Endoscopic resection therapies for rectal neuroendocrine tumors: A systematic review and meta-analysis

Transcription

1 bs_bs_banner doi:.1111/jgh META-ANALYSIS AND SYSTEMATIC REVIEW for rectal neuroendocrine tumors: A systematic review and meta-analysis Xin Zhou,* Haiting Xie,* Lingduo Xie,* Jing Li, Weihua Cao and Wei Fu* *Department of General Surgery, and Department of Epidemiology and Biostatistics, School of Public Health, Peking University Health Science Center, Beijing, China Key words endoscopic mucosal resection, endoscopic submucosal dissection, gastrointestinal endoscopy, rectal neuroendocrine tumor. Accepted for publication 22 August Correspondence Mr Wei Fu, Department of General Surgery, Peking University Third Hospital, Beijing 0191, China. fuwei0720@sohu.com Funding: None. Competing interests: The authors declare that they have no conflict of interest. Abstract Background and Aim: Among various endoscopic resection therapies, including conventional endoscopic mucosal resection (EMR) only with a snare after submucosal injection, modified EMR (m-emr) with other assistant devices such as a ligation band or a suction cap, and endoscopic submucosal dissection (ESD), we aimed to study which is the best choice for rectal neuroendocrine tumors. Methods: A broad literature research was performed, and a systematic review and metaanalysis were conducted. Results: Ten retrospective studies with 650 patients were included. Complete resection rates were significantly higher in the ESD group compared with the EMR group (relative risk [RR] 0.89, 95% confidence interval [CI] [0.79, 0.99]), in the m-emr group compared with the conventional EMR group (RR 0.72, 95% CI [0.60, 0.86]), and was comparable between the ESD group and the m-emr group (RR 1.03, 95% CI [0.95, 1.11]). Procedure time was significantly longer in the ESD group than in the EMR group (standard mean differences 1.37, 95% CI [ 1.99, 0.75]), but there was no significant difference between that of the m-emr group and ESD group (standard mean differences 1.50, 95% CI [ 3.14, 0.14]). Local recurrence occurred in five cases in the EMR group (5/328) and did not occur in the ESD group (0/209). Conclusions: ESD or m-emr techniques could be applied to rectal neuroendocrine tumors with indications for endoscopic treatment. m-emr procedures appear to be comparable with ESD in the treatment of rectal neuroendocrine tumors. However, the findings have to be carefully interpreted due to the lower level of evidence. Introduction In recent years, the incidence of rectal neuroendocrine tumors has increased drastically with the advances in detection modalities, especially in gastrointestinal endoscopy. 1 4 According to the European Neuroendocrine Tumor Society (ENETS) 2012 Consensus Guidelines, 5 well-differentiated rectal neuroendocrine tumors below 2 cm without muscularis invasion or lymph node involvement could be resected endoscopically. And various treatment modalities other than radical surgery have evolved, including endoscopic polypectomy, endoscopic mucosal resection (EMR), endoscopic submucosal dissection (ESD), and transanal endoscopic microsurgery (TEM). 6 9 In addition, some new techniques have been developed, which are derived from conventional EMR procedures, including EMR using a band ligation device (EMR- B), EMR with a cap-fitted panendoscope (EMR-C), and endoscopic submucosal resection with a ligation device (ESMR-L); all these new techniques share the same step of creating a pseudopedicle before resection 7,8,,11 and were classified as modified EMR (m-emr) in our study. The aim of the current systematic review and meta-analysis was to determine which endoscopic procedure is the best choice for the treatment of rectal neuroendocrine tumors. Methods Search strategy. A comprehensive database research of Pubmed, Embase, and Medline, and the Cochrane Central Library was performed to identify articles published up to February The search term combinations consisted of medical subject heading (MeSH) terms, text words, and word variants for endoscopy, EMR, ESD, rectal neuroendocrine tumors, and rectal carcinoids (see Appendix I). Only articles published in English were considered for inclusion, and reference bibliographies of all retrieved articles were searched manually for other possible studies. Journal of Gastroenterology and Hepatology 29 (2014)

