Context.-Endocrine pancreatic tumors (EPTs) are rare lesions with varying biological behavior. Establishing malignancy is a challenge for clinicians and pathologists.
Objective.-To establish the role of proliferative, apoptotic, angiogenic, and hormonal markers as predictors of malignancy in EPTs.
Design.-Paraffin-embedded EPT samples were studied for prognostic markers.
Patients.-Twenty-one consecutive patients with a diagnosis of ept.
Main Outcome Measures.-The proliferative fraction (topoisomerase II[alpha]), microvascular density (CD34), vascular endothelial growth factor expression, and estrogen receptor-beta (ER[beta]) expression were studied by immunohistochemistry on all EPTs. Apoptosis was also assessed with terminal deoxynucleotidyl transferase nick-end labeling.
Results.-We identified 13 benign and 8 malignant tumors. Topoisomerase II[alpha] was significantly increased in malignant tumors (P = .001), while there were no differences in apoptosis, microvascular density, or vascular endothelial growth factor expression in association with malignancy. No correlation could be identified between microvascular density and vascular endothelial growth factor expression, and ER[beta] was not detected. A receiver operating characteristic curve for topoisomerase Ha disclosed that above a labeling index of 13, the test had 88% sensitivity and 100% specificity for predicting malignancy.
Conclusion.-Cellular proliferation measured with topoisomerase II[alpha] is a simple prognostic marker for malignancy in EPTs, unlike apoptosis, angiogenesis, or the presence of ER[beta], which were not associated with malignant behavior. These findings designate a defined field for future research on endocrine pancreatic carcinogenesis and a possible target for chemotherapeutic agents.
(Arch Pathol Lab Med. 2004;128:426-429)
Endocrine pancreatic tumors (EPTs) constitute a group of neoplasias derived from Langerhans islets. They represent 1% to 2% of all pancreatic tumors,1 mainly affect adults, and present with a wide variety of clinical pictures.2 Although EPTs may have a slow growth pattern and benign course, malignancy occurs in 5% to 10% of insulinomas and in up to 60% of gastrinomas. The majority of the remaining functional tumors may also behave in a malignant fashion. The nonfunctional tumors have a frequency of malignancy as high as 90%.2,3
Invasion of adjacent organs and presence of distant metastases are definite criteria in establishing malignancy in EPTs, since classic morphologic features, such as nuclear hyperchromasia and pleomorphism, prominent nucleoli, infiltrative margins, perineural infiltration, and lymph or blood vessel invasion, may be absent even in metastasizing tumors.2,4,5 Also, these features appear in only one third to one half of all EPTs,5 and metastases may become apparent many months or even years after the removal of a humor limited to the pancreas, requiring prolonged clinical follow-up to be certain of its benign nature.2
Several predictive markers for malignancy and/or survival have been investigated, with inconsistent results. Cellular proliferation assessed by different methods has been identified previously as a prognostic factor in EPTs. Accordingly, malignancy and poor survival were associated with expression of Ki-67 in all EPTs and with topoisomerase II[alpha] (topo II[alpha]) in nonfunctional tumors.6-7 On the other hand, it has been suggested that apoptosis suppression is necessary for endocrine tumorigenesis,8,9 although its association with a malignant behavior has not been evaluated.
Angiogenesis and vascular endothelial growth factor (VEGF) have been recognized as necessary for tumor growth and invasion.10 Experimentally, these factors are indispensable for progression to malignancy in pancreatic islet cell tumors."11,12 Expression of VEGF has been shown in EPTs already; however, its prognostic significance has not been widely assessed." Similarly, sex-hormone receptors have been identified in pancreatic islet cells and tumors,14 and a prognostic role for the progesterone receptor in EPTs has been investigated, with uncertain results.6,15 The recently cloned human [beta]-estrogen receptor was proposed as a marker of good prognosis in some tumors,16,17 although its utility in EPTs has not been analyzed.
The aim of the present study was to investigate the association between the biological behavior of EPTs and markers of proliferative activity, apoptosis, and angiogenesis, as well as estrogen receptor [beta] (ER[beta]) expression.
