J Gastrointest Surg DOI 10.1007/s11605-014-2480-x
ORIGINAL ARTICLE
Prognostic Significance of Neuroendocrine Differentiation in Colorectal Adenocarcinoma After Radical Operation: a Meta-analysis Yu-Jie Zeng & Wei Lai & Lu Liu & Heng Wu & Xing-Xi Luo & Jie Wang & Zhong-Hua Chu
Received: 14 October 2013 / Accepted: 6 February 2014 # 2014 The Society for Surgery of the Alimentary Tract
Abstract Background The phenomenon of neuroendocrine differentiation has been observed in colorectal adenocarcinoma. However, the ability of neuroendocrine differentiation to predict the outcome of colorectal adenocarcinoma remains controversial. Methods We conducted an extensive search of research studies related to neuroendocrine differentiation using scientific databases, including the PubMed, Embase, OVID, BIOSIS Previews, and Cochrane Central Register of Controlled Trials (up to July, 2013), according to the established search terms. RevMan version 5.2 statistical program was used to analyze the data. An odds ratio (OR) with a 95 % confidence interval (CI) was used for the dichotomous data. Results Eleven studies with a total of 1,587 patients were included. Patients with neuroendocrine differentiation who underwent a radical operation had a lower 5-year survival rate (pooled OR 0.60, 95 % CI 0.37–0.97) compared with those without neuroendocrine differentiation, with evidence of moderate heterogeneity (I2 =37 %, p=0.10). A sensitivity analysis and metaregression showed that the different classification criteria of neuroendocrine differentiation used in these studies were the main source of heterogeneity. When the strong positive rates of neuroendocrine differentiation indicators between the higher (stage III + IV) and the lower (stage I + II) clinical stages were compared, the pooled OR was 1.84 (703 patients; 95 % CI 0.98–3.43) without evidence of heterogeneity (I2 =0 %, p=0.89). However, comparisons between consecutive stages showed different ORs: stage II vs. I (203 patients; OR=0.52, 95 % CI 0.17–1.56), stage III vs. II (569 patients; OR=2.27, 95 % CI 1.03–4.98), and stage IV vs. III (375 patients; OR=1.81, 95 % CI 1.00–3.29). Conclusion The patients with strong positive indicators of neuroendocrine differentiation had a lower 5-year survival rate. The ability to detect neuroendocrine indicators using conventional methods could improve the prognosis judgment of colorectal adenocarcinoma. Keywords Neuroendocrine differentiation . Colorectal adenocarcinoma . Prognosis . Meta-analysis
Introduction The phenomenon of neuroendocrine differentiation has been observed in neoplasia of several non-neuroendocrine organs, Yu-Jie Zeng and Wei Lai contribute equally to this work. Y.|t| 0.37 0.721 0.68 0.516 −2.02 0.083 0.56 0.593
[95 % confidence interval] −1.453219 1.996322 −1.394744 2.530068 −4.78393 0.3742877 −1.241141 2.012466
Regression equation ψ=α+βy+γz+ε(_cons) was built to define the grading criteria of bleeding events. x, y, and z are independent variables. α, β, γ, and ε are coefficient variations that are associated with the four classification criteria—classification 1 to 4. Assigned (1,0,0), (0,1,0), (0,0,1), or (0,0,0) to (x,y,z), the equation presents the four classification criteria separately. The estimated between-study variance (tau2 ) was reduced from 0.23 to 0.001 without the influence of classification 3
J Gastrointest Surg
Fig. 3 Forest plots of indicator strong positive vs. indicator negative on the 5-year survival rate in the colorectal adenocarcinoma by subgroup analysis according to the different classifications
three studies (Fig. 4a). The pooled OR was 1.84 (703 patients; 95 % CI 0.98–3.43) with no evidence of heterogeneity (I2 =0 %, p=0.89). Although the 95% CI of the OR was mainly located on the part of the axis favoring the higher stages, there was no statistical significance for strong positive rate of indicator between the higher and lower clinical TNM stages. Comparisons of strong positive indicator rate (group III) between consecutive individual stages resulted in the following data: stage II vs. I (203 patients; OR=0.52, 95 % CI 0.17– 1.56), stage III vs. II (569 patients; OR=2.27, 95 % CI 1.03– 4.98), and stage IV vs. III (375 patients; OR=1.81, 95 % CI 1.00–3.29) (Fig. 4b–d). Though only the OR for stage III vs. stage II was statistically significant, the higher clinical TNM stage seems to show a higher positive rate of neuroendocrine differentiation indicator. Publication Bias Funnel plots were used to estimate the publication bias of the meta-analysis. As shown in Fig. 5, the shape of the funnel plot did not reveal an obvious asymmetry.
