董紅霞,梁浩
不同國家人群幽門螺桿菌感染與結(jié)直腸癌相關(guān)性的Meta分析
董紅霞,梁浩
目的 探討不同洲別及經(jīng)濟狀況下人群幽門螺桿菌(Hp)感染與結(jié)直腸癌發(fā)病的相關(guān)性。方法 計算機檢索PubMed、EMBASE、High Wire Press、Ovid、Medline和EBSCO等數(shù)據(jù)庫,對符合納入標(biāo)準(zhǔn)的Hp感染與結(jié)直腸癌發(fā)病關(guān)系的文獻(xiàn)應(yīng)用RevMan 5.3軟件進(jìn)行Meta分析,同時按照洲別(亞洲、歐洲和美洲)和經(jīng)濟發(fā)展情況(發(fā)達(dá)國家與發(fā)展中國家)對各分析結(jié)果的OR值及95%CI進(jìn)行合并計算。結(jié)果 共納入文獻(xiàn)23篇,總樣本量182 561例,其中Hp陽性88 378例,Hp陰性94 183例。Meta分析結(jié)果顯示Hp感染與結(jié)直腸癌的發(fā)病明顯相關(guān)(OR=1.42,95%CI 1.38~1.46)。從不同洲別來看,亞洲國家人群相關(guān)性較低(OR=1.29,95%CI 1.13~1.48),美洲國家人群相關(guān)性最高(OR=1.44,95%CI 1.39~1.48)。從經(jīng)濟發(fā)展情況來看,發(fā)展中國家人群相關(guān)性較低(OR=1.17,95%CI 1.01~1.37),發(fā)達(dá)國家人群相關(guān)性較高(OR=1.43,95%CI 1.39~1.47)。結(jié)論 Hp感染是結(jié)直腸癌發(fā)生的危險因素,Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性與地理位置及經(jīng)濟發(fā)展情況有關(guān)。
螺桿菌,幽門;結(jié)直腸腫瘤;Meta分析
結(jié)直腸癌是一種常見的惡性腫瘤,由于其發(fā)生是多因素、多步驟交互作用的結(jié)果,預(yù)防和治療都較困難。幽門螺旋桿菌(Helicobacter pylori,Hp)作為一種常見的致癌病原體,與胃癌及相關(guān)淋巴瘤的關(guān)系已得到公認(rèn)。目前,有研究結(jié)果提出Hp感染與結(jié)直腸癌的發(fā)生相關(guān),Hp感染是結(jié)直腸癌的誘發(fā)因素[1-2]。本研究對國內(nèi)外發(fā)表的有關(guān)Hp感染與結(jié)直腸癌關(guān)系的臨床研究進(jìn)行系統(tǒng)評價,探討不同國家人群Hp感染與結(jié)直腸癌的相關(guān)性,以期為結(jié)直腸癌的病因探索和防治提供參考依據(jù)。
1.1 檢索策略 計算機檢索PubMed、EMBASE、High Wire Press、Ovid、Medline和EBSCO等數(shù)據(jù)庫。檢索詞包括:Helicobacter pylori和colorectal carcinoma、colorectal cancer、colon cancer、colonic neoplasms。
1.2 納入及排除標(biāo)準(zhǔn) 納入標(biāo)準(zhǔn):①2014年10月之前發(fā)表的,收錄在PubMed、EMBASE、High Wire Press、Ovid、Medline和EBSCO的文獻(xiàn);②涉及Hp與結(jié)直腸癌的關(guān)系,并提供病例組和對照組人數(shù);③Hp感染至少經(jīng)過一種診斷方法判斷。排除標(biāo)準(zhǔn):①只有摘要而缺乏全文的文獻(xiàn)資料以及數(shù)據(jù)不完整無法利用的文獻(xiàn);②重復(fù)發(fā)表的文獻(xiàn);③無對照組的文獻(xiàn)以及綜述性文獻(xiàn);④實驗有明顯設(shè)計缺陷和偏倚太大的文獻(xiàn),如未按病例對照研究設(shè)計等。
1.3 資料提取 所有檢索策略通過多次預(yù)檢索后確定,最初共檢索到635篇文獻(xiàn)。經(jīng)過簡單的題目篩選排除大部分不符合條件的文獻(xiàn)后,對剩余所獲文獻(xiàn)摘要進(jìn)行閱讀,在排除明顯不符合納入標(biāo)準(zhǔn)的文獻(xiàn)后,對可能符合納入標(biāo)準(zhǔn)的文獻(xiàn)進(jìn)行全文閱讀,最終納入23篇符合納入標(biāo)準(zhǔn)的文獻(xiàn)。