MATERIALS AND METHODS
Twenty-one cases with a confirmed diagnosis of EPT were studied. Tumor samples were selected from the surgical pathology files when containing representative tissue for immunohistochemistry. Demographic and clinical data (laboratory data, imaging findings, surgical and histopathology reports, evolution, and outcome) for all 21 patients were obtained from the hospital records, when available.
Tumors were classified as functional when clinical symptoms related to hormone overproduction associated with serum elevation and immunohistochemical detection of the corresponding hormone were documented, and as nonfunctional when these conditions were absent, regardless of the presence of immuno-staining for any hormone. Malignancy was established when invasion into neighboring organs or distant metastases was demonstrated by imaging techniques, surgery, and/or pathologic examination.
Four-micrometer sections were cut from each fixed tumor sample and stored at -20°C until use. All staining procedures were performed according to the avidin-biotin complex method. Briefly, samples were deparaffinized in xylol and rehydrated in sequential alcohol baths. Antigen retrieval was performed with a microwave pressure cooker using 10mM citrate buffer at pH 6.0 (CD34 and VEGF) and 10mM EDTA buffer at pH 8.0 (ER[beta] and topo II[alpha]). Slides were then allowed to cool at room temperature and were stained using the avidin-biotin-peroxidase complex detection method (DAB basic detection kit, Ventana Medical Systems Inc, Tucson, Ariz) in a Ventana NexES automated stainer.
The following primary antibodies were used for immunohistochemistry: mouse monoclonal antibody raised against human topo II[alpha] (Dako, Glostrup, Denmark; dilution 1:50), CD34 (Dako; dilution 1:50), VEGF (R & D Systems, Minneapolis, Minn; dilution 1:100), and ER[beta] (Genetex, Inc, San Antonio, Tex; dilution 1: 20). For topo II[alpha] and ER[beta], a labeling index was obtained according to the average number of tumor cells with evident nuclear staining in 10 high-power fields (X40). In the angiogenesis analysis, each section was examined under low power to identify intratumoral microvascular "hot spots"; sections were then observed at X40 magnification. Structures with the morphologic features of microvessels that stained with the chromogen, irrespective of whether a lumen was present, were counted. The average from 10 microscopic fields was considered the microvessel-labeling index. The intensity of cytoplasmic VEGF staining within the tumor was scored semiquantitatively as absent (0), weak (1), moderate (2), or strong (3), and the proportion of cells staining was noted using a high-power field (X40). Positive controls for topo II[alpha], CD34, and VEGF staining were tonsillar tissue, while endometrium was used for ER[beta]. Cases stained without the primary antibodies were used as negative controls to rule out nonspecific staining. Immunohistochemical analysis was performed by an experienced pathologist (A.G.D.) who had no knowledge of the patients' clinical data.
Apoptotic cells were identified by in situ detection of DNA fragmentation, using the terminal deoxynucleotidyl transferase nick-end labeling (TUNEL) method. Staining was performed on deparaffinized slides following incubation with protease for 15 minutes at 37°C, washing with Tris, and quenching of endogenous peroxidase by incubation in 3.0% (vol/vol) hydrogen peroxide in phosphate-buffered saline for 5 minutes. Single-stranded DNA was identified using a commercially available kit (ApopTag Plus; Intergen Co, New York, NY), according to the manufacturer's protocol, and the tissue was counterstained with hematoxylin for 1 minute. Sections were dehydrated in ethanol, cleared in xylene, and mounted with glass coverslips. The apoptotic labeling index was defined as the average number of apoptotic cells in 10 random fields (X40).
