Solon, Jacqueline G. MD*; Burke, John P. PhD*; Walsh, Stewart R. FRCSEd*,†; Coffey, J. Calvin PhD*,†
Up to 80% of patients with Crohn’s Disease (CD) will require at least one surgical intervention over the course of their life.1 A significant proportion will undergo more than one intestinal resection.2 Up to 90% of patients will have endoscopic features of disease recurrence within a year of surgery,3 whereas almost half will develop clinical recurrence within 5 years4 and up to three quarters will require repeat surgical resection within 10 years.5
The determinants of disease recurrence following intestinal resection for CD include smoking, site and extent of disease as well as genetic make-up and altered immune response.6–8 However, there is inconsistency among studies that characterize risk factors for surgical recurrence. It is possible that biomarkers may resolve this issue by providing objective indices of disease activity and prediction of relapse. Recent data demonstrate that therapy using either azathioprine9 or infliximab after resection10 significantly reduces the risk of endoscopic recurrence. However, these medications have potential side effects such as leucopenia, pancreatitis, allergic reactions, infections, or malignancy.11–13 Accurate and reproducible biomarkers of disease recurrence would thus be helpful in tailoring postoperative medical therapy appropriately.
The association of NOD2 (nucleotide-binding oligomerisation domain 2)/CARD15 (caspase activation and recruitment domain 15) polymorphism present on chromosome 16 q12 with CD was first reported at the same time by 2 groups in 2001.14,15 NOD2 encodes an intracellular receptor peptidoglycan, which is found within the cell walls of bacteria.16 The 3 polymorphisms associated with CD (R702W [rs139104022], G908R [rs2066845] and L1007finsC [rs2066847]) impair cellular responses to peptidoglycan, thereby altering the host’s inflammatory immune response. Heterozygous carriers of a polymorphism have a 1.75-fold to 4-fold increase risk of disease development, whereas homozygosity increases the risk by a factor of 11 to 27.17 NOD2 polymorphisms have also been associated with ileal involvement, fibrostenosing disease, and an increased rate of surgical intervention.6,18–21
Although NOD2 polymorphisms are associated with an increased risk of surgical resection, their ability to predict patients at risk of postoperative disease recurrence remains equivocal.22–24 To address this, a meta-analysis of all studies that have examined the effect of NOD2 polymorphism on surgical recurrence was undertaken.
MATERIALS AND METHODS
Literature Search and Study Selection
A systematic search of Medline, Embase, and Ovid was performed for all studies published comparing expression of NOD2 polymorphisms with recurrence by entering the following in the search algorithm: Crohn’s AND NOD OR NOD2 AND recurrence. The Cochrane Central Register of Controlled Trials was also searched for articles that compared surgical recurrence with NOD2 polymorphism expression. The latest search was done on May 28, 2012. Two authors (J.G.S. and J.P.B.) independently examined the title and abstract of citations, and the full texts of potentially eligible trials were obtained; disagreements were resolved by discussion. The reference list of retrieved articles was further screened for additional eligible publications. When data were unclear or incomplete, the corresponding author was contacted to clarify data extraction.
Studies comparing surgical recurrence requiring re-operations in patients with CD and NOD2 polymorphism expression were eligible for inclusion. The primary endpoint of this meta-analysis was recurrence of disease requiring surgical intervention. Noncomparative studies and review articles with no original data were excluded. Whenever more than 1 publication reported results for the same group of patients, only the report with the longest follow-up was included to avoid data duplication. There were no language restrictions.
Data Extraction and Outcomes
The following information regarding each eligible trial was recorded: author’s names, journal, year of publication, study design, follow-up duration, total number of patients, patient ethnicity, and smoking prevalence. The following information was recorded from both arms of each eligible study: the total number of patients undergoing initial surgical resection, the number of patients expressing NOD2 polymorphisms and the subtype of polymorphism, and the number of expressing and nonexpressing patients who developed surgical recurrence.
