin expressions to clinicopathological features and prognosis ABD AL-RAHMAN MOHAMMAD FODA, AMIRA KAMAL EL-HAWARY and AZZA ABDEL AZIZ Pathology Department, Faculty of Medicine, Mansoura University, Mansoura, Egypt

Foda AA-RM, El-Hawary AK, Aziz AA. Colorectal adenocarcinoma with mucinous component: relation of MMP-13, EGFR and E-cadherin expressions to clinicopathological features and prognosis. APMIS 2015; 123: 502–508. The aim of this study was to compare colorectal adenocarcinoma with mucinous component, ordinary adenocarcinoma (OA) and mucinous adenocarcinoma (MA) regarding clinicopathological parameters, survival, EGFR, MMP-13, and E-cadherin. We studied tumor tissue specimens from 28 patients with adenocarcinoma with mucinous component, 47 with OA, and 56 with MA, who underwent radical surgery from January 2007 to January 2012 at the Gastroenterology Centre, Mansoura University, Egypt. High density manual tissue microarrays were constructed and immunohistochemistry for EGFR, MMP-13, and E-cadherin was done. Colorectal adenocarcinoma with mucinous component (AWMC) was significantly associated with more perineural invasion, lower EGFR, and MMP-13 expressions than OA, with no difference in E-cadherin expression. Conversely, only microscopic abscess formation was significantly more with colorectal AWMC than MC with no difference in EGFR, MMP-13 and E-cadherin expression between both groups. Colorectal AWMC showed a better survival than MA with no difference with OA. In a univariate analysis, EGFR, MMP-13, and E-cadherin expressions did not show a significant impact on disease-free or overall survival in patients with colorectal AWMC. Colorectal AWMC remains a vague entity that resembles OA in some clinicopathological and molecular respects as well as MA. Key words: Colorectal adenocarcinoma with mucinous component; MMP-13; EGFR; E-cadherin. Amira Kamal El-Hawary, Pathology Department, Faculty of Medicine, Mansoura University, El-Dakahlia, Mansoura 35511, Egypt. e-mail: [email protected]

Colorectal adenocarcinoma with mucinous component (AWMC) is a histological subtype of ordinary adenocarcinoma (OA) that shows intra or extracellular mucin secretion 50% of the tumor (2). Many studies had examined the pathological features of MA and OA. However, only few prior studies focused on comparing AWMC and MA (3). Some claimed that MA and AWMC do not differ from OA with respect to outcome and histological predictors of outcome (4). In contrast, others reported that AWMC and MA have distinct clinicopathological features compared with OA (3). A number of immunohistochemical markers have been investigated in CRC and in some cases, these Received 21 October 2014. Accepted 1 February 2015

502

antibodies showed differential rates of expression among different morphologic types of carcinoma. MMP-13 plays an important role in the tumor development and progression, and thus becomes a useful indicator for clinical assessment of biological behavior and prognosis in CRC patients (5). EGFR and E-cadherin are two of the most studied proteins involved in CRC promotion, progression and treatment (6). To the best of our knowledge, the clinicopathological features of AWMC have not been thoroughly studied in relation to OA and MA. Moreover, MMP-13, EGFR, and E-cadherin expressions hadn’t been investigated in colorectal AWMC. In this study, we aimed to compare AWMC, OA, and MA regarding clinicopathological, histological parameters, survival, EGFR (as a marker of proliferation), MMP-13 (as a marker

MMP13, EGFR, E-CADHERIN IN COLORECTAL ADENOCARCINOMA

of invasion), and E-cadherin (as a marker of adhesion). MATERIAL AND METHODS Cases This retrospective study was carried out in surgical pathology lab at Gastroenterology Center, Mansoura, Egypt. Files of all resected CRC cases were revised during the period from January 2007 to December 2011 and cases with mucinous activity were selected and revised (84 cases). They included 28 cases with AWMC (mucinous component 50% of the tumor). Another 47 cases of OA (without mucinous component) were chosen randomly for comparison from the same period. Cases with incomplete clinical data and those that were composed completely of pools of mucin with very few epithelial cells were excluded.

Clinical parameters and histopathological evaluation All clinicopathological data of these 131 cases were revised with re-examination of all their slides. This included: age, gender, location, size, shape, multiplicity, histological type, grade, depth of invasion (T), tumor margins, lymphovascular invasion, extramural perineural invasion, peri- and intra-tumoral lymphocytic infiltration, extent of neutrophilic infiltrate, nearby and distant mucosa, whether the tumor is on top of adenoma or not, number of lymph node (LN) metastases (N), distant metastasis (M), TNM staging, state of surgical cut margins, associated schistosomiasis, exact site of the ova and any other finding.