Data are expressed as medians and ranges. Statistical comparisons were made using the Mann-Whitney U test for continuous variables and the Fisher exact test or x^sup 2^ analysis for categorical variables. For significant associations identified in the univariate analysis, receiver operating characteristic (ROC) curves were performed to determine the ideal cutoff value and operational characteristics of the test. Correlation coefficient and Spearman rank correlation tests were applied, as appropriate, to test dependence between variables. Differences were considered significant at P
Demography and Biological Behavior
Eight tumors occurred in men and 13 were diagnosed in women, with a median age at diagnosis of 45 years (range, 17-70 years). Thirteen of the EPTs were benign and 8 were malignant, with no difference regarding age or sex. Twelve of the tumors were functional (9 insulinomas and 3 gastrinomas), and 9 were nonfunctional. Malignancy occurred in 1 of the functional tumors (insulinoma) and in 7 of the nonfunctional tumors (P = .002). Two benign insulinomas were classified as multiple endocrine neoplasia type I (MEN-1). Four patients with malignant EPTs died (2 during hospitalization, and 2 more 3 and 8 months after discharge); the remaining 4 patients abandoned outpatient follow-up after 5 to 25 months. Eleven cases with benign EPTs were followed for at least 2 years without evidence of tumor recurrence (the remaining patients were followed for 19 months and 1 month).
Results of topo II[alpha], CD34, and apoptosis (TUNEL) studies are shown in Table 1. No immunostaining was observed for ER[beta]. The only marker associated with malignancy was topo II[alpha] (Figure). Semiquantitative expression of VEGF is demonstrated in Table 2. There was no association for VEGF, even when neoplasias with strong immunostaining were compared with the remaining tumors (P = .13).
No correlation between microvascular labeling index (CD34) and VEGF expression was observed (r^sub s^ = 0.25; P = .28). However, when only benign EPTs were analyzed, a moderate correlation was found (r^sub s^ = 0.53; P = .05). On the other hand, topo II[alpha] and apoptosis were not related (r = 0.05; P = .82), even when comparing benign and malign tumors separately.
The ROC curve analysis of the topo II[alpha] labeling index demonstrated an area under the curve of 0.947 (95% confidence interval, 0.754-0.992). When a labeling index of 13 was established as the ideal cutoff value, a sensitivity value of 88% and specificity of 100% were found, with positive and negative predictive values of 100% and 93%, respectively. The only functional tumor found to be malignant had a topo II[alpha] labeling index of 22.
Establishment of malignancy in EPTs is a difficult endeavor when tumor infiltration or metastases are not found, because apparently benign tumors can show malignant behavior after months or even years of observation.2 A marker capable of distinguishing malignancy would anticipate the need for close observation and early adjuvant chemotherapy, improving survival. Tumor development is considered a multistep process, involving multiple molecular abnormalities and leading to the transition from a normal to a malignant cellular state.18 Some of these abnormalities include proliferative, apoptotic, and angiogenic pathways, as well as alterations in sex-hormone signaling.14,18 In the present study, we tested 5 molecular markers involved in tumorigenesis, as predictors of malignancy, in a group of 21 patients with EPTs. All but 1 of the benign cases were followed for at least 19 months without evidence of recurrence; thus, the possibility of being malignant and misclassified is remote.
The DNA topoisomerases are necessary for normal cell functioning by controlling DNA conformation, replication, recombination, and/or other transcriptional events. The [alpha] isoform of topoisomerase II predominates mainly in proliferating cells, making it a marker of cell proliferation19 that correlates with aggressive clinical behavior in other neoplasms.19,20 In this study, the topo II[alpha]-labeling index was associated with malignant behavior in EPTs, which is in agreement with other studies in which the tumoral growth fraction was assessed using proliferating cell nuclear antigen and/or Ki-67.5,6,21,22 Similar findings were also reported in a previous study using antibodies to topo II[alpha]; however, only nonfunctional tumors were studied, excluding the functional counterparts that are actually more common in clinical practice.4,23,24 The current work found that topo II[alpha] was also able to distinguish malignancy in functional EPTs. Furthermore, this marker showed high sensitivity and specificity at a labeling index greater than 13, which allows us to propose it as a useful marker for evaluating the tumoral growth fraction and to predict malignancy in EPTs.