All pooled outcome measures were determined using fixed and random-effects models as described by DerSimonian and Laird25 and the odds ratio (OR) was estimated with its variance and 95% confidence interval (CI). The random-effects analysis weighted the natural logarithm of each study's OR by the inverse of its variance plus an estimate of the between-study variance in the presence of between-study heterogeneity. As previously described,26 heterogeneity between ORs for the same outcome between different studies was assessed. This was through the use of the I2 inconsistency test and chi-square–based Cochran’s Q statistic27 test in which P < 0·05 is taken to indicate the presence of significant heterogeneity. The predictive performance of NOD2 was assessed with a summary receiver operating characteristic (ROC) curve. We analyzed the data using the MIDAS (Meta-analytical Integration of Diagnostic Accuracy Studies) command. Analyses were conducted using Statsdirect version 2.5.6 (StatsDirect Ltd, Cheshire, United Kingdom). Pooled sensitivity and specificity values together with the area under the summary ROC values were calculated using Meta-Disc version 1.4 (Universidad Complutense, Madrid, Spain).
Six published articles that directly compared NOD2 polymorphism expression with surgical recurrence in CD were identified20,22–24,28,29 (Table 1). A total of 40 potentially eligible studies were identified initially; with 34 eventually excluded from the meta-analysis based on title and abstract or due to duplication of data sets (Fig. 1). All included studies were published within the last 10 years and were published by European centers. The spectrum of patients was generally reflective of clinical practice in all studies (Table 2).
Risk of Surgical Recurrence
In total, 1003 patients with CD who underwent intestinal resection were assessed for NOD2 single nucleotide polymorphism expression. Three-hundred forty patients carried at least 1 NOD2 polymorphism (33.9%) and 146 patients (43%) expressed the L1007finsC variant, 88 (26%) expressed the G908R variant, whereas 156 (46%) expressed the R702W variant (either as simple or compound heterozygotes or homozygotes). A total of 335 patients (33.4%) experienced a surgical recurrence over an average median follow-up period of 12.2 years (range, 5.6–16.2 years).
Overall, 39.1% of NOD2 carriers experienced a surgical recurrence relative to 30.5% of non-NOD2 patients. On fixed-effects analysis, this led to a significantly increased risk of recurrence (OR: 1.44, 95% CI: 1.09–1.90, P = 0.012) but not on random-effects analysis (OR: 1.58, 95% CI: 0.97–2.57, P = 0.064) (Fig. 2). There was significant across trial study heterogeneity (Cochran Q: 12.36, P = 0.030, I2: 59.6%). This resulted in a pooled sensitivity of 0.397 (95% CI: 0.344–0.452) and a specificity of 0.690 (95% CI: 0.654–0.725). The sensitivity and specificity of the presence of any NOD2 polymorphism for surgical recurrence in each study are displayed on the summary ROC diagram (Fig. 3). The summary ROC curve and the summary operating point (with 95% confidence contour) are shown in relation to the individual studies. A perfect test has an area under the summary ROC curve close to 1, and a poor test has an area under the summary ROC curve close to 0.5 (a coin flip has a diagnostic accuracy of 0.5). The area under the summary ROC curve was 0.638 for surgical recurrence, consistent with poor diagnostic accuracy.
Subgroup analysis of individual NOD2 polymorphisms was next performed. On random-effects analysis, the R702W variant was not associated with surgical recurrence (OR: 1.65, 95% CI: 0.96–2.83, P = 0.071) (Fig. 4A) with a pooled sensitivity of 0.204 (95% CI: 0.140–0.282) and a specificity of 0.860 (95% CI: 0.814–0.898). The G908R variant was not associated with surgical recurrence (OR: 1.01, 95% CI: 0.53–1.91, P = 0.974) (Fig. 4B) with a pooled sensitivity of 0.117 (95% CI: 0.068–0.183) and a specificity of 0.885 (95% CI: 0.842–0.919). The L1007finsC variant was not associated with surgical recurrence (OR: 1.00, 95% CI: 0.44–2.76, P = 0.831) (Fig. 4C) with a pooled sensitivity of 0.153 (95% CI: 0.097–0.225) and a specificity of 0.832 (95% CI: 0.784–0.874).