1.4 病例資料分析 納入總樣本量為182 561例,其中Hp陽性88 378例,Hp陰性94 183例;Hp陽性腫瘤患者13 691例,Hp陰性腫瘤患者10 604例。其中,所納入的病例按照所屬洲的不同分為亞洲、歐洲和美洲國家人群,按照所屬國家經(jīng)濟情況分為發(fā)達(dá)國家和發(fā)展中國家人群。按照兩種不同的分類對各國Hp感染與結(jié)直腸癌的相關(guān)性進(jìn)行分析。
1.5 統(tǒng)計學(xué)處理 采用RevMan 5.3軟件進(jìn)行Meta分析。采用χ2檢驗評估是否存在統(tǒng)計學(xué)異質(zhì)性,I2檢驗評估異質(zhì)性的大小,若I2>50%,表明存在統(tǒng)計學(xué)異質(zhì)性。對結(jié)果測量的二分類變量采用比值比(OR),兩者效應(yīng)量均用95%可信區(qū)間(CI)表示,P<0.05為差異有統(tǒng)計學(xué)意義。計算合并OR值及95%CI,繪制森林圖,并對發(fā)表性偏倚進(jìn)行漏斗圖分析。
表1 納入研究的基本信息Tab.1 Baseline characteristics of the included studies
2.1 納入文獻(xiàn)基本信息 共有23篇文獻(xiàn)符合納入標(biāo)準(zhǔn),納入文獻(xiàn)的第一作者、發(fā)表年份、所屬國家、研究類型、Hp檢測方法等基本信息見表1。納入文章的總病例數(shù)、Hp陽性病例數(shù)、結(jié)直腸癌病例數(shù)以及OR值、95% CI見圖1。
2.2 不同洲別Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性分析
2.2.1 相關(guān)性分析 共納入了23篇文章,屬于11個不同國家,分為3個不同洲別,即亞洲、美洲和歐洲。Meta分析結(jié)果顯示,亞洲國家人群Hp感染與結(jié)直腸癌發(fā)病合并OR值為1.29(95%CI 1.13~1.48,χ2=42.08,P<0.00001),歐洲國家人群Hp感染與結(jié)直腸癌發(fā)病合并OR值為1.33(95%CI 1.20~1.48,χ2=23.71,P=0.008),美洲國家人群Hp感染與結(jié)直腸癌發(fā)病合并OR值為1.44(95%CI 1.39~1.48,χ2=5.40,P=0.25,圖2),表明亞洲國家人群Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性較低,其他洲別相關(guān)性較高,具有一定的地域差異。
圖1 納入研究的樣本數(shù)以及Hp感染與結(jié)直腸癌的相關(guān)性Fig.1 The samples of the included studies and the correlation between Hp infection and colorectal cancer
2.2.2 發(fā)表偏倚 按照洲屬不同進(jìn)行Meta分析,其漏斗圖分布不對稱(圖3),亞洲國家的顯示效應(yīng)值相對集中在圖形右側(cè),歐洲國家集中于中上部,而美洲國家則集中于左側(cè),說明入選文獻(xiàn)可能存在一定的發(fā)表偏倚。其原因可能與各研究中的病例數(shù)及各研究采用的Hp檢測方法不同有關(guān)。
2.3 不同經(jīng)濟發(fā)展情況下Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性分析
2.3.1 相關(guān)性分析 共納入23篇文獻(xiàn),屬于11個不同國家,共分為兩種不同經(jīng)濟發(fā)展情況,即發(fā)達(dá)國家和發(fā)展中國家。Meta分析結(jié)果顯示,發(fā)達(dá)國家Hp感染與結(jié)直腸癌發(fā)病合并OR值為1.43(95%CI 1.39~1.47,χ2=44.69,P=0.0003),發(fā)展中國家Hp感染與結(jié)直腸癌發(fā)病合并OR值為1.17(95%CI 1.01~1.37,χ2=24.30,P<0.0001,圖4),表明發(fā)展中國家人群Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性較低,而發(fā)達(dá)國家的相關(guān)性較高。