Suppression of apoptosis is considered essential for tumor development and/or progression.9,18 Experimental studies have demonstrated the crucial role of antiapoptotic (Bcl-2, Bcl-x^sub L^) and proliferative/apoptotic (c-Myc) signals in islet cell tumorigenesis.9 Also, expression of Bcl-2 has been reported in 45% of EPTs,25 and induction of apoptosis has been observed in tumor biopsies of patients treated with somatostatin analogs.26 However, the decreased TUNEL-labeling index in malignant EPTs did not reach statistical significance, nor was any correlation found between apoptosis and cellular proliferation. Our results do not support apoptosis as a key factor for malignancy in EPTs, as it has been shown in other neoplasias,27 and an independent alteration on cellular proliferation seems to trigger the malignant behavior of these tumors.
The role of angiogenesis and angiogenic factors has been widely studied in models of pancreatic [beta]-cell carcinogenesis. Transformation of premalignant lesions into solid and invasive tumors requires an angiogenic switch in which VEGF plays a pivotal role, although the precise molecular mechanisms involved in the process are poorly understood.11,12,28
Microvascular density did not differ between benign and malignant tumors. Results for VEGF were similar, although the semiquantitative analysis found strong immunostaining only in malignant tumors. To our knowledge, microvascular density had not been assessed in EPTs on clinical grounds, and the lack of association between this parameter and tumor behavior contrasts with what has been demonstrated in other neoplasms (ie, breast and prostate).29,30 Terris et al13 found VEGF expression in 16 of 20 EPTs, without correlation with tumor stage.13 Taken together, these findings suggest that angiogenesis and VEGF expression are not as crucial for tumor invasion in EPTs, and other factors of malignancy may be more important. Also, this is in agreement with the recent finding of the role of VEGF on a well-characterized model of [beta]cell carcinogenesis, in which it was responsible for early angiogenesis, progression from adenoma to carcinoma, and accelerated tumor growth, but did not increase tumor invasion or metastasis. These authors concluded that VEGF-mediated angiogenesis is required, but not sufficient, for the progression to tumor malignancy.31 These findings are especially noteworthy in EPTs, in which the main definition of malignancy derives from invasion and metastasis, thus precluding the role of vascular density and VEGF as early markers of malignancy. Other factors of malignancy, such as lymphangiogenesis and VEGF-C, cannot be discarded.32
Correlation between microvascular density and VEGF was found only in benign tumors. This association can be due to the dependence of early angiogenesis on VEGF (the angiogenic switch) that may be lost as carcinogenesis progresses, reflecting the less critical role of VEGF on an already initiated process,31 the need for other angiogenic factors (ie, matrix metalloproteinases), or the loss of antiangiogenic regulators.12,33
Finally, we found no ER[beta] immunostaining in any EPT, excluding its utility as a marker of benign behavior, as suggested in breast, ovarian, and colon cancer,16,17,34 or its possible role as a regulator of apoptosis and angiogenesis, according to some proapoptotic and antiangiogenic properties previously attributed to this receptor.35,36
In conclusion, topo II[alpha] is a potentially useful marker of malignancy that can help the clinician to formulate a more specific approach for the search for metastasis and that can facilitate the early administration of adjuvant chemotherapy and/or somatostatin analogs. Furthermore, the cellular content of this enzyme can be used to identify the sensitivity to topo II-targeted chemotherapeutic agents.37 The identification of cellular proliferation as the only factor associated with malignancy designates a specific field for more in-depth studies on endocrine pancreatic carcinogenesis.
1. Gamboa-Dominguez A. Patologia y ultraestructura de los tumores endocrine. In: Herrera MF, Uscanga LF, Robles-Diaz G, Campuzano-Fernandez M, eds. Pancreas. Mexico City, Mexico: McGraw-Hill Interamericana; 2000:439-448.
2. Modlin IM, Schmid SW, Tang LH, Farhadi J, Buchler M. Endocrine tumors of the pancreas. In: Dervenis CG, Bassi C. Pancreatic Tumors: Achievements and Prospective. Stuttgart, Germany: Georg Thieme Verlag; 2000:332-353.
3. Canto Jairala JA, Herrera MF, Gamboa-Dominguez A, et al. Endocrine tumors of the pancreas at a Mexican institution. Rev Invest Clin. 1997;49:25-30.
4. Kenny BD, Sloan JM, Hamilton PW, Watt PC, Johnston CF, Buchanan KD. The role of morphometry in predicting prognosis in pancreatic islet cell tumors. Cancer. 1989;64:460-465.