Discussion The phenomenon of neuroendocrine differentiation is common in adenocarcinoma tissues. For example, neuroendocrine differentiation has been extensively investigated in prostate cancer and shown to be correlated with disease progression and poor prognosis.22 However, the ability of neuroendocrine differentiation to predict long-term outcome in colorectal adenocarcinoma is still debatable. To our knowledge, this question has not been discussed in any published study to date. The diagnosis for neuroendocrine differentiation depends primarily on the positive rate of the immunohistochemical indicator—CgA, which has been recognized as the most specific indicator of neuroendocrine tissue.23 To date, there is no uniform criterion to diagnose and classify neuroendocrine differentiation in colorectal adenocarcinoma. However, we have deduced a criterion from the published studies. First, there was approximately 2 % of the neuroendocrine cells in normal colorectal epithelium cells.24 Second, it was shown that the
J Gastrointest Surg
Fig. 4 a Forest plots of combined TNM stage III + IV vs. I + II on the indicator strong positive rate. b Forest plots of combined TNM stage II vs. I on the indicator strong positive rate. c Forest plots of combined TNM
stage III vs. II on the indicator strong positive rate. d Forest plots of combined TNM stage IV vs. III on the indicator strong positive rate
mixed adenoneuroendocrine carcinoma (MANEC), which was first named by 2010 WHO classification of tumors of the digestive system,25 consisted of both the adenocarcinoma and neuroendocrine carcinoma cells, and the proportion of each was higher than 30 %. Therefore, we propose that the definition of neuroendocrine differentiation in colorectal adenocarcinoma is the proportion of positive neuroendocrine differentiation indicators ranging from 2 to 30 %. We chose studies that included patients who underwent radical operation but no preoperative radiotherapy or chemotherapy, in order to avoid the impact of advanced disease on 5year survival. Although there could be a selection bias, the
purpose of our study was to discuss the patients who underwent radical operation. None of the selected studies mentioned any recurrent disease or further treatment, such as repeated surgery or palliative chemotherapy, and only two studies mentioned post-operation adjuvant chemotherapy. Therefore, we assumed the same disease progression in all studies. The purpose of this meta-analysis was to compare the 5year survival rates of patients with or without neuroendocrine differentiation who underwent colorectal adenocarcinoma radical operation. The strong positive indicator group was named group III. There were four different classification criteria in these studies, and each classification was
J Gastrointest Surg
Fig. 5 Funnel plot analysis of publication bias for included studies
divided into three groups, as shown in Table 1. Six studies considered at least 1 positive cell (stained with CgA)/mm2 in a colorectal cancer specimen as strong positive neuroendocrine differentiation (classification 1), while three studies considered at least 2 % positive cells (stained with CgA) in a colorectal cancer specimen as strong positive neuroendocrine differentiation (classification 2). The other two studies considered at least 5 positive cells (stained with CgA)/mm2 and 1–20 % positive cells (stained with CgA) in the colorectal cancer specimen as strong positive neuroendocrine differentiation (classification 3 and classification 4, respectively). The combined results from the 11 studies showed that the presence of neuroendocrine differentiation was significantly associated with a lower 5-year survival rate in colorectal adenocarcinoma. Based on these findings, further research is necessary to determine how neuroendocrine differentiation affects the prognosis. Because there was limited data and only limited significance of group II in all of the classifications, we did not compare the 5-year survival rates between group II and group I. By performing subgroup analysis, sensitivity analysis, and meta-regression, we found that the study “Gen-you You 2003,”18 which was the only study in classification 3, was the main source of heterogeneity. Because classification 3, which contained only one study, had been shown to be highly heterogeneous compared with other classifications, we removed this classification and found no heterogeneity (I2 =0 %, p=0.64) for the remaining studies. The pooled OR of the overall effect was 0.50 (95 % CI 0.34–0.72, tau2 =0.001, p=0.0003). The criterion for strong positive differentiation of classification 1 was “more than 1 positive cells/mm2 of colorectal cancer was stained with CgA” and for classification 2 was “more than 2 % immunoreactive cells.” These two classifications are appropriate for the definition we proposed earlier. In addition, it was found that the I2 for the studies included was 0 %, which suggested that these two classifications of neuroendocrine differentiation did not
result in evident heterogeneity. Therefore, we considered classification 1 and classification 2 as the superior clinical choices for use, and our results are based on these two classifications. The clinical TNM stage is the principal standard for estimating prognosis in colorectal adenocarcinoma. Our results seem to show that the higher clinical TNM stage may correlate with the higher positive rate of neuroendocrine differentiation indicator. This correlation might explain the results of our meta-analysis. However, only a comparison of tumors with the same TNM stage with and without neuroendocrine differentiation can confirm this conclusion. The studies included in this meta-analysis were not based on random control trials. Based on the NOS scores of the included studies, their quality was considered good. Although there were four classifications of neuroendocrine differentiation in these studies, most of the studies belonged to classification 1 and classification 2. Moreover, the shape of the funnel plot did not reveal any obvious asymmetry. This meta-analysis is a promising study that provides evidence that the neuroendocrine differentiation in colorectal adenocarcinoma is associated with a poor 5-year survival rate. Somatostatin analogues are widely used in patients with neuroendocrine tumor, who suffer from carcinoid syndrome. Carcinoid syndrome is due to the active substance secreted by neuroendocrine cells.26 Furthermore, the somatostatin analogues have been shown to significantly increase the time to tumor progression compared with a placebo in patients with functionally active and inactive metastatic midgut neuroendocrine tumors.27 Therefore, the application of somatostatin analogues in colorectal adenocarcinoma with neuroendocrine differentiation could be an entirely new area of research.