Analysis of the funnel plot (Fig. 5) demonstrated across trial publication bias, however, the Begg-Mazumdar (P = 0.136) and Egger (P = 0.182) tests did not reach statistical significance.
Epidemiologic data indicate that the heterogeneity of CD may in part relate to genetic factors.30–34 Thus, considerable research has focused on the identification of genetic determinants of disease progression. Most of the discrepancies in the literature on predictors of postoperative recurrence are because of variable definitions of CD recurrence (endoscopic, histological, clinical, surgical, etc.) as well as length and type of follow-up. Although NOD2 polymorphisms are associated with increased risk of disease development14,15,35 (in particular, fibrostenosing disease21), controversy persists in relation to association with postoperative recurrence.22,23,29 The current study has shown that carriers of one or more NOD2 polymorphisms had a 58% increased risk of developing surgical recurrence compared with noncarriers; however, this did not reach statistical significance and was associated with significant across study heterogeneity.
This meta-analysis is based on all currently available published data comparing NOD2 polymorphism expression and surgical recurrence, thereby providing a strong appraisal of its predictive value. However, there are a number of important limitations. The past decade has seen a dramatic increase in gene association studies and a contemporaneous increase in experimental design variation.36 The STrengthening the REporting of Genetic Association studies (STREGA) initiative37 expanded the STrengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist to improve the interrogation of evidence from genetic studies and thereby standardize reporting of results. As the studies contained in this meta-analysis did not conform to STREGA guidelines this limits extrapolations from data contained. In addition, the studies contained small to moderate numbers of patients in mainly single centers. Taken together, these factors have contributed to a considerable level of heterogeneity. The only available data for inclusion were from European centers. Therefore, the results presented herein may apply only to a European population and cannot be extrapolated to other populations. The analysis is also limited by the absence of data to allow for evaluation of confounding factors such as smoking status, fistulating disease, ileal involvement, or did it clearly define what was considered surgical recurrence (stricturoplasty versus re-resection, etc).
The fact the fixed-effects analysis was significant and the random-effects analysis was not is not surprising. A fixed analysis assumes that there is no heterogeneity between individual trials and all trials are sampling the same population of patients,38,39 this is not the case in the current study. Furthermore, despite obvious graphical publication bias, the Begg-Mazumdar and Egger tests did not reach statistical significance. As the power of these methods is low when small numbers of studies are included40 this is to be expected. Likewise, tests for funnel plot asymmetry when there are fewer than 10 studies in the meta-analysis may not be able to distinguish chance from real asymmetry.41.
Of the 3 common NOD2 polymorphisms associated with the development of CD, homozygosity for the L1007finsC variant is characterized by earlier onset and more aggressive disease compared with other variants.42,43 Three of the 6 studies defined surgical recurrence in relation to the 3 polymorphisms and so a subset analysis was performed comparing the expression of each polymorphism with re-operation rates. None of the 3 genetic variants were associated with surgical recurrence with odds ratios ranging from 1.00 to 1.65. A subset analysis of homozygosity versus heterozygosity of gene expression was not possible to perform as these data could not be extracted from the literature.
This study reviewed the currently available literature to determine if it supports the relationship between the expression of NOD2 polymorphisms and recurrence of CD requiring operative intervention. Although currently available data do not support NOD2 genetic polymorphisms as biomarkers of surgical recurrence in patients with CD, this is possibly because of the small numbers of patients genotyped across the studies and the heterogeneous nature of the populations studied. Therefore, it may be suggested the data support the conduction of a large multicenter homogenous study designed to ensure sufficient patient accrual as well as consistency in data reporting. It is noteworthy that the 58% cumulative increased risk of surgical recurrence is identical to the risk of index resection induced by being a NOD2 carrier.43 The current findings indicate that there are insufficient data to support a role for NOD2 in predicting surgical recurrence, but do not exclude a role for it. It is more likely that the reason no association with surgical recurrence was found is because of the small numbers of patients genotyped across the studies and the heterogeneous nature of the populations studied rather than a genuine lack of association.