5. Pelosi G, Zamboni G, Doglioni C, et al. Immunodetection of proliferating cell nuclear antigen assesses the growth fraction and predicts malignancy in endocrine tumors of the pancreas. Am J Surg Pathol. 1992;16:1215-1225.
6. Pelosi G, Bresaola E, Bogina G, et al. Endocrine tumors of the pancreas: Ki67 immunoreactivity on paraffin sections is an independent predictor for malignancy: a comparative study with proliferating-cell nuclear antigen and progesterone receptor protein immunostaining, mitotic index, and other clinicopathologic variables. Hum Pathol. 1996;27:1124-1134.
7. Chang HJ, Batts KP, Lloyd RV, et al. Prognostic significance of p27, Ki-67, and topoisomerase IIa expression in clinically nonfunctioning pancreatic endocrine tumors. Endocr Pathol. 2000;11:229-241.
8. Hager JH, Hanahan D. Tumor cells utilize multiple pathways to down-modulate apoptosis: lessons from a mouse model of islet cell carcinogenesis. Ann N Y Acad Sci. 1999;887:150-163.
9. Pelengaris S, Khan M. Oncogenic co-operation in beta-cell tumorigenesis. Endocr Relat Cancer. 2001;8:307-314.
10. Folkman J. Seminars in Medicine of Beth Israel Hospital, Boston: clinical applications of research on angiogenesis. N Engl J Med. 1995;333:1757-1763.
11. Bergers G, Javaherian K, Lo KM, Folkman J, Hanahan D. Effects of angiogenesis inhibitors on multistage carcinogenesis in mice. Science. 1999;284:808812.
12. Christofori G, Naik P, Hanahan D. Vascular endothelial growth factor and its receptors, ftl-1 and flk-1, are expressed in normal pancreatic islets and throughout islet cell tumorigenesis. Mol Endocrinol. 1995;9:1760-1770.
13. Terris B, Scoazec JY, Rubbia L, et al. Expression of vascular endothelial growth factor in digestive endocrine tumors. Histopathology. 1998;32:133-138.
14. Robles-Diaz G, Duarte-Rojo A. Pancreas: a sex steroid-dependent tissue. Isr Med Assoc J. 2001;3:364-368.
15. Viale G, Doglioni C, Gambacorta M, Zamboni G, Coggi G, Bordi C. Progesterone receptor immunoreactivity in pancreatic endocrine tumors. Cancer. 1992;70:2268-2277.
16. Foley EF, Jazaeri AA, Shupnik MA, Jazaeri O, Rice LW. Selective loss of estrogen beta in malignant colon. Cancer Res. 2000;60:245-248.
17. Iwao K, Miyoshi Y, Egawa C, Ikeda N, Nogushi S. Quantitative analysis of estrogen receptor-beta mRNA and its variants in human breast. Int J Cancer. 2000; 88:733-736.
18. Compagni A, Christofori G. Recent advances in research on multistage tumorigenesis. Br J Cancer. 2000;83:1-5.
19. Lee A, LiVolsi VA, Baloch ZW. Expression of DNA topoisomerase IIalpha in thyroid neoplasia. Mod Pathol. 2000;13:396-400.
20. Dingemans AM, Witlox MA, Stallaert RA, van der Valk P, Postmus PE, Giaccone G. Expression of DNA topoisomerase IIalfa and topoisomerase IIbeta genes predict survival and response to chemotherapy in patients with small cell lung cancer. Clin Cancer Res. 1999;5:2048-2058.
21. Clarke MR, Baker EE, Weyant RJ, Hill L, Carfy SE. Proliferative activity in pancreatic endocrine tumors: association with function, metastases and survival. Endocr Pathol. 1997;8:181-187.
22. Jorda M, Ghorab Z, Fernandez G, Nassiri M, Hanly A, Nadji M. Low nuclear proliferative activity is associated with nonmetastatic islet cell tumors. Arch Pathol Lab Med. 2003;127:196-199.