Conclusion In conclusion, this meta-analysis is the first report to discuss the impact of neuroendocrine differentiation on the 5-year survival rate of patients with colorectal adenocarcinoma. We showed that a strong positive CgA in colorectal adenocarcinoma is associated with a lower 5-year survival rate compared with the negative CgA group. We showed that detecting the neuroendocrine indicators conventionally could potentially improve the prognosis judgment of colorectal adenocarcinoma. We believe that an authoritative and uniform criterion for the classification of neuroendocrine differentiation can be developed and accepted all over the world, which is based on the existing literature. Investigations of the exact mechanism of how neuroendocrine differentiation impacts the prognosis of colorectal adenocarcinoma could create new avenues for research.
J Gastrointest Surg Acknowledgments assistance.
We are grateful to Zonghua Li for statistical 13.
Funding This study was supported by grants from the Science and Technology Project of Guangdong Province (no. 2012B050600014), Natural Science Fund of Guangdong Province (no. S2012010009161), and National Natural Foundation of China (no. 81001306).
14.
15.
References 1. Matei D V, Renne G, Pimentel M, Sandri M T, Zorzino L, Botteri E, De Cicco C, Musi G, Brescia A, Mazzoleni F, Tringali V, Detti S, de Cobelli O. Neuroendocrine differentiation in castration-resistant prostate cancer: a systematic diagnostic attempt[J]. Clin Genitourin Cancer. 2012, 10(3): 164–173. 2. Cho Y B, Yang S S, Lee W Y, Song S Y, Kim S H, Shin H J, Yun S H, Chun H K. The clinical significance of neuroendocrine differentiation in T3-T4 node-negative colorectal cancer[J]. Int J Surg Pathol. 2010, 18(3): 201–206. 3. Travis W D. Lung tumours with neuroendocrine differentiation[J]. Eur J Cancer. 2009, 45 Suppl 1: 251–266. 4. Miremadi A, Pinder S E, Lee A H, Bell J A, Paish E C, Wencyk P, Elston C W, Nicholson R I, Blamey R W, Robertson J F, Ellis I O. Neuroendocrine differentiation and prognosis in breast adenocarcinoma[J]. Histopathology. 2002, 40(3): 215–222. 5. Staren E D, Gould V E, Jansson D S, Hyser M, Gooch G T, Economou S G. Neuroendocrine differentiation in “poorly differentiated” colon carcinomas[J]. Am Surg. 1990, 56(7): 412–419. 6. Grabowski P, Schindler I, Anagnostopoulos I, Foss H D, Riecken E O, Mansmann U, Stein H, Berger G, Buhr H J, Scherubl H. Neuroendocrine differentiation is a relevant prognostic factor in stage III–IV colorectal cancer[J]. Eur J Gastroenterol Hepatol. 2001, 13(4): 405–411. 7. de Bruine A P, Wiggers T, Beek C, Volovics A, von Meyenfeldt M, Arends J W, Bosman F T. Endocrine cells in colorectal adenocarcinomas: incidence, hormone profile and prognostic relevance[J]. Int J Cancer. 1993, 54(5): 765–771. 8. Gulubova M, Vlaykova T. Chromogranin A-, serotonin-, synaptophysin- and vascular endothelial growth factor-positive endocrine cells and the prognosis of colorectal cancer: an immunohistochemical and ultrastructural study[J]. J Gastroenterol Hepatol. 2008, 23(10): 1574–1585. 9. Grabowski P, Schonfelder J, Ahnert-Hilger G, Foss H D, Stein H, Berger G, Zeitz M, Scherubl H. Heterogeneous expression of neuroendocrine marker proteins in human undifferentiated carcinoma of the colon and rectum[J]. Ann N Y Acad Sci. 2004, 1014: 270–274. 10. Hamada Y, Oishi A, Shoji T, Takada H, Yamamura M, Hioki K, Yamamoto M. Endocrine cells and prognosis in patients with colorectal carcinoma[J]. Cancer. 1992, 69(11): 2641–2646. 11. Indinnimeo M, Cicchini C, Memeo L, Stazi A, Provenza C, Ricci F, Mingazzini P L. Correlation between chromogranin-A expression and pathological variables in human colon carcinoma[J]. Anticancer Res. 2002, 22(1A): 395–398. 12. Shinji S, Naito Z, Ishiwata T, Tanaka N, Furukawa K, Suzuki H, Seya T, Kan H, Tsuruta H, Matsumoto S, Matsuda A, Teranishi N, Ohaki Y, Tajiri T. Neuroendocrine cell differentiation of poorly