The authors would like to thank Dr Jean-Paul Achkar and Prof Claudio Fiocchi (Dept. of Gastroenterology, Digestive Disease Institute, Cleveland Clinic Foundation, Cleveland, OH) for their critical review and comments on this manuscript.
1. Sachar DB. Recurrence rates in Crohn's disease: predicting the future and predicting the past. Gut. 2006;55:1069–1070.
2. Bernell O, Lapidus A, Hellers G. Risk factors for surgery and postoperative recurrence in Crohn's disease. Ann Surg. 2000;231:38–45.
3. Olaison G, Smedh K, Sjodahl R. Natural course of Crohn's disease after ileocolic resection: endoscopically visualised ileal ulcers preceding symptoms. Gut. 1992;33:331–335.
4. Rutgeerts P, Geboes K, Vantrappen G, et al.. Natural history of recurrent Crohn's disease at the ileocolonic anastomosis after curative surgery. Gut. 1984;25:665–672.
5. Lock MR, Farmer RG, Fazio VW, et al.. Recurrence and reoperation for Crohn's disease: the role of disease location in prognosis. N Engl J Med. 1981;304:1586–1588.
6. Hampe J, Grebe J, Nikolaus S, et al.. Association of NOD2 (CARD 15) genotype with clinical course of Crohn's disease: a cohort study. Lancet. 2002;359:1661–1665.
7. Cunningham MF, Docherty NG, Coffey JC, et al.. Postsurgical recurrence of ileal Crohn's disease: an update on risk factors and intervention points to a central role for impaired host-microflora homeostasis. World J Surg. 2010;34:1615–1626.
8. Ahmed T, Rieder F, Fiocchi C, et al.. Pathogenesis of postoperative recurrence in Crohn's disease. Gut. 2011;60:553–562.
9. D'Haens GR, Vermeire S, Van Assche G, et al.. Therapy of metronidazole with azathioprine to prevent postoperative recurrence of Crohn's disease: a controlled randomized trial. Gastroenterology. 2008;135:1123–1129.
10. Regueiro M, Schraut W, Baidoo L, et al.. Infliximab prevents Crohn's disease recurrence after ileal resection. Gastroenterology. 2009;136:441–450; e441;quiz 716.
11. Present DH, Meltzer SJ, Krumholz MP, et al.. 6-Mercaptopurine in the management of inflammatory bowel disease: short- and long-term toxicity. Ann Intern Med. 1989;111:641–649.
12. Prefontaine E, Macdonald JK, Sutherland LR. Azathioprine or 6-mercaptopurine for induction of remission in Crohn's disease. Cochrane Database Syst Rev. 2009:CD000545.
13. Siegel CA, Hur C, Korzenik JR, et al.. Risks and benefits of infliximab for the treatment of Crohn's disease. Clin Gastroenterol Hepatol. 2006;4:1017–1024; quiz 1976.
14. Hugot JP, Chamaillard M, Zouali H, et al.. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease. Nature. 2001;411:599–603.
15. Ogura Y, Bonen DK, Inohara N, et al.. A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease. Nature. 2001;411:603–606.
16. Inohara N, Ogura Y, Fontalba A, et al.. Host recognition of bacterial muramyl dipeptide mediated through NOD2. Implications for Crohn's disease. J Biol Chem. 2003;278:5509–5512.
17. Economou M, Trikalinos TA, Loizou KT, et al.. Differential effects of NOD2 variants on Crohn's disease risk and phenotype in diverse populations: a metaanalysis. Am J Gastroenterol. 2004;99:2393–2404.
18. Lesage S, Zouali H, Cezard JP, et al.. CARD15/NOD2 mutational analysis and genotype-phenotype correlation in 612 patients with inflammatory bowel disease. Am J Hum Genet. 2002;70:845–857.
19. Cuthbert AP, Fisher SA, Mirza MM, et al.. The contribution of NOD2 gene mutations to the risk and site of disease in inflammatory bowel disease. Gastroenterology. 2002;122:867–874.
20. Ahmad T, Armuzzi A, Bunce M, et al.. The molecular classification of the clinical manifestations of Crohn's disease. Gastroenterology. 2002;122:854–866.