23. Heitz PU, Kasper M, Polak JM, Kloppel G. Pancreatic endocrine tumors. Hum Pathol. 1982;13:263-271.
24. Broughan TA, Leslie JD, Soto JM, Hermann RE. Pancreatic islet cell tumors. Surgery. 1986;99:671-678.
25. Wang DG, Johnston CF, Buchanan KD. Oncogene expression in gastroenteropancreatic neuroendocrine tumors. Cancer. 1997;80:668-675.
26. Eriksson B, Oberg K. Summing up 15 years of somatostatin analog therapy in neuroendocrine tumors: future outlook. Ann Oncol. 1999;10(suppl 2):S31S38.
27. Tannapfel A, Geissler F, Kockerling F, Katalinic A, Hauss J, Wittekind C. Apoptosis and proliferation in relation to histopathological variables and prognosis in hepatocellular carcinoma. J Pathol. 1999;187:439-445.
28. Inoue M, Hager JM, Ferrara N, Gerber HP, Hanahan D. VEGF-A has a critical, nonredundant role in angiogenic switching and pancreatic cell carcinogenesis. Cancer Cell. 2002;1:193-202.
29. Hansen S, Overgaard J, Rose C, et al. Independent prognostic value of angiogenesis and the level of plasminogen activator inhibitor type 1 in breast cancer patients. Br J Cancer. 2003;88:102-108.
30. Mehta R, Kyshtoobayeva A, Kurosaki T, et al. Independent association of angiogenesis index with outcome in prostate cancer. Clin Cancer Res. 2001;7: 81-88.
31. Gannon G, Mandriota SJ, Cui L, Baetens D, Pepper MS, Christofori G. Overexpression of vascular endothelial growth factor-A165 enhances tumor angiogenesis but not metastasis during beta-cell carcinogenesis. Cancer Res. 2002; 62:603-608.
32. Mandriota SJ, Jussila L, Jeltsch M, et al. Vascular endothelial growth factorC-mediated lymphangiogenesis promotes tumour metastasis. EMBO J. 2001;20: 672-682.
33. Dumortier J, Ratineau C, Roche C, Lombard-Bohas C, Chayvialle JA, Scoazec JY. Angiogenesis and endocrine tumors. Bull Cancer. 1999;86:148-153.
34. Rutherford R, Brown WD, Sapi E, Aschkenazi S, Munoz A, Mor G. Absence of estrogen receptor-beta expression in metastatic ovarian cancer. Obstet Gynecol. 2000;96:417-421.
35. Qiu Y, Waters CE, Lewis AE, Langman MJ, Eggo MC. Oestrogen-induced apoptosis in colonocytes expressing oestrogen receptor beta. J Endocrinol. 2002; 174:369-377.
36. Nilsson S, Kuiper G, Gustafsson JA. ERbeta: a novel estrogen receptor offers the potential for new drug development. Trends Endocrinol Metab. 1998;9: 387-395.
37. Yabuki N, Sasano H, Kato K, et al. Immunohistochemical study of DNA topoisomerase II in human gastric disorders. Am J Pathol. 1996;149:997-1007.
Jose Luis Diaz-Rubio, MD; Andres Duarte-Rojo, MD; Milena Saqui-Salces, Eng; Armando Camboa-Dominguez, MD, MSci; Guillermo Robles-Diaz, MD
Accepted for publication November 24, 2003.
From the Departments of Gastroenterology (Drs Diaz-Rubio, DuarteRojo, and Robles-Diaz) and Pathology (Drs Saqui-Salces and GamboaDominguez), Instituto Nacional de Ciencias Medicas y Nutricion "Salvador Zubiran," Mexico City, Mexico.
The authors have no relevant financial interest in the products or companies described in this article.
Reprints: Guillermo Robles-Diaz, MD, Department of Gastroenterology, Instituto Nacional de Ciencias Medicas y Nutricion "Salvador Zubiran," Vasco de Quiroga 15, Delegacion Tlalpan, Mexico, DF 14000, Mexico (e-mail: firstname.lastname@example.org).
Copyright College of American Pathologists Apr 2004
Provided by ProQuest Information and Learning Company. All rights Reserved