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
differentiated colorectal adenocarcinoma correlates with liver metastasis[J]. Int J Oncol. 2006, 29(2): 357–364. Volante M, Marci V, Andrejevic-Blant S, Tavaglione V, Sculli M C, Tampellini M, Papotti M. Increased neuroendocrine cells in resected metastases compared to primary colorectal adenocarcinomas[J]. Virchows Arch. 2010, 457(5): 521–527. Mori M, Mimori K, Kamakura T, Adachi Y, Ikeda Y, Sugimachi K. Chromogranin positive cells in colorectal carcinoma and transitional mucosa[J]. J Clin Pathol. 1995, 48(8): 754–758. Ferrero S, Buffa R, Pruneri G, Siccardi A G, Pelagi M, Lee A K, Coggi G, Bosari S. The prevalence and clinical significance of chromogranin A and secretogranin II immunoreactivity in colorectal adenocarcinomas[J]. Virchows Arch. 1995, 426(6): 587–592. Foley E F, Gaffey M J, Frierson H J. The frequency and clinical significance of neuroendocrine cells within stage III adenocarcinomas of the colon[J]. Arch Pathol Lab Med. 1998, 122(10): 912– 914. Lloyd R V, Schroeder G, Bauman M D, Krook J E, Jin L, Goldberg R M, Farr G J. Prevalence and Prognostic Significance of Neuroendocrine Differentiation in Colorectal Carcinomas[J]. Endocr Pathol. 1998, 9(1): 35–42. Yao G Y, Zhou J L, Lai M D, Chen X Q, Chen P H. Neuroendocrine markers in adenocarcinomas: an investigation of 356 cases[J]. World J Gastroenterol. 2003, 9(4): 858–861. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in metaanalyses[J]. Eur J Epidemiol. 2010, 25(9): 603–605. Ospina P A, Nydam D V, Diciccio T J. Technical note: The risk ratio, an alternative to the odds ratio for estimating the association between multiple risk factors and a dichotomous outcome[J]. J Dairy Sci. 2012, 95(5): 2576–2584. Salim A, Mackinnon A, Griffiths K. Sensitivity analysis of intentionto-treat estimates when withdrawals are related to unobserved compliance status[J]. Stat Med. 2008, 27(8): 1164–1179. Sagnak L, Topaloglu H, Ozok U, Ersoy H. Prognostic significance of neuroendocrine differentiation in prostate adenocarcinoma[J]. Clin Genitourin Cancer. 2011, 9(2): 73–80. Kloppel G. Classification and pathology of gastroenteropancreatic neuroendocrine neoplasms[J]. Endocr Relat Cancer. 2011, 18 Suppl 1: S1–S16. Lewin K J. The endocrine cells of the gastrointestinal tract. The normal endocrine cells and their hyperplasias. Part I[J]. Pathol Annu. 1986, 21 Pt 1: 1–27. Bosman Ft C F H R. WHO classification of tumours of the digestive system. 4th edition. Lyon: International Agency for Research on Cancer[Z]. 2010. Oberg K, Kvols L, Caplin M, Delle F G, de Herder W, Rindi G, Ruszniewski P, Woltering E A, Wiedenmann B. Consensus report on the use of somatostatin analogs for the management of neuroendocrine tumors of the gastroenteropancreatic system[J]. Ann Oncol. 2004, 15(6): 966–973. Rinke A, Muller H H, Schade-Brittinger C, Klose K J, Barth P, Wied M, Mayer C, Aminossadati B, Pape U F, Blaker M, Harder J, Arnold C, Gress T, Arnold R. Placebo-controlled, double-blind, prospective, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors: a report from the PROMID Study Group[J]. J Clin Oncol. 2009, 27(28): 4656–4663.