2 Records identified through database searching (n = 519) Additional records identified via other sources (n = 0) Duplicates removed (n = 118) Evaluated by title and abstract (n = 401) Evaluated in details (n = 26) Records included (n = ) Removed by title and abstract (n = 375) -Studies not relevant (n = 356) -Case reports (n = 11) -Review articles (n = 7) -Letters (n = 1) Records excluded after detailed evaluation (n = 16) -Only 1 approach reported (n = 14) -Data could not be extracted (n = 2) Figure 1 Flow diagram for identification of eligible studies. Study selection. Inclusion criteria for studies were the following: (i) original article with extractable data, and (ii) methodology in which various endoscopic procedures were compared. If the same data were published twice or more, either the study of higher quality or the most recent publication was included. If primary data were not published, the authors were contacted to obtain detailed information. Case reports, review articles, editorials, and letters were excluded, as well as studies on only one endoscopic procedure. Titles and abstracts of articles identified by our search strategy were screened by two independent reviewers. If any discrepancies existed, a third reviewer was involved to make the decision. For all articles meeting the criteria, full articles were obtained and reviewed by the same two reviewers in the same way. Data extraction. Data on study characteristics, methodological quality, and outcomes were extracted by two independent reviewers according to a standard form designed in advance. A third reviewer was employed to resolve disagreements. The outcome data of interest were tumor size, complete resection rate, procedure time, postoperative bleeding, perforation, and local recurrence. Tumor size was defined as the endoscopically measured tumor size before resection. Among various criteria regarding completeness of resection, histologically complete resection was adopted, which was defined as resections in which the lateral and vertical edges of the resected specimens were free of tumor invasion. If this histological definition was not specified in the article, the complete resection criteria that were most similar to this definition were adopted. Bleeding complications were classified as immediate or delayed. Delayed postoperative bleeding was defined as hematochezia or melena in the patient after completion of endoscopic procedures that required endoscopic or radiological hemostasis or transfusions. Only delayed postoperative bleeding was analyzed. Endoscopic procedures were classified into three categories according to the description of authors in their articles: conventional EMR, m-emr, and ESD. Conventional EMR was defined as EMR only with a snare after submucosal injection. m-emr was defined as EMR with other assistant devices, such as a ligation band or a suction cap. If data on various EMR procedures were given in one study, they were merged as one in the comparison of EMR and ESD. Dichotomous variables were merged directly, while continuous variables were merged using the formula described in the Cochrane handbook. 12 Methodological quality assessment. Because all included studies were retrospective, the Newcastle-Ottawa Scale was used for methodological quality assessment by two independent reviewers. 13 A full score of nine points were used to cover the topics of this review. Statistical analysis and data synthesis. Review Manager (RevMan) software (version 5.0.; The Nordic Cochrane Centre, Copenhagen, Denmark) was used for this meta-analysis. Analyses were carried out according to the endoscopic procedures: EMR versus ESD, m-emr versus ESD, and conventional EMR versus m-emr. Relative risk or risk ratio (RR) with a 95% CI was used to compare complete resections among the various endoscopic procedures. Risk difference (RD) was used to evaluate complications or tumor recurrence because complications or recurrences may not have occurred in some groups. The standard mean differences (STD) with 95% CI were used for continuous variables. The Mantel-Haenszel model was used for pooling RR or RD, while the inverse variance model was used for weighted mean difference or standardized mean difference. Statistical heterogeneity was assessed using the I 2 statistic, and a P-value of less than 0.1 was considered statistically significant. Heterogeneity was considered to be present if the I 2 > is greater than 50%. The fixed-effect model was used if heterogeneity existed, and a random-effects model used if not. Significance was set at P = 0.05 in both models. Funnel plots were used to screen for publication bias. Asymmetry of the funnel plots were tested by Egger s regression, and a P-value of < 0.1 was considered to be significant. Results Search results and study selection. After abstract screening and full-text review, studies with a total of 650 patients were included in this meta-analysis (Fig. 1, study 260 Journal of Gastroenterology and Hepatology 29 (2014)