2.3.2 發(fā)表偏倚 按照經(jīng)濟發(fā)展情況的不同進(jìn)行Meta分析,其漏斗圖分布不對稱(圖5),發(fā)達(dá)國家的顯示效應(yīng)值相對集中在圖形的中上部,而發(fā)展中國家的集中于右下部,說明入選文獻(xiàn)可能存在一定的發(fā)表偏倚。其原因可能與各研究中病例數(shù)及各研究采用的Hp檢測方法不同有關(guān)。
Hp是目前已知最常見的胃部感染的細(xì)菌之一,其與胃癌、黏膜相關(guān)淋巴瘤的關(guān)系已獲公認(rèn)[25-26]。但近年來發(fā)達(dá)國家的很多文獻(xiàn)指出,Hp感染與結(jié)直腸癌之間有一定的相關(guān)性[2,27],Rokkas等[28]通過對文獻(xiàn)進(jìn)行Meta分析指出,Hp感染可增加結(jié)直腸癌的發(fā)生風(fēng)險(OR=1.41,95%CI 1.24~1.60);而Wu 等[29]經(jīng)過Meta分析認(rèn)為,Hp感染確實增加了結(jié)直腸癌的發(fā)生率(OR=1.66,95%CI 1.39~1.97)。但是對于不同國家地區(qū)或經(jīng)濟發(fā)展情況不同的國家,Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性比較尚未見報道。本研究在調(diào)研以往Hp感染與結(jié)直腸癌相關(guān)性文獻(xiàn)的基礎(chǔ)上,對不同國家地區(qū)間或經(jīng)濟發(fā)展情況不同國家間二者的相關(guān)性進(jìn)行比較,以期得到Hp感染與結(jié)直腸癌發(fā)病相關(guān)性的地域性或經(jīng)濟發(fā)展情況差異。
本Meta分析共納入23篇文獻(xiàn),分屬于11個不同的國家,按照洲屬不同分為亞洲、歐洲和美洲國家,按照經(jīng)濟發(fā)展情況分為發(fā)達(dá)國家和發(fā)展中國家。其中,土耳其雖然橫跨歐亞兩大洲,但其經(jīng)濟政治等比較接近歐洲,因此列為歐洲國家進(jìn)行數(shù)據(jù)分析。而經(jīng)濟發(fā)展情況分類中,以色列情況比較特殊,它在2005年才屬于發(fā)達(dá)國家之列,因此將以色列2000年和2001年發(fā)表的文章列為發(fā)展中國家,而將其2014年發(fā)表的文章列為發(fā)達(dá)國家。Meta分析結(jié)果提示,Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性與地理位置及經(jīng)濟發(fā)展情況有關(guān)。從地理位置上看,亞洲地區(qū)相關(guān)性較低,美洲國家相關(guān)性較高,從經(jīng)濟發(fā)展情況來看,發(fā)展中國家相關(guān)性較低,發(fā)達(dá)國家相關(guān)性較高,這一點與Hp的感染情況恰恰相反[30],提示仍有很多未知因素影響Hp感染與結(jié)直腸癌的關(guān)系。
圖2 不同洲別國家Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性比較Fig.2 Comparison of colorectal cancer-Hp infection correlation in different continents
圖3 不同洲別國家Hp感染與結(jié)直腸癌發(fā)病相關(guān)性的漏斗圖Fig.3 Funnel plot of colorectal cancer-Hp infection correlation in different continents