ORIGINAL ARTICLE
Prognostic Significance of Neuroendocrine Differentiation in Colorectal Adenocarcinoma After Radical Operation: a Meta-analysis Yu-Jie Zeng & Wei Lai & Lu Liu & Heng Wu & Xing-Xi Luo & Jie Wang & Zhong-Hua Chu
Received: 14 October 2013 / Accepted: 6 February 2014 # 2014 The Society for Surgery of the Alimentary Tract
Abstract Background The phenomenon of neuroendocrine differentiation has been observed in colorectal adenocarcinoma. However, the ability of neuroendocrine differentiation to predict the outcome of colorectal adenocarcinoma remains controversial. Methods We conducted an extensive search of research studies related to neuroendocrine differentiation using scientific databases, including the PubMed, Embase, OVID, BIOSIS Previews, and Cochrane Central Register of Controlled Trials (up to July, 2013), according to the established search terms. RevMan version 5.2 statistical program was used to analyze the data. An odds ratio (OR) with a 95 % confidence interval (CI) was used for the dichotomous data. Results Eleven studies with a total of 1,587 patients were included. Patients with neuroendocrine differentiation who underwent a radical operation had a lower 5-year survival rate (pooled OR 0.60, 95 % CI 0.37–0.97) compared with those without neuroendocrine differentiation, with evidence of moderate heterogeneity (I2 =37 %, p=0.10). A sensitivity analysis and metaregression showed that the different classification criteria of neuroendocrine differentiation used in these studies were the main source of heterogeneity. When the strong positive rates of neuroendocrine differentiation indicators between the higher (stage III + IV) and the lower (stage I + II) clinical stages were compared, the pooled OR was 1.84 (703 patients; 95 % CI 0.98–3.43) without evidence of heterogeneity (I2 =0 %, p=0.89). However, comparisons between consecutive stages showed different ORs: stage II vs. I (203 patients; OR=0.52, 95 % CI 0.17–1.56), stage III vs. II (569 patients; OR=2.27, 95 % CI 1.03–4.98), and stage IV vs. III (375 patients; OR=1.81, 95 % CI 1.00–3.29). Conclusion The patients with strong positive indicators of neuroendocrine differentiation had a lower 5-year survival rate. The ability to detect neuroendocrine indicators using conventional methods could improve the prognosis judgment of colorectal adenocarcinoma. Keywords Neuroendocrine differentiation . Colorectal adenocarcinoma . Prognosis . Meta-analysis
Introduction The phenomenon of neuroendocrine differentiation has been observed in neoplasia of several non-neuroendocrine organs, Yu-Jie Zeng and Wei Lai contribute equally to this work. Y.|t| 0.37 0.721 0.68 0.516 −2.02 0.083 0.56 0.593
[95 % confidence interval] −1.453219 1.996322 −1.394744 2.530068 −4.78393 0.3742877 −1.241141 2.012466
Regression equation ψ=α+βy+γz+ε(_cons) was built to define the grading criteria of bleeding events. x, y, and z are independent variables. α, β, γ, and ε are coefficient variations that are associated with the four classification criteria—classification 1 to 4. Assigned (1,0,0), (0,1,0), (0,0,1), or (0,0,0) to (x,y,z), the equation presents the four classification criteria separately. The estimated between-study variance (tau2 ) was reduced from 0.23 to 0.001 without the influence of classification 3
J Gastrointest Surg
Fig. 3 Forest plots of indicator strong positive vs. indicator negative on the 5-year survival rate in the colorectal adenocarcinoma by subgroup analysis according to the different classifications
three studies (Fig. 4a). The pooled OR was 1.84 (703 patients; 95 % CI 0.98–3.43) with no evidence of heterogeneity (I2 =0 %, p=0.89). Although the 95% CI of the OR was mainly located on the part of the axis favoring the higher stages, there was no statistical significance for strong positive rate of indicator between the higher and lower clinical TNM stages. Comparisons of strong positive indicator rate (group III) between consecutive individual stages resulted in the following data: stage II vs. I (203 patients; OR=0.52, 95 % CI 0.17– 1.56), stage III vs. II (569 patients; OR=2.27, 95 % CI 1.03– 4.98), and stage IV vs. III (375 patients; OR=1.81, 95 % CI 1.00–3.29) (Fig. 4b–d). Though only the OR for stage III vs. stage II was statistically significant, the higher clinical TNM stage seems to show a higher positive rate of neuroendocrine differentiation indicator. Publication Bias Funnel plots were used to estimate the publication bias of the meta-analysis. As shown in Fig. 5, the shape of the funnel plot did not reveal an obvious asymmetry.