3 characteristics listed in Table 1) Two articles by Kim KM et al. 21 and Kim HH et al. 22 compared different endoscopic procedures in two seemingly overlapping cohorts of patients: one study was on EMR versus ESD, and the other was on EMR versus m-emr. To avoid duplication bias, the former was included when EMR was compared with ESD, while the latter was included when EMR was compared with m-emr. Various EMR procedures were compared with ESD in nine studies involving 537 patients ,23 m-emr was compared with ESD in four studies involving 188 patients, 15,19,21,23 and conventional EMR was compared with m-emr in two studies involving 120 patients. 15,22 In addition, polypectomy was reported in two studies with only 19 patients. 16,17 Results of methodological quality assessment are listed in Table 2; the methodological quality was not ideal in all included studies. All included studies were retrospective in nature due to the relative rarity of the disease. All included studies enrolled consecutive patients who were pathologically diagnosed and without lymph node involvement on computed tomography (CT) scanning or endoscopic ultrasonography (EUS). All included studies failed to state that they controlled for additional factors except for endoscopic procedures. At least two endoscopic procedure groups were compared in each study. All included studies did regular followup, as can be seen from Table 1; for some cases, follow-up time might be not long enough for recurrence to occur. Data analysis Tumor size. Tumor size was not significantly different in all comparisons among the three groups: EMR versus ESD (STD 0.01, 95% CI [ 0.20, 0.19]; P = 0.95), m-emr versus ESD ( 0.03 [ 0.50, 0.44]; P = 0.90), and EMR versus m-emr (0.17 [ 0.22, 0.57]; P = 0.39) (Fig. 2). Complete resection rate. Complete resection rate was significantly higher in the ESD group than in the EMR group (RR 0.89, 95% CI [0.79, 0.99]; P = 0.03), in the m-emr group than the conventional EMR group (RR 0.72, 95% CI [0.60, 0.86]; P = ), and was comparable between the ESD group and the m-emr group (RR 1.03, 95% CI [0.95, 1.11]; P = 0.52). In the two studies that reported outcomes of polypectomy, 16,17 complete resection was only achieved in 1/5 and /14 patients, which was far less than corresponding conventional EMR (22/26, 43/58) (Fig. 3). Data on further treatment for incompletely resected cases are shown in Table 3. Procedure time. Procedure time was significantly longer in the ESD group compared with the EMR group (STD 1.37, 95% CI [ 1.99, 0.75]; P < ). However, the difference between the m-emr group and the ESD group (STD 1.50, 95% CI [ 3.14, 0.14]; P = 0.07) did not reach statistical significance, neither did the difference between the conventional EMR group and the m-emr group (STD 0.17, 95% CI [ 0.19, 0.53]; P = 0.35) (Fig. 4). Postoperative bleeding. Postoperative bleeding occurred in two cases in the EMR group (2/328), one case in the m-emr Table 1 Characteristics of studies included in this analysis Follow-up time (month) Study type Period Country Patients Polypectomy EMR m-emr ESD Size limit (mm) No. Author Publication year 1 Onozato Y et al Retrospective Japan EMR 33.2 (13.6) ESD 80.8 (25.6) 2 Park HW et al Retrospective Korea EMR 33.0 (3 117) ESD 12.6 (6 24) 3 Lee DS et al Retrospective Korea NG EMR 22 (14 35) ESD 13 (9 38) 4 Zhou PH et al Retrospective China NG 20 EMR 42.6 (26.1) ESD 18.7 (.6) 5 Kim KM et al Retrospective Korea (6 59) 6 Choi CW et al Retrospective Korea NG 7 Kim HH et al Retrospective Korea NG 8 Zhao ZF et al Retrospective China (9.76) 9 Sung HY et al Retrospective Korea (5 7) Niimi K et al Retrospective Japan EMR 24.0 (32.5) ESD 65.1 ± 57 EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection; m-emr, modified EMR; NG, not given. Journal of Gastroenterology and Hepatology 29 (2014)