Tissue microarray (TMA) construction Three manual TMA blocks were constructed using modified mechanical pencil tip method as previously described by Foda (7). Three representative cores of 0.8 mm diameter were punched from each case (one near the surface of the tumor, another at the center of the tumor and the last at the advancing edge of the tumor). The cores were punched from the most representative areas considering tumor heterogeneity from mucinous and non-mucinous areas. Sections from TMA blocks were prepared (4 lm thickness) for routine H&E. Other sections were prepared on charged slides for immunohistochemistry.

Immunohistochemistry After trimming, TMA blocks were sectioned with a microtome to obtain sections of 4 lm thickness. Deparaffinized sections were incubated for 30 min with 0.3% hydrogen peroxide in methanol and microwave heated for 30 min in EDTA buffer solution, pH 8.0. Subsequently, an indirect immunoperoxidase technique was applied, using antibodies against monoclonal mouse antihuman MMP-13 (Clone Ab1, Ready-to-Use, Cat. #MS825-R7), monoclonal mouse anti-human EGFR Ab-10 (Clone 111.6, Cat. #MS-378-R7, 7.0 mL, Ready-to-Use © 2015 APMIS. Published by John Wiley & Sons Ltd

for Immunohistochemistry) and monoclonal mouse antihuman E-Cadherin Ab-3 (Clone 36B5, Cat. #MS-1479R7, 7.0 mL, Ready-to-Use for Immunohistochemistry), all from Thermo Fisher Scientific, Lab Vision Corporation (Fremont, CA, USA). Primary antibody was left to react for 30 min at room temperature. Immunoperoxidase method was performed using ImmunoPure UltraSensitive ABC Peroxidase (catalog no. 32052; Thermo Scientific, Rockford, IL, USA), using diaminobenzidine as chromogen.

Evaluation of immunohistochemistry MMP-13, EGFR and E-cadherin and expressions were semi-quantitatively assessed for each case according to previously published methods. Both intensity and percentage of positive cells were considered. Cytoplasmic staining was considered positive for MMP-13. Cytoplasmic staining was considered positive as well as membranous staining for EGFR (8, 9) and E-cadherin (10, 11). Immunoreactivity was evaluated independently by the three authors (Foda AA, Abdel Aziz A and El-Hawary AK) who were blinded regarding the classification and prognosis of the tumors. The proximity of the scores between the three observers was >95%. In the remaining cases in which discrepancies had been noted, the differences were resolved by reevaluation and discussion to reach consensus agreements. Cytoplasmic MMP-13 staining intensity was graded as 0 (negative), 1 (weak), 2 (moderate), and 3 (strong), and the percentage of positive-stained cells was graded as 0 (50%) for each core (12). Intensity of both EGFR and E-cadherin staining was graded as: 0 (negative), 1 (weak), 2 (moderate), and 3 (strong), and the percentage of positive-stained cells was graded as: 0 (0%), 1 (1–10%), 2 (11–50%), 3 (>50%) for each core (8–13). For all immune stains, the final score for each core was determined by the combining intensity and percent scores (0–6). The mean of the three cores for each case was calculated. For data analysis, tumors were classified as negative (score 0), low expression group (score 1–3), and high expression group (score 4–6).

Statistical analysis Data were analyzed, applying SPSS, version 16.0 for Windows (SPSS Inc, IBM, Chicago, IL, USA). Chi-square (v2) test was used to test significant differences between AWMC cases and other groups in clinicopathological and histological parameters, as well as in MMP-13, EGFR and E-cadherin expressions. Survival data were analyzed using Kaplan–Meier test. A comparison of survival curves was carried out using the log-rank test. A two-tailed p ≤ 0.05 was considered significant in all tests.

RESULTS In total, 131 CRC cases were analyzed. Ages ranged from 20 to 78 years (mean, 53.44 years). The participants were 81 men (61.8%) and 50 women