Discussion The phenomenon of neuroendocrine differentiation is common in adenocarcinoma tissues. For example, neuroendocrine differentiation has been extensively investigated in prostate cancer and shown to be correlated with disease progression and poor prognosis.22 However, the ability of neuroendocrine differentiation to predict long-term outcome in colorectal adenocarcinoma is still debatable. To our knowledge, this question has not been discussed in any published study to date. The diagnosis for neuroendocrine differentiation depends primarily on the positive rate of the immunohistochemical indicator—CgA, which has been recognized as the most specific indicator of neuroendocrine tissue.23 To date, there is no uniform criterion to diagnose and classify neuroendocrine differentiation in colorectal adenocarcinoma. However, we have deduced a criterion from the published studies. First, there was approximately 2 % of the neuroendocrine cells in normal colorectal epithelium cells.24 Second, it was shown that the
J Gastrointest Surg
Fig. 4 a Forest plots of combined TNM stage III + IV vs. I + II on the indicator strong positive rate. b Forest plots of combined TNM stage II vs. I on the indicator strong positive rate. c Forest plots of combined TNM
stage III vs. II on the indicator strong positive rate. d Forest plots of combined TNM stage IV vs. III on the indicator strong positive rate
mixed adenoneuroendocrine carcinoma (MANEC), which was first named by 2010 WHO classification of tumors of the digestive system,25 consisted of both the adenocarcinoma and neuroendocrine carcinoma cells, and the proportion of each was higher than 30 %. Therefore, we propose that the definition of neuroendocrine differentiation in colorectal adenocarcinoma is the proportion of positive neuroendocrine differentiation indicators ranging from 2 to 30 %. We chose studies that included patients who underwent radical operation but no preoperative radiotherapy or chemotherapy, in order to avoid the impact of advanced disease on 5year survival. Although there could be a selection bias, the
purpose of our study was to discuss the patients who underwent radical operation. None of the selected studies mentioned any recurrent disease or further treatment, such as repeated surgery or palliative chemotherapy, and only two studies mentioned post-operation adjuvant chemotherapy. Therefore, we assumed the same disease progression in all studies. The purpose of this meta-analysis was to compare the 5year survival rates of patients with or without neuroendocrine differentiation who underwent colorectal adenocarcinoma radical operation. The strong positive indicator group was named group III. There were four different classification criteria in these studies, and each classification was
J Gastrointest Surg
Fig. 5 Funnel plot analysis of publication bias for included studies
divided into three groups, as shown in Table 1. Six studies considered at least 1 positive cell (stained with CgA)/mm2 in a colorectal cancer specimen as strong positive neuroendocrine differentiation (classification 1), while three studies considered at least 2 % positive cells (stained with CgA) in a colorectal cancer specimen as strong positive neuroendocrine differentiation (classification 2). The other two studies considered at least 5 positive cells (stained with CgA)/mm2 and 1–20 % positive cells (stained with CgA) in the colorectal cancer specimen as strong positive neuroendocrine differentiation (classification 3 and classification 4, respectively). The combined results from the 11 studies showed that the presence of neuroendocrine differentiation was significantly associated with a lower 5-year survival rate in colorectal adenocarcinoma. Based on these findings, further research is necessary to determine how neuroendocrine differentiation affects the prognosis. Because there was limited data and only limited significance of group II in all of the classifications, we did not compare the 5-year survival rates between group II and group I. By performing subgroup analysis, sensitivity analysis, and meta-regression, we found that the study “Gen-you You 2003,”18 which was the only study in classification 3, was the main source of heterogeneity. Because classification 3, which contained only one study, had been shown to be highly heterogeneous compared with other classifications, we removed this classification and found no heterogeneity (I2 =0 %, p=0.64) for the remaining studies. The pooled OR of the overall effect was 0.50 (95 % CI 0.34–0.72, tau2 =0.001, p=0.0003). The criterion for strong positive differentiation of classification 1 was “more than 1 positive cells/mm2 of colorectal cancer was stained with CgA” and for classification 2 was “more than 2 % immunoreactive cells.” These two classifications are appropriate for the definition we proposed earlier. In addition, it was found that the I2 for the studies included was 0 %, which suggested that these two classifications of neuroendocrine differentiation did not
result in evident heterogeneity. Therefore, we considered classification 1 and classification 2 as the superior clinical choices for use, and our results are based on these two classifications. The clinical TNM stage is the principal standard for estimating prognosis in colorectal adenocarcinoma. Our results seem to show that the higher clinical TNM stage may correlate with the higher positive rate of neuroendocrine differentiation indicator. This correlation might explain the results of our meta-analysis. However, only a comparison of tumors with the same TNM stage with and without neuroendocrine differentiation can confirm this conclusion. The studies included in this meta-analysis were not based on random control trials. Based on the NOS scores of the included studies, their quality was considered good. Although there were four classifications of neuroendocrine differentiation in these studies, most of the studies belonged to classification 1 and classification 2. Moreover, the shape of the funnel plot did not reveal any obvious asymmetry. This meta-analysis is a promising study that provides evidence that the neuroendocrine differentiation in colorectal adenocarcinoma is associated with a poor 5-year survival rate. Somatostatin analogues are widely used in patients with neuroendocrine tumor, who suffer from carcinoid syndrome. Carcinoid syndrome is due to the active substance secreted by neuroendocrine cells.26 Furthermore, the somatostatin analogues have been shown to significantly increase the time to tumor progression compared with a placebo in patients with functionally active and inactive metastatic midgut neuroendocrine tumors.27 Therefore, the application of somatostatin analogues in colorectal adenocarcinoma with neuroendocrine differentiation could be an entirely new area of research.