4 group (1/90), and three cases in the ESD group (3/209); the differences were not statistically significant. All bleeding cases were controlled endoscopically using electrocoagulation probes and metal clips. Perforation. Perforation occurred in two cases in the EMR group (2/328), one case in the m-emr group (1/90), and three Table 2 Methodological quality of studies included in this analysis No. Author Selection Comparability Outcome Total OnozatoY et al Park HW et al Lee DS et al Zhou PH et al Kim KM et al Choi CW et al Kim HH et al Zhao ZF et al Sung HY et al Niimi K et al cases in the ESD group (3/209); the differences were not statistically significant. All the perforated cases were managed by conservative medical treatment with or without endoscopic closure using clips. Recurrence. Local recurrence occurred in five cases in the EMR group (5/328), but did not occur in the ESD group (0/209) and m-emr group (0/90). Of the five recurrences, one case underwent additional EMR under EUS, three cases underwent salvage ESD, and one case underwent transanal surgery. Only one case, in the study of Niimi et al., 19 was found to have distant metastasis. The original lesion was 7 mm in size and was resected by complete resection. There had been no evidence of lymphovascular invasion or muscularis propria invasion. Discussion Since 1907, when Oberndorfer 24 put forward the definition of carcinoid, this term has been used to denote neuroendocrine tumors originating from a variety of tissues. For many years, carcinoids were considered to be indolent in their behavior. Although the ENETS 2012 Consensus Guidelines 5 emphasized the EMR versus ESD EMR ESD Study or Subgroup Mean SD Total Mean SD Total Weight IV, Fixed, 95% Cl Year Zhou et al % 0.24 [ 0.85, 0.36] 20 Lee et al % 0.14 [ 0.61, 0.33] 20 Onozato et al % 0.57 [ 1.34, 0.20] 20 Park et al % 0.22 [ 0.21, 0.65] 20 Choi et al % 0.45 [ 0.96, 0.06] 2012 Kim et al % 0.34 [ 0.04, 0.72] 2012 Niimi et al % 0.17 [ 0.64, 0.97] 2012 IV, Fixed, 95% Cl Total (95% Cl) % 0.01 [ 0.20, 0.19] Heterogeneity: Chi 2 =.26, df = 6 (P = 0.11); I 2 = 42% Test for overall effect: Z = 0.07 (P = 0.95) m-emr versus ESD m-mer ESD Study or Subgroup Mean SD Total Mean SD Total Weight IV, Ramdom, 95% Cl Year Choi et al % 0.45 [ 0.96, 0.06] 2012 Kim et al % 0.23 [ 0.20, 0.66] 2012 Niimi et al % 0.17 [ 0.64, 0.97] IV, Ramdom, 95% Cl Total (95% Cl) % 0.03 [ 0.50, 0.44] Heterogeneity: Tau 2 = 0.09; Chi 2 = 4.22, df = 2 (P = 0.12); I 2 = 53% Test for overall effect: Z = 0.12 (P = 0.90) EMR versus m-emr EMR m-emr Study or Subgroup Mean SD Total Mean SD Total Weight IV, Fixed, 95% Cl Kim et al % 0.17 [ 0.22, 0.57] Total (95% Cl) % 0.17 [ 0.22, 0.57] Heterogeneity: Not applicable Test for overall effect: Z = 0.86 (P = 0.39) IV, Fixed, 95% Cl Figure 2 Forest plot of tumor size by different endoscopic procedures. EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection; m-emr, modified EMR. 262 Journal of Gastroenterology and Hepatology 29 (2014)

5 EMR versus ESD Study or Subgroup Park et al. 20 Zhou et al. 20 Onozato et al. 20 Lee et al. 20 Choi et al Zhao et al Sung et al Niimi et al Kim et al EMR ESD Risk Ratio Events Total Events Total Weight M-H, Random, 95% Cl % 6.0% 6.1% 8.4% 11.2% 13.1% 9.2% 12.4% 20.2% 0.79 [0.65, 0.96] 0.53 [0.36, 0.79] 1.09 [0.74, 1.60] 0.78 [0.57, 1.06] 1.03 [0.81, 1.30] 0.92 [0.75, 1.13] 0.80 [0.60, 1.07] 1.07 [0.86, 1.33] 0.92 [0.84, 1.01] Total (95% Cl) % 0.89 [0.79, 0.99] Total events Heterogeneity: Tau 2 = 0.01, Chi 2 = 17.47, df = 8 (P = 0.03); I 2 = 54% Test for overall effect: Z = 2.14 (P = 0.03) m-emr versus ESD Study or Subgroup Choi et al Kim et al Niimi et al Zhao et al Year m-emr ESD Risk Ratio Events Total Events Total Weight M-H, Fixed, 95% Cl % 47.3% 13.2% 12.0% 1.03 [0.81, 1.30] 1.02 [0.96, 1.09] 1.07 [0.86, 1.33] 1.00 [0.83, 1.20] Risk Ratio M-H, Random, 95% Cl Risk Ratio M-H, Fixed, 95% Cl Total (95% Cl) % 1.03 [0.95, 1.11] Total events Heterogeneity: Chi 2 = 0.27, df = 3 (P = 0.97); I 2 = 0% Test for overall effect: Z = 0.65 (P = 0.52) EMR versus m-emr Study or Subgroup Kim et al Zhao et al EMR m-emr Risk Ratio Events Total Events Total Weight M-H, Fixed, 95% Cl Total (95% Cl) % 0.72 [0.60, 0.86] Total events Heterogeneity: Chi 2 = 0.50, df = 1 (P = 0.48); I 2 = 0% Test for overall effect: Z = 3.56 (P = ) % [0.57, 0.86] 0.81 [0.57, 1.14] Risk Ratio M-H, Fixed, 95% Cl Figure 3 Forest plot of complete resection by different endoscopic procedures. EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection; m-emr, modified EMR. importance of endoscopic treatment in rectal neuroendocrine tumors, neither these guidelines nor those of the North American Neuroendocrine Tumor Society 20 Consensus 25 specified which endoscopic treatment modalities should be used in practice. After conducting a meta-analysis to compare EMR with ESD in the treatment of rectal neuroendocrine tumors, Zhong et al. recommended the use of ESD. 26 However, only four studies involving 246 patients published in the same year were included in their review. In addition, in their review, conventional EMR and m-emr were lumped together in comparison with ESD without considering advances in EMR techniques. Thus, we conducted this systematic review to compare various endoscopic procedures in the treatment of rectal neuroendocrine tumors. Complete resection is crucial in guaranteeing a curative treatment for rectal neuroendocrine tumors. 