503

MMP13, EGFR, E-CADHERIN IN COLORECTAL ADENOCARCINOMA

but comparable to OA it was nearly the same. Most studies reported that MA was associated with poor prognosis (23–26), while others found no correlation between this histological subtype and clinical outcome (27–30). However, better survival of AWMC in relation to that of MA can be explained by association of AWMC with some good prognostic factors as positive microscopic abscess formation. There was a lot of debate about the interobserver agreement of the distinction between AWMC and MA. Depending on survival assays, the cut-off value of 50% mucinous component of the tumor was proposed. Our results confirmed that the reproducibility of the distinction between these two entities is sufficient and AWMC showed a significantly better prognosis than MA. However, this distinction has to be further validated on the molecular level. Consistent with our results, Ogino et al. (31) observed differences in frequencies of K-ras mutations and O-6-methylguanine DNA methyl-transferase (MGMT) according to the degree of mucinous component. K-ras mutations were detected more frequently in AWMC than in OA, and loss of MGMT was observed more frequently in AWMC compared with MA. Kaur et al. (32) found a significant association between abnormal mismatch repair (MMR) proteins expression and mucinous, signet ring and poorly differentiated CRC in contrast to OA cases which showed normal MMR proteins expression. The expression of MMR proteins should also be studied in AWMC to further enrich the distinction between these CRC subtypes. In conclusion, AWMC remains a vague entity that resembles OA in some clinicopathological and molecular respects as well as MA. Further molecular studies are needed to group AWMC with either OA or MA or to be considered as a distinct subtype. Validation of the significance of the cut-off value of 50% mucinous component of the tumor is also to be considered, depending on molecular and prognostic factors.

REFERENCES 1. Hamilton SR, Bosman FT, Boffetta P, Ilyas M, Morreau H, Nakamura S-I, et al. Tumors of the colon and rectum. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, editors. WHO Classification of Tumors of the Digestive System. Lyon: IARC Press, 2010:132–82. 2. Hyngstrom J, Hu CY, Xing Y, You YN, Feig BW, Skibber JM, et al. Clinicopathology and outcomes for mucinous and signet ring colorectal adenocarcinoma: analysis from the National Cancer data base. Ann Surg Oncol 2012;19:2814–21.

© 2015 APMIS. Published by John Wiley & Sons Ltd

3. Lee DW, Han SW, Lee HJ, Rhee YY, Bae JM, Cho NY, et al. Prognostic implication of mucinous histology in colorectal cancer patients treated with adjuvant FOLFOX chemotherapy. Br J Cancer 2013;108:1978– 84. 4. Langner C, Harbaum L, Pollheimer MJ, Kornprat P, Lindtner RA, Schlemmer A, et al. Mucinous differentiation in colorectal cancer–indicator of poor prognosis? Histopathology 2012;60:1060–72. 5. Yang B, Gao J, Rao Z, Shen Q. Clinicopathological significance and prognostic value of MMP-13 expression in colorectal cancer. Scand J Clin Lab Invest 2012;72:501–5. 6. Theodoropoulos GE, Karafoka E, Papailiou JG, Stamopoulos P, Zambirinis CP, Bramis K, et al. P53 and EGFR expression in colorectal cancer: a reappraisal of ‘old’ tissue markers in patients with long follow-up. Anticancer Res 2009;29:785–91. 7. Foda AA. No-cost manual method for preparation of tissue microarrays having high quality comparable to semiautomated methods. Appl Immunohistochem Mol Morphol 2013;21:271–4. 8. Bibeau F, Boissiere-Michot F, Sabourin JC, Gourgou-Bourgade S, Radal M, Penault-Llorca F, et al. Assessment of epidermal growth factor receptor (EGFR) expression in primary colorectal carcinomas and their related metastases on tissue sections and tissue microarray. Virchows Arch 2006;449:281–7. 9. Sauer T. Epidermal growth factor receptor gene and immunohistochemical expression in colorectal carcinomas. Diagn Histopathol 2007;14:94–8. 10. Khoursheed MA, Mathew TC, Makar RR, Louis S, Asfar SK, Al-Sayer HM, et al. Expression of E-cadherin in human colorectal cancer. Surgeon 2003;1:86–91. 11. Hahn-stromberg V, Edvardsson H, Bodin L, Franzen L. Disturbed expression of E-cadherin, beta-catenin and tight junction proteins in colon carcinoma is unrelated to growth pattern and genetic polymorphisms. APMIS 2008;116:253–63. 12. Foda AA, El-Hawary AK, Abdel-Aziz A. Differential expression of matrix metalloproteinase-13 in mucinous and nonmucinous colorectal carcinomas. Ann Diagn Pathol 2013;17:347–51. 13. Jie D, Zhongmin Z, Guoqing L, Sheng L, Yi Z, Jing W, et al. Positive Expression of LSD1 and negative expression of E-cadherin correlate with metastasis and poor prognosis of colon cancer. Dig Dis Sci 2013;58:1581–9. 14. Vlad IC, S ß -Cadariu PA, Irimie A, Cazacu M. Prognostic Significance of perineural invasion in patients with rectal cancer using R environment for statistical computing and graphics. Appl Med Inform 2012;31:13–20. 15. Foda AA, El-Hawary AK, Abdel Aziz A, Hosni A, Zalata KR, Gado AI. Rare mucinous colorectal adenocarcinoma: analysis of the epidemiological factors in relation to survival in Egyptian patients. Am J Cancer Epidemiol Prev 2014;2:10–9. 16. Uehara K, Nakanishi Y, Shimoda T, Taniguchi H, Akasu T, Moriya Y. Clinicopathological significance of microscopic abscess formation at the invasive margin of advanced low rectal cancer. Br J Surg 2007;94:239–43. 17. Yamaguchi T, Taniguchi H, Fujita S, Sekine S, Yamamoto S, Akasu T, et al. Clinicopathological charac-