Conclusion In conclusion, this meta-analysis is the first report to discuss the impact of neuroendocrine differentiation on the 5-year survival rate of patients with colorectal adenocarcinoma. We showed that a strong positive CgA in colorectal adenocarcinoma is associated with a lower 5-year survival rate compared with the negative CgA group. We showed that detecting the neuroendocrine indicators conventionally could potentially improve the prognosis judgment of colorectal adenocarcinoma. We believe that an authoritative and uniform criterion for the classification of neuroendocrine differentiation can be developed and accepted all over the world, which is based on the existing literature. Investigations of the exact mechanism of how neuroendocrine differentiation impacts the prognosis of colorectal adenocarcinoma could create new avenues for research.
J Gastrointest Surg Acknowledgments assistance.
We are grateful to Zonghua Li for statistical 13.
Funding This study was supported by grants from the Science and Technology Project of Guangdong Province (no. 2012B050600014), Natural Science Fund of Guangdong Province (no. S2012010009161), and National Natural Foundation of China (no. 81001306).
14.
15.
References 1. Matei D V, Renne G, Pimentel M, Sandri M T, Zorzino L, Botteri E, De Cicco C, Musi G, Brescia A, Mazzoleni F, Tringali V, Detti S, de Cobelli O. Neuroendocrine differentiation in castration-resistant prostate cancer: a systematic diagnostic attempt[J]. Clin Genitourin Cancer. 2012, 10(3): 164–173. 2. Cho Y B, Yang S S, Lee W Y, Song S Y, Kim S H, Shin H J, Yun S H, Chun H K. The clinical significance of neuroendocrine differentiation in T3-T4 node-negative colorectal cancer[J]. Int J Surg Pathol. 2010, 18(3): 201–206. 3. Travis W D. Lung tumours with neuroendocrine differentiation[J]. Eur J Cancer. 2009, 45 Suppl 1: 251–266. 4. Miremadi A, Pinder S E, Lee A H, Bell J A, Paish E C, Wencyk P, Elston C W, Nicholson R I, Blamey R W, Robertson J F, Ellis I O. Neuroendocrine differentiation and prognosis in breast adenocarcinoma[J]. Histopathology. 2002, 40(3): 215–222. 5. Staren E D, Gould V E, Jansson D S, Hyser M, Gooch G T, Economou S G. Neuroendocrine differentiation in “poorly differentiated” colon carcinomas[J]. Am Surg. 1990, 56(7): 412–419. 6. Grabowski P, Schindler I, Anagnostopoulos I, Foss H D, Riecken E O, Mansmann U, Stein H, Berger G, Buhr H J, Scherubl H. Neuroendocrine differentiation is a relevant prognostic factor in stage III–IV colorectal cancer[J]. Eur J Gastroenterol Hepatol. 2001, 13(4): 405–411. 7. de Bruine A P, Wiggers T, Beek C, Volovics A, von Meyenfeldt M, Arends J W, Bosman F T. Endocrine cells in colorectal adenocarcinomas: incidence, hormone profile and prognostic relevance[J]. Int J Cancer. 1993, 54(5): 765–771. 8. Gulubova M, Vlaykova T. Chromogranin A-, serotonin-, synaptophysin- and vascular endothelial growth factor-positive endocrine cells and the prognosis of colorectal cancer: an immunohistochemical and ultrastructural study[J]. J Gastroenterol Hepatol. 2008, 23(10): 1574–1585. 9. Grabowski P, Schonfelder J, Ahnert-Hilger G, Foss H D, Stein H, Berger G, Zeitz M, Scherubl H. Heterogeneous expression of neuroendocrine marker proteins in human undifferentiated carcinoma of the colon and rectum[J]. Ann N Y Acad Sci. 2004, 1014: 270–274. 10. Hamada Y, Oishi A, Shoji T, Takada H, Yamamura M, Hioki K, Yamamoto M. Endocrine cells and prognosis in patients with colorectal carcinoma[J]. Cancer. 1992, 69(11): 2641–2646. 11. Indinnimeo M, Cicchini C, Memeo L, Stazi A, Provenza C, Ricci F, Mingazzini P L. Correlation between chromogranin-A expression and pathological variables in human colon carcinoma[J]. Anticancer Res. 2002, 22(1A): 395–398. 12. Shinji S, Naito Z, Ishiwata T, Tanaka N, Furukawa K, Suzuki H, Seya T, Kan H, Tsuruta H, Matsumoto S, Matsuda A, Teranishi N, Ohaki Y, Tajiri T. Neuroendocrine cell differentiation of poorly