5 The resection of ESD tends to be wider and deeper than EMR in practice, and in the current study ESD was comparable with m-emr in complete resection rate; this may be explained by the treatment principle. In the treatment of rectal neuroendocrine tumors, only tumors smaller than 2 cm could be treated endoscopically, and in all studies only tumors smaller than 15 mm without muscularis invasion were enrolled. m-emr was found to be superior to conventional EMR in this review, which might be due to advances in EMR techniques, such as the band ligation device and cap-fitted panendoscope. Both of these new devices permit formation of a longer pseudopedicle, which allows wider and deeper resection. Journal of Gastroenterology and Hepatology 29 (2014)

6 Table 3 Treatment after incomplete resection No. Author Incomplete resection Treatment given 1 Onozato Y et al Additional EMCT in 7 2 Park HW et al Radical surgery in 1 3 Lee DS et al Radical surgery in 3, transanal surgery in 1, follow-up in 14 4 Zhou PH et al Transanal surgery in 1, salvage ESD in 2 5 Kim KM et al. 21 TEM in 5,follow-up in 4 6 Choi CW et al Follow-up in 11 7 Kim HH et al NG 8 Zhao ZF et al NG 9 Sung HY et al Salvage EMR in 2, TEM in 2, radical surgery in 2 Niimi K et al Follow-up in 1 EMCT, endoscopic microwave coagulation therapy; EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection; TEM, transanal endoscopic microsurgery; NG, not given. EMR versus ESD EMR ESD Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% Cl Year Onozato et al % 3.36 [ 4.49, 2.23] 20 Zhou et al % 0.97 [ 1.61, 0.34] 20 Lee et al % 0.70 [ 1.18, 0.21] 20 Park et al % 2.12 [ 2.65, 1.59] 20 Kim et al % 0.24 [ 0.62, 0.14] 20 Zhao et al % 1.40 [ 2.25, 0.55] 2012 Niimi et al % 0.88 [ 1.72, 0.03] 2012 Choi et al % 1.80 [ 2.40, 1.19] 2012 IV, Random, 95% Cl Total (95% Cl) % 1.37 [ 1.99, 0.75] Heterogeneity: Tau 2 = 0.68, Chi 2 = 58.00, df = 7 (P < ); I 2 = 88% Test for overall effect: Z = 4.30 (P < ) m-emr versus ESD m-emr ESD Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% Cl Choi et al % 1.80 [ 2.40, 1.19] Kim et al % 0.54 [0.11, 0.98] Niimi et al % 0.88 [ 1.72, 0.03] Zhao et al % 4.57 [ 6.38, 2.77] Total (95% Cl) % 1.50 [ 3.14, 0.14] Heterogeneity: Tau 2 = 2.53, Chi 2 = 59.75, df = 3 (P < ); I 2 = 95% Test for overall effect: Z = 1.80 (P = 0.07) EMR versus m-emr EMR m-emr Study or Subgroup Mean SD Total Mean SD Total Weight IV, Fixed, 95% Cl Kim et al % 0.23 [ 0.16, 0.63] Zhao et al % 0.14 [ 1.02, 0.74] Total (95% Cl) % 0.17 [ 0.19, 0.53] Heterogeneity: Chi 2 = 0.58, df = 1 (P = 0.45); I 2 = 0% Test for overall effect: Z = 0.93 (P = 0.35) IV, Random, 95% Cl IV, Fixed, 95% Cl Figure 4 Forest plot of procedure time. EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection; m-emr, modified EMR. 264 Journal of Gastroenterology and Hepatology 29 (2014)

7 Also, there might be selection bias between the two procedures because the endoscopists tended to resect larger lesions using ESD. However, endoscopically measured tumor size before resection was not significantly different among these groups in the current study. Further treatment of incompletely resected tumors consists of annual follow-up for grade 1 (G1) tumors, and transanal excision for G2/G3 tumors. 5 One reason in favor of the annual follow-up may be the indolent behavior of small G1 tumors. Another reason lies in the uncertainty of judging a cut margin because of the burning effect of electrosurgical devices on the residual tumor cells in a positive margin, which may make an aggressive approach not necessary. In the five incompletely resected cases by Kim et al., 21 pathologic assessment of specimens by salvage TEM found only changes due to scarring without remnant tumor. This might also partly explain the relatively low recurrence rate (5/328) compared with the relatively high incomplete resection rate in the EMR group aside from the salvage treatment. Delayed bleeding and perforation were comparable between the EMR and ESD groups. They rarely happened and were successfully managed by conservative medical treatment. In the metaanalysis by Cao et al. 27 comparing EMR and ESD for tumors of the entire gastrointestinal tract, ESD was found to have higher procedure-related