507

MMP13, EGFR, E-CADHERIN IN COLORECTAL ADENOCARCINOMA

A

B

Fig. 1. Survival of the three groups of cases; adenocarcinoma with mucinous component, mucinous adenocarcinoma and ordinary adenocarcinoma: (A) disease-free survival, (B) overall survival.

AWMC showed significantly negative/lower expression of MMP-13 than OA (p = 0.002). Conversely, AWMC showed significantly higher expression of MMP-13 than MA (p = 0.014). AWMC showed significantly negative/lower expression of EGFR than OA (p = 0.007), but not MA (p = 0.876). On the other hand, there were no significant differences in E-cadherin expressions between AWMC, OA, and MA (p = 0.638 and p = 0.432, respectively; Table 4; Fig. 2). To clarify the prognostic impact of MMP-13, EGFR, and E-cadherin expression on survival of AWMC cases, univariate analyses were carried out. Neither MMP-13, EGFR nor E-cadherin expresTable 4. Differential expressions of MMP-13, EGFR, and E-cadherin in 28 cases of ordinary adenocarcinoma with mucinous component in relation to other groups AWMC OA MA (n = 28) (n = 47) (n = 56) n (%) n (%) n (%) MMP-13 expression Negative/low 16 (57.1) 10 (21.3) 46 (82.1) High 12 (42.9) 37 (78.7) 10 (17.9) v2 = 9.966, v2 = 6.035, p = 0.0021 p = 0.0141 EGFR expression Negative/low 12 (42.9) 7 (14.9) 25 (44.6) High 16 (57.1) 40 (85.1) 31 (55.4) v2 = 7.254, v2 = 0.024, p = 0.0071 p = 0.876 E-cadherin expression Negative/low 4 (14.3) 5 (10.6) 12 (21.4) High 24 (85.7) 42 (89.4) 44 (78.6) 2 v = 0.221, v2 = 0.618, p = 0.638 p = 0.432 AWMC, Ordinary adenocarcinoma with mucinous component; MA, Mucoid adenocarcinoma; OA, Ordinary adenocarcinoma. 1 p ≤ 0.05 is significantly different with ordinary adenocarcinoma with mucoid component group. © 2015 APMIS. Published by John Wiley & Sons Ltd

sions showed a significant impact on DFS or OS in patients with AWMC (data not shown).

DISCUSSION To the best of our knowledge, this is the first study of clinicopathological, histological features, and survival of AWMC. The frequency of this tumor subtype is currently largely unknown due to lack of consensus about its exact definition. We hypothesized that AWMC may be a specific histological subtype of CRC different from MA and OA. Our results showed that AWMC was similar to MA in some respects as well as to OA. AWMC was similar to MA and not OA in association with more perineural invasion. To the best of our knowledge, we are the first to demonstrate a significant association between mucinous histology and perineural invasion. Vlad et al. (14) demonstrated that perineural invasion positivity correlates with aggressive behavior in rectal cancer. To be added here, increasing mucin secretion by the tumor also imparts more aggressive behavior by the tumor. On the other hand, AWMC was similar to OA and not MA in association with positive microscopic abscess formation. Microscopic abscess formation was defined as the presence of debris and leukocytes (mainly neutrophils) at the invasive margin and the surface of the tumor (15), and was considered to be associated with more favorable outcome in rectal cancers (16). Yamaguchi et al. (17) reported that MA was significantly associated with negative microscopic abscess formation. Our data confirmed these findings, about 63% of MA cases in our study were associated with absence of microscopic abscess formation, which was signifi-

505

FODA et al.

A

B

C

D

Fig. 2. Ordinary adenocarcinoma with mucinous component: (A) H&E staining showing

Colorectal adenocarcinoma with mucinous component: relation of MMP-13, EGFR, and E-cadherin expressions to clinicopathological features and prognosis.

The aim of this study was to compare colorectal adenocarcinoma with mucinous component, ordinary adenocarcinoma (OA) and mucinous adenocarcinoma (MA) ...
515KB Sizes 2 Downloads 7 Views