從數(shù)據(jù)分析結(jié)果來看,Meta分析的異質(zhì)性(I2)都較高,且經(jīng)過亞組(按照洲屬不同和經(jīng)濟發(fā)展情況不同)處理,多數(shù)I2仍舊>50%,說明具有異質(zhì)性。漏斗圖分析結(jié)果表明有發(fā)表偏倚存在,且經(jīng)過不同分組后,其偏倚情況未發(fā)生改變。這在一定程度上降低了評價結(jié)果的可靠性。但是需要指出的是,本次研究是對現(xiàn)有的文獻(xiàn)資料進(jìn)行Meta分析,所納入的23篇文獻(xiàn)來自全世界11個不同國家,分屬3個大洲,其判斷指標(biāo)、病例數(shù)、年齡、種族、環(huán)境、地域、經(jīng)濟或其他因素都會影響Hp的流行及其對結(jié)直腸癌的作用。在關(guān)注到發(fā)達(dá)國家Hp感染與結(jié)直腸癌發(fā)病相關(guān)性較高的同時,也需警惕我們國家結(jié)直腸癌的發(fā)生率正逐年升高,且Hp的感染率不降反升。由于發(fā)展中國家研究文獻(xiàn)較少,覆蓋的人群不夠廣泛,合并的結(jié)果也可能存在一定局限性,尚需更為合理可靠的大規(guī)模隨機雙盲對照試驗進(jìn)一步驗證。
圖4 發(fā)達(dá)國家與發(fā)展中國家Hp感染與結(jié)直腸癌發(fā)病的相關(guān)性比較Fig.4 Comparison of colorectal cancer-Hp infection correlation in developed and developing countries
圖5 發(fā)達(dá)國家與發(fā)展中國家Hp感染與結(jié)直腸癌的漏斗圖Fig.5 Funnel plot of colorectal cancer associated with Hp infection in developed and developing countries
[1] Zhang Y, Hoffmeister M, Weck MN, et al. Helicobacter pylori infection and colorectal cancer risk: evidence from a large population-based case-control study in Germany[J]. Am J Epidemiol, 2012, 175(5): 441-450.
[2] Zumkeller N, Brenner H, Zwahlen M, et al. Helicobacter pylori infection and colorectal cancer risk: a meta-analysis[J]. Helicobacter, 2006, 11(2): 75-80.
[3] Talley NJ, Zinsmeister AR, Weaver A, et al. Gastric adenocarcinoma and Helicobacter pylori infection[J]. J Natl Cancer Inst, 1991, 83(23): 1734-1739.
[4] Penman ID, el-Omar E, Ardill JE, et al. Plasma gastrin concentrations are normal in patients with colorectal neoplasia and unaltered following tumor resection[J]. Gastroenterology, 1994, 106(5): 1263-1270.
[5] Moss SF, Neugut AI, Garbowski GC, et al. Helicobacter pylori seroprevalence and colorectal neoplasia: evidence against an association[J]. J Natl Cancer Inst, 1995, 87(10): 762-763.
[6] Meucci G, Tatarella M, Vecchi M, et al. High prevalence of Helicobacter pylori infection in patients with colonic adenomas and carcinomas[J]. J Clin Gastroenterol, 1997, 25(4): 605-607.
[7] Thorburn CM, Friedman GD, Dickinson CJ, et al. Gastrin and colorectal cancer: a prospective study[J]. Gastroenterology, 1998, 115(2): 275-280.