bleeding (odds ratio [OR] 2.20, [95% CI 1.58, 3.07]) and perforation rates (OR 4.09, 95% CI [2.47, 6.80]). The differences between the current study and theirs could be explained by different indications for endoscopic procedures. In their analysis, different tumors were resected with ESD procedures, including early cancers which needed more comprehensive resections and lesions larger than 20 mm, while in the treatment of rectal neuroendocrine tumors, only rectal neuroendocrine tumors smaller than 16 mm without muscularis propria invasion were resected with ESD. Since m-emr and ESD were found to be comparable in the treatment of rectal neuroendocrine tumors in the current study, socioeconomic factor might be another aspect to consider, that is, the selection of endoscopic procedures should include considerations of the cost benefit. There may be differences in the covering of cost in medical care between insurance systems of different countries, and hospitals may defray the cost of operation expenditure that exceeds the insurance system payment in some countries, which might influence the choice of endoscopic procedures in a socioeconomic point of view. Seven studies in the current analysis enrolled patients with tumors smaller than mm in diameter, and three studies employed 15 mm as an upper limit for endoscopic resection. 16,17,20 The metastatic rate has been estimated to be less than 3% for rectal neuroendocrine tumors < 1 cm in diameter, and between % and 15% for a lesion between 1 cm and 2 cm; a significantly higher metastatic risk, considered to be between 60% and 80%, has been recorded in lesions larger than 2 cm in diameter. 5 All included studies ruled out patients with lymph node involvement on CT scanning or EUS to ensure curative resection. Other factors reported to be associated with metastasis include invasion of muscularis propria, lymphovascular invasion, atypical endoscopic features such as central depression, a high mitotic rate, and Ki-67 index In the meta-analysis by Zhou et al. studying factors associated with lymph node involvement in rectal neuroendocrine tumors, tumor size > 1 cm, invasion of muscularis propria, venous invasion, and central depression of the tumor body were considered to be risk factors of lymph node involvement, and EUS and magnetic resonance imaging were recommended for preoperative assessment to rule out lymph node involvement or invasion of muscularis propria in high-risk patients. 33 When encountered with tumors with these characteristics, a decision to perform endoscopic procedures should be prudently made. Limitations. All included studies were from Asian countries, which might be due to the relatively lower incidence of neuroendocrine tumors in Western countries. Whether the clinical outcomes of patients undergoing endoscopic procedures in the West were different from those in the East could not be analyzed in the current study. Other main limitations and possible biases included the fact that all studies were retrospective in nature and collected patient data across a wide study interval. Because of the relative rarity of this disease, it was very difficult for methodologically high-quality controlled clinical trials to be implemented. Biases might have been caused by the innovation of operative instruments and techniques, such as the advances in EMR procedures. Also, learning curve effects may play a role in introducing biases. In addition, there was conspicuous heterogeneity between included studies due to differences in indications for endoscopic procedures, time counting, and definitions of complete resection. Even though strict definitions of tumor size, complete resection, bleeding, and recurrence were applied, and a random-effects model was used, the precision of overall effect sizes could not be absolutely free from influence or exaggeration. Finally, we only included studies published in English, which might have caused selection biases. Conclusions. In conclusion, the available clinical evidence suggests that ESD or m-emr techniques could be applied to rectal neuroendocrine tumors with indications for endoscopic treatment. m-emr procedures appear to be comparable with ESD in the treatment of rectal neuroendocrine tumors with indications for endoscopic resection. Large prospective multicenter studies are needed to confirm these results obtained from low level of evidence. References 1 Tsikitis VL, Wertheim BC, Guerrero MA. Trends of incidence and survival of gastrointestinal neuroendocrine tumors in the United States: a seer analysis. J. Cancer 2012; 3: Ellis L, Shale MJ, Coleman MP. Carcinoid tumors of the gastrointestinal tract: trends in incidence in England since Am. J. Gastroenterol. 20; 5: Hauso O, Gustafsson BI, Kidd M et al. Neuroendocrine tumor epidemiology: contrasting Norway and North America. Cancer 2008; 113: Ito T, Sasano H, Tanaka M et al. Epidemiological study of gastroenteropancreatic neuroendocrine tumors in Japan. J. Gastroenterol. 20; 45: Caplin M, Sundin A, Nillson O et al. ENETS Consensus Guidelines for the management of patients with digestive neuroendocrine neoplasms: colorectal neuroendocrine neoplasms. Neuroendocrinology 2012; 95: Journal of Gastroenterology and Hepatology 29 (2014)