21. Abreu MT, Taylor KD, Lin YC, et al.. Mutations in NOD2 are associated with fibrostenosing disease in patients with Crohn's disease. Gastroenterology. 2002;123:679–688.
22. Maconi G, Colombo E, Sampietro GM, et al.. CARD15 gene variants and risk of reoperation in Crohn's disease patients. Am J Gastroenterol. 2009;104:2483–2491.
23. Alvarez-Lobos M, Arostegui JI, Sans M, et al.. Crohn's disease patients carrying Nod2/CARD15 gene variants have an increased and early need for first surgery due to stricturing disease and higher rate of surgical recurrence. Ann Surg. 2005;242:693–700.
24. Buning C, Genschel J, Buhner S, et al.. Mutations in the NOD2/CARD15 gene in Crohn's disease are associated with ileocecal resection and are a risk factor for reoperation. Aliment Pharmacol Ther. 2004;19:1073–1078.
25. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188.
26. Dunne C, Burke JP, Morrow M, et al.. Effect of margin status on local recurrence after breast conservation and radiation therapy for ductal carcinoma in situ. J Clin Oncol. 2009;27:1615–1620.
27. Lau J, Ioannidis JP, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med. 1997;127:820–826.
28. Onnie CM, Fisher SA, Prescott NJ, et al.. Diverse effects of the CARD15 and IBD5 loci on clinical phenotype in 630 patients with Crohn's disease. Eur J Gastroenterol Hepatol. 2008;20:37–45.
29. Renda MC, Orlando A, Civitavecchia G, et al.. The role of CARD15 mutations and smoking in the course of Crohn's disease in a Mediterranean area. Am J Gastroenterol. 2008;103:649–655.
30. Polito JM 2nd, Childs B, Mellits ED, et al.. Crohn's disease: influence of age at diagnosis on site and clinical type of disease. Gastroenterology. 1996;111:580–586.
31. Satsangi J, Grootscholten C, Holt H, et al.. Clinical patterns of familial inflammatory bowel disease. Gut. 1996;38:738–741.
32. Bayless TM, Tokayer AZ, Polito JM 2nd, et al.. Crohn's disease: concordance for site and clinical type in affected family members—potential hereditary influences. Gastroenterology. 1996;111:573–579.
33. Colombel JF, Grandbastien B, Gower-Rousseau C, et al.. Clinical characteristics of Crohn's disease in 72 families. Gastroenterology. 1996;111:604–607.
34. Achkar JP, Duerr R. The expanding universe of inflammatory bowel disease genetics. Curr Opin Gastroenterol. 2008;24:429–434.
35. Hampe J, Cuthbert A, Croucher PJ, et al.. Association between insertion mutation in NOD2 gene and Crohn's disease in German and British populations. Lancet. 2001;357:1925–1928.
36. Ioannidis JP, Gwinn M, Little J, et al.. A road map for efficient and reliable human genome epidemiology. Nat Genet. 2006;38:3–5.
37. Little J, Higgins JP, Ioannidis JP, et al.. Strengthening the Reporting of Genetic Association Studies (STREGA): an extension of the STROBE statement. PLoS Med. 2009;6:e22.
38. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22:719–748.
39. Mahid SS, Hornung CA, Minor KS, et al.. Systematic reviews and meta-analysis for the surgeon scientist. Br J Surg. 2006;93:1315–1324.
40. Egger M, Davey Smith G, Schneider M, et al.. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634.
41. Higgins JP, Altman DG, Gotzsche PC, et al.. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928.
42. Seiderer J, Schnitzler F, Brand S, et al.. Homozygosity for the CARD15 frameshift mutation 1007fs is predictive of early onset of Crohn's disease with ileal stenosis, entero-enteral fistulas, and frequent need for surgical intervention with high risk of re-stenosis. Scand J Gastroenterol. 2006;41:1421–1432.
43. Adler J, Rangwalla SC, Dwamena BA, et al.. The prognostic power of the NOD2 genotype for complicated Crohn's disease: a meta-analysis. Am J Gastroenterol. 2011;106:699–712.
© Crohn's & Colitis Foundation of America, Inc.