[8] Fireman Z, Trost L, Kopelman Y, et al. Helicobacter pylori: seroprevalence and colorectal cancer[J]. Isr Med Assoc J, 2000, 2(1): 6-9.
[9] Hartwich J, Konturek SJ, Pierzchalski P, et al. Molecular basis of colorectal cancer-role of gastrin and cyclooxygenase-2[J]. Med Sci Monit, 2001, 7(6): 1171-1181.
[10] Shmuely H, Passaro D, Figer A, et al. Relationship between Helicobacter pylori CagA status and colorectal cancer[J]. Am J Gastroenterol, 2001, 96(12): 3406-3410.
[11] Siddheshwar RK, Gray JC, Kelly SB. Plasma levels of progastrinbut not amidated gastrin or glycine extended gastrin are elevated in patients with colorectal carcinoma[J]. Gut, 2001, 48(1): 47-52.
[12] Limburg PJ, Stolzenberg-Solomon RZ, Colbert LH, et al. Helicobacter pylori seropositivity and colorectal cancer risk: a prospective study of male smokers[J]. Cancer Epidemiol Biomarkers Prev, 2002, 11(10 Pt 1): 1095-1099.
[13] Mizuno S, Morita Y, Inui T, et al. Helicobacter pylori infection is associated with colon adenomatous polyps detected by highresolution colonoscopy[J]. Int J Cancer, 2005, 117(6): 1058-1059.
[14] Fujimori S, Kishida T, Kobayashi T, et al. Helicobacter pylori infection increases the risk of colorectal adenoma and adenocarcinoma, especially in women[J]. J Gastroenterol, 2005, 40(9): 887-893.
[15] Machida-Montani A, Sasazuki S, Inoue M, et al. Atrophic gastritis, Helicobacter pylori, and colorectal cancer risk: a casecontrol study[J]. Helicobacter, 2007, 12(4): 328-332.
[16] D'Onghia V, Leoncini R, Carli R, et al. Circulating gastrin and ghrelin levels in patients with colorectal cancer: correlation with tumour stage, Helicobacter pylori infection and BMI[J]. Biomed Pharmacother, 2007, 61(2-3): 137-141.
[17] Jones M, Helliwell P, Pritchard C, et al. Helicobacter pylori in colorectal neoplasms: is there an aetiological relationship [J]? World J Surg Oncol, 2007, 5: 51.
[18] Zumkeller N, Brenner H, Chang-Claude J, et al. Helicobacter pylori infection, interleukin-1 gene polymorphisms and the risk of colorectal cancer: evidence from a case-control study in Germany[J]. Eur J Cancer, 2007, 43(8): 1283-1289.
[19] Wu IC, Wu DC, Yu FJ, et al. Association between Helicobacter pylori seropositivity and digestive tract cancers[J]. World J Gastroenterol, 2009, 15(43): 5465-5471.
[20] Engin AB, Karahalil B, Engin A, et al. Oxidative stress, Helicobacter pylori, and OGG1 Ser326Cys, XPC Lys939Gln, and XPD Lys751Gln polymorphisms in a Turkish population with colorectal carcinoma[J]. Genet Test Mol Biomarkers, 2010, 14(4): 559-564.
[21] Abbass K, Gul W, Beck G, et al. Association of Helicobacter pylori infection with the development of colorectal polyps and colorectal carcinoma[J]. South Med J, 2011, 104(7): 473-476.
[22] Strofilas A, Lagoudianakis EE, Seretis C, et al. Association of Helicobacter pylori infection and colon cancer[J]. J Clin Med Res, 2012, 4(3): 172-176.
[23] Sonnenberg A, Genta RM. Helicobacter pylori is a risk factor for colonic neoplasms[J]. Am J Gastroenterol, 2013, 108(2): 208-215.
[24] Shmuely H, Melzer E, Braverman M, et al. Helicobacter pylori infection is associated with advanced colorectal neoplasia[J]. Scand J Gastroenterol, 2014, 49(1): 35-42.