8 6 Suzuki S, Ishii N, Uemura M et al. Endoscopic submucosal dissection (ESD) for gastrointestinal carcinoid tumors. Surg. Endosc. 2012; 26: Imada-Shirakata Y, Sakai M, Kajiyama T et al. Endoscopic resection of rectal carcinoid tumors using aspiration lumpectomy. Endoscopy 1997; 29: Okamoto Y, Fujii M, Tateiwa S et al. Treatment of multiple rectal carcinoids by endoscopic mucosal resection using a device for esophageal variceal ligation. Endoscopy 2004; 36: Kinoshita T, Kanehira E, Omura K, Tomori T, Yamada H. Transanal endoscopic microsurgery in the treatment of rectal carcinoid tumor. Surg. Endosc. 2007; 21: Mashimo Y, Matsuda T, Uraoka T et al. Endoscopic submucosal resection with a ligation device is an effective and safe treatment for carcinoid tumors in the lower rectum. J. Gastroenterol. Hepatol. 2008; 23: Ono A, Fujii T, Saito Y et al. Endoscopic submucosal resection of rectal carcinoid tumors with a ligation device. Gastrointest. Endosc. 2003; 57: Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of Interventions Version [updated February 2008]. The Cochrane Collaboration, Cited 19 Nov Available from URL: 13 Wells GA, Shea B, O Connell D et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality if nonrandomized studies in meta-analyses Cited 12 Dec Available from URL: 14 Zhou PH, Yao LQ, Qin XY et al. Advantages of endoscopic submucosal dissection with needle-knife over endoscopic mucosal resection for small rectal carcinoid tumors: a retrospective study. Surg. Endosc. 20; 24: Zhao ZF, Zhang N, Ma SR et al. A comparative study on endoscopy treatment in rectal carcinoid tumors. Surg. Laparosc. Endosc. Percutan. Tech. 2012; 22: Sung HY, Kim SW, Kang WK et al. Long-term prognosis of an endoscopically treated rectal neuroendocrine tumor: -year experience in a single institution. Eur. J. Gastroenterol. Hepatol. 2012; 24: Park HW, Byeon JS, Park YS et al. Endoscopic submucosal dissection for treatment of rectal carcinoid tumors. Gastrointest. Endosc. 20; 72: Onozato Y, Kakizaki S, Iizuka H et al. Endoscopic treatment of rectal carcinoid tumors. Dis. Colon Rectum 20; 53: Niimi K, Goto O, Fujishiro M et al. Endoscopic mucosal resection with a ligation device or endoscopic submucosal dissection for rectal carcinoid tumors: an analysis of 24 consecutive cases. Dig. Endosc. 2012; 24: Lee DS, Jeon SW, Park SY et al. The feasibility of endoscopic submucosal dissection for rectal carcinoid tumors: comparison with endoscopic mucosal resection. Endoscopy 20; 42: Kim KM, Eo SJ, Shim SG et al. Treatment outcomes according to endoscopic treatment modalities for rectal carcinoid tumors. Clin. Res. Hepatol. Gastroenterol. 2013; 37: Kim HH, Park SJ, Lee SH et al. Efficacy of endoscopic submucosal resection with a ligation device for removing small rectal carcinoid tumor compared with endoscopic mucosal resection: analysis of 0 cases. Dig. Endosc. 2012; 24: Choi CW, Kang DH, Kim HW et al. Comparison of endoscopic resection therapies for rectal carcinoid tumor: endoscopic submucosal dissection versus endoscopic mucosal resection using band ligation. J. Clin. Gastroenterol. 2013; 47: Oberndorfer S. Karzinoide tumoren des dunndarms. Frankf. Z. Pathol. 1907; 1: Anthony LB, Strosberg JR, Klimstra DS et al. The NANETS consensus guidelines for the diagnosis and management of gastrointestinal neuroendocrine tumors (nets): well-differentiated nets of the distal colon and rectum. Pancreas 20; 39: Zhong DD, Shao LM, Cai JT. Endoscopic mucosal resection (EMR) versus endoscopic submucosal dissection (ESD) for rectal carcinoid tumours: a systematic review and meta-analysis. Colorectal Dis. 2013; 15: Cao Y, Liao C, Tan A et al. Meta-analysis of endoscopic submucosal dissection versus endoscopic mucosal resection for tumors of the gastrointestinal tract. Endoscopy 2009; 41: Shields CJ, Tiret E, Winter DC. Carcinoid tumors of the rectum: a multi-institutional international collaboration. Ann. Surg. 20; 252: Kim BN, Sohn DK, Hong CW et al. Atypical endoscopic features can be associated with metastasis in rectal carcinoid tumors. Surg. Endosc. 