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
differentiated colorectal adenocarcinoma correlates with liver metastasis[J]. Int J Oncol. 2006, 29(2): 357–364. Volante M, Marci V, Andrejevic-Blant S, Tavaglione V, Sculli M C, Tampellini M, Papotti M. Increased neuroendocrine cells in resected metastases compared to primary colorectal adenocarcinomas[J]. Virchows Arch. 2010, 457(5): 521–527. Mori M, Mimori K, Kamakura T, Adachi Y, Ikeda Y, Sugimachi K. Chromogranin positive cells in colorectal carcinoma and transitional mucosa[J]. J Clin Pathol. 1995, 48(8): 754–758. Ferrero S, Buffa R, Pruneri G, Siccardi A G, Pelagi M, Lee A K, Coggi G, Bosari S. The prevalence and clinical significance of chromogranin A and secretogranin II immunoreactivity in colorectal adenocarcinomas[J]. Virchows Arch. 1995, 426(6): 587–592. Foley E F, Gaffey M J, Frierson H J. The frequency and clinical significance of neuroendocrine cells within stage III adenocarcinomas of the colon[J]. Arch Pathol Lab Med. 1998, 122(10): 912– 914. Lloyd R V, Schroeder G, Bauman M D, Krook J E, Jin L, Goldberg R M, Farr G J. Prevalence and Prognostic Significance of Neuroendocrine Differentiation in Colorectal Carcinomas[J]. Endocr Pathol. 1998, 9(1): 35–42. Yao G Y, Zhou J L, Lai M D, Chen X Q, Chen P H. Neuroendocrine markers in adenocarcinomas: an investigation of 356 cases[J]. World J Gastroenterol. 2003, 9(4): 858–861. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in metaanalyses[J]. Eur J Epidemiol. 2010, 25(9): 603–605. Ospina P A, Nydam D V, Diciccio T J. Technical note: The risk ratio, an alternative to the odds ratio for estimating the association between multiple risk factors and a dichotomous outcome[J]. J Dairy Sci. 2012, 95(5): 2576–2584. Salim A, Mackinnon A, Griffiths K. Sensitivity analysis of intentionto-treat estimates when withdrawals are related to unobserved compliance status[J]. Stat Med. 2008, 27(8): 1164–1179. Sagnak L, Topaloglu H, Ozok U, Ersoy H. Prognostic significance of neuroendocrine differentiation in prostate adenocarcinoma[J]. Clin Genitourin Cancer. 2011, 9(2): 73–80. Kloppel G. Classification and pathology of gastroenteropancreatic neuroendocrine neoplasms[J]. Endocr Relat Cancer. 2011, 18 Suppl 1: S1–S16. Lewin K J. The endocrine cells of the gastrointestinal tract. The normal endocrine cells and their hyperplasias. Part I[J]. Pathol Annu. 1986, 21 Pt 1: 1–27. Bosman Ft C F H R. WHO classification of tumours of the digestive system. 4th edition. Lyon: International Agency for Research on Cancer[Z]. 2010. Oberg K, Kvols L, Caplin M, Delle F G, de Herder W, Rindi G, Ruszniewski P, Woltering E A, Wiedenmann B. Consensus report on the use of somatostatin analogs for the management of neuroendocrine tumors of the gastroenteropancreatic system[J]. Ann Oncol. 2004, 15(6): 966–973. Rinke A, Muller H H, Schade-Brittinger C, Klose K J, Barth P, Wied M, Mayer C, Aminossadati B, Pape U F, Blaker M, Harder J, Arnold C, Gress T, Arnold R. Placebo-controlled, double-blind, prospective, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors: a report from the PROMID Study Group[J]. J Clin Oncol. 2009, 27(28): 4656–4663.