[25] Tao W, Zhang N, Yang L. Relationship between Helicobacter pylori infection and serum pepsinogen levels in patients with chronic gastritic diseases or gastric cancer[J]. Med J Chin PLA, 2012, 37(4): 350-353. [陶偉, 張寧, 楊力. 慢性胃病及胃癌中幽門螺桿菌感染與血清胃蛋白酶原水平變化的關(guān)系[J]. 解放軍醫(yī)學(xué)雜志, 2012, 37(4): 350-353.]
[26] Liu J, Li ZS. Helicobacter pylori infection and gastric cancer[J]. Chin J Pract Intern Med, 2014, 34(5): 538-544. [劉炯, 李兆申.幽門螺桿菌感染與胃癌發(fā)生[J]. 中國實用內(nèi)科雜志, 2014, 34(5): 538-544.]
[27] Chen YS, Xu SX, Ding YB, et al. Helicobacter pylori infection and the risk of colorectal adenoma and adenocarcinoma: an updated meta-analysis of different testing methods[J]. Asian Pac J Cancer Prev, 2013, 14(12): 7613-7619.
[28] Rokkas T, Sechopoulos P, Pistiolas D, et al. The relationship of Helicobacter pylori infection and colon neoplasia, on the basis of meta-analysis[J]. Eur J Gastroenterol Hepatol, 2013, 25(11): 1286-1294.
[29] Wu Q, Yang ZP, Xu P, et al. Association between Helicobacter pylori infection and the risk of colorectal neoplasia: a systematic review and meta-analysis[J]. Colorectal Dis, 2013, 15(7): e352-e364.
[30] Malaty HM. Epidemiology of Helicobacter pylori infection[J]. Best Pract Res Clin Gastroenterol, 2007, 21(2): 205-214.
Correlation between colorectal cancer and Helicobacter pylori infection in different countries: A meta-analysis
DONG Hong-xia, LIANG Hao*
Department of Gastroenterology, General Hospital of PLA, Beijing 100853, China
*
, E-mail: lianghao301@163.com
ObjectiveTo explore the relationship between Helicobacter pylori (Hp) infection and colorectal cancer in different continents or economic conditions.MethodsPublished case-control studies dealing with the correlation of colorectal cancer with Hp infection were retrieved from PubMed, EMBASE, High Wire Press, Ovid, Medline and EBSCO. Meta-analysis was performed by using RevMan 5.3 software. We selected the OR and 95% CI as indicators of the analysis according to different continents (Asia, Europe, and America) and economic conditions (developed countries and developing countries).ResultsA total of 23 studies dealing with the correlation of colorectal cancer with Hp infection were included in the present meta-analysis, and there was a total sample of 182,561 patients, including 88,378 cases in Hp positive group and 94,183 cases in Hp negative group. The results of meta-analysis showed the OR was 1.42 (95%CI 1.38-1.46). Geographically, the correlation was low in Asian countries (OR=1.29, 95%CI 1.13-1.48), and was highest in American countries (OR=1.44, 95%CI 1.39-1.48). According to the economic conditions, the correlation was low in developing countries (OR=1.17, 95%CI 1.01-1.37), and was higher in developed countries (OR=1.43, 95%CI 1.39-1.47).ConclusionsHp infection is a risk factor for colorectal cancer. It seems that there is a close relation ship between its incidence and geography, and also economic condition.
Helicobacter pylori; colorectal neoplasms; meta-analysis
R735.35;R735.37
A
0577-7402(2015)03-0236-06
10.11855/j.issn.0577-7402.2015.03.13
2014-11-05;
2015-02-13)
(責(zé)任編輯:熊曉然)
董紅霞,主治醫(yī)師,碩士研究生。主要從事幽門螺桿菌感染引發(fā)的相關(guān)疾病研究
100853 北京 解放軍總醫(yī)院消化科(董紅霞、梁浩)
梁浩,E-mail:lianghao301@163.com