2008; 22: Shim KN, Yang SK, Myung SJ et al. Atypical endoscopic features of rectal carcinoids. Endoscopy 2004; 36: Shimizu T, Tanaka S, Haruma K et al. Growth characteristics of rectal carcinoid tumors. Oncology 2000; 59: Konishi T, Watanabe T, Kishimoto J et al. Prognosis and risk factors of metastasis in colorectal carcinoids: results of a nationwide registry over 15 years. Gut 2007; 56: Zhou X, Xie H, Xie L, Li J, Fu W. Factors associated with lymph node metastasis in radically resected rectal carcinoids: a systematic review and meta-analysis. J. Gastrointest. Surg. 2013; 17: [Epub ahead of print]. 266 Journal of Gastroenterology and Hepatology 29 (2014)

9 Appendix I Search strategies Database Search strategies No. of records Pubmed #1 Endoscopy, Gastrointestinal [Mesh] 203 #2 Endoscopy, Digestive System [Mesh] #3 Endoscopy [Mesh] #4 Surgical Procedure* AND,Endoscopic #5 Procedure* AND, Endoscopic Surgical* #6 endoscopic mucosal resection #7 endoscopic submucosal dissection #8 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 #9 neuroendocrine tumor* AND, rectal # neuroendocrine tumor* AND, rectum #11 neuroendocrine tumour* AND, rectal #12 neuroendocrine tumour* AND, rectum #13 neuroendocrine cancer* AND, rectal #14 neuroendocrine cancer* AND, rectum #15 neuroendocrine carcinoma* AND, rectal #16 neuroendocrine carcinoma* AND, rectum #17 neuroendocrine malignan* AND, rectal #18 neuroendocrine malignan* AND, rectum #19 neuroendocrine neoplasm* AND, rectal #20 neuroendocrine neoplasm* AND, rectum #21 carcinoid* AND, rectal #22 carcinoid* AND, rectum #23 carcinoid tumor* AND, rectal #24 carcinoid tumor* AND, rectum #25 carcinoid tumour* AND, rectal #26 carcinoid tumour* AND, rectum #27 carcinoid cancer* AND, rectal #28 carcinoid cancer* AND, rectum #29 carcinoid carcinoma* AND, rectal #30 carcinoid carcinoma* AND, rectum #31 carcinoid malignan* AND, rectal #32 carcinoid malignan* AND, rectum #33 carcinoid neoplasm* AND, rectal #34 carcinoid neoplasm* AND, rectum #35 #9 OR # OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 #36 #8 OR #35 Journal of Gastroenterology and Hepatology 29 (2014)

10 Appendix I Continued Database Search strategies No. of records The Cochrane Central Register of Controlled Trails #1 MeSH descriptor: [Endoscopy] explode all trees 4 #2 MeSH descriptor: [Endoscopy, Gastrointestinal] explode all trees #3 MeSH descriptor: [Endoscopy, Digestive System] explode all trees #4 endoscopic mucosal resection :ti,ab,kw #5 endoscopic submucosal dissection :ti,ab,kw #6 #1 or #2 or #3 or #4 or #5 #7 MeSH descriptor: [Neuroendocrine Tumors] explode all trees #8 MeSH descriptor: [Carcinoid Tumor] explode all trees #9 neuroendocrine near tumor*:ti,ab,kw # neuroendocrine near tumour*:ti,ab,kw #11 neuroendocrine near cancer*:ti,ab,kw #12 neuroendocrine near carcinoma*:ti,ab,kw #13 neuroendocrine near neoplasm*:ti,ab,kw #14 neuroendocrine near malignan*:ti,ab,kw #15 carcinoid*:ti,ab,kw #16 carcinoid near tumor*:ti,ab,kw #17 carcinoid near tumour*:ti,ab,kw #18 carcinoid near cancer*:ti,ab,kw #19 carcinoid near neoplasm*:ti,ab,kw #20 carcinoid near malignan*:ti,ab,kw #21 #7 or # or #9 or # or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 #22 rectum:ti,ab,kw #23 rectal*:ti,ab,kw #24 #22 or #23 #25 #6 and #21 and #24 Medline and Embase #1 endoscopy /exp 312 #2 digestive tract endoscopy /exp #3 gastrointestinal endoscopy ;/exp #4 endoscopic therapy ;/exp #5 endoscopic mucosal resection ;/exp or endoscopic mucosal resection ; #6 endoscopic submucosal dissection ;/exp or endoscopic submucosal dissection #7 #1 or #2 or #3 or #4 or #5 or #6 #8 neuroendocrine tumor /exp #9 carcinoid /exp # neuroendocrine NEAR/1 tumour* #11 neuroendocrine NEAR/1 tumor* #12 neuroendocrine NEAR/1 cancer* #13 neuroendocrine NEAR/1 carcinoma* #14 neuroendocrine NEAR/1 neoplasm* #15 neuroendocrine NEAR/1 malignan* #16 carcinoid* #17 carcinoid NEAR/1 tumour* #18 carcinoid NEAR/1 tumor* #19 carcinoid NEAR/1 cancer* #20 carcinoid NEAR/1 carcinoma* #21 carcinoid NEAR/1 neoplasm* #22 carcinoid NEAR/1 malignan* #23 #8 or #9 or # or #11 or #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19 or #20 or #21 or #22 #24 rectal /exp OR rectal #25 rectum /exp OR rectum #26 #24 or #25 #27 #7 and # 23 and # Journal of Gastroenterology and Hepatology 29 (2014)

11 本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 ( 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具