HuGENet™ Case Study
NOD2 and Crohn Disease
How does this finding implicate NOD2 in Crohn disease?
Could NOD2 be associated with Crohn disease
-- in the 17 people who did not inherit the 3020insC variant from their parents?
-- in people from families without other affected members?
The NOD2 3020insC variant was highly associated with Crohn disease (CD) in families in which at least one parent had the variant (13% of all families in the study).
Other NOD2 variants—or other genes---could be involved in other cases of CD. The authors noted that because 365 of the 416 families had more than one affected member, the association of NOD2 3020insC with “sporadic” CD requires further study.
No conclusions can be drawn from these data about the relative risk for CD in people with the variant, or the fraction of CD cases attributable to the variant.
At minimum, what other information about cases and controls would be important to know?Answer 2
CD can be difficult to diagnose, and no agreement exists about diagnostic criteria. Therefore, knowing how CD was diagnosed in cases (ie, the case definition) and how it was excluded in controls is particularly important. More information is needed about the control populations: were they population-based or hospital-based? Were participants asked about family history of IBD?
Cases were from clinical populations studied at three major medical referral centers in Chicago, Baltimore, and Pittsburgh. Allele frequencies in controls from San Francisco were similar to those in Chicago and Baltimore. However, the German control group had a somewhat (though not statistically) higher frequency of the 3020insC variant. Could any of these people have had undiagnosed CD?
How different were the distributions of the NOD2 3020insC variant in cases and controls?Answer 3
The NOD2 3020insC allele was found among controls, but only about half as often as among cases. The allele frequency among Baltimore cases (6.8%) was about twice as high as in Baltimore controls (3.2%), but not very different from German controls (5.3%). These data demonstrate a clear association of the NOD2 3020insC variant with CD, but suggest it is neither a necessary nor sufficient cause.
How strong was the association between NOD2 3020insC and Crohn disease? Are heterozygotes at increased risk?Answer 4
NOD2 3020insC homozygotes were 18 times more likely than people without this variant to have CD, but accounted for only 11 of 416 cases. Heterozygotes appeared to be at slightly increased risk, but this result was not statistically significant (95% CI 0.8-2.5). Approximately 8% of controls were heterozygotes.
What assumptions are required to estimate an attributable fraction from these data? If the result is correct, what does it suggest about the role of NOD2 in Crohn disease at the population level?Answer 5
Using case-control study results to estimate attributable fraction is reasonable only if the cases and controls are comparable and population-based; we can't really assume this from the data provided. Although NOD2 3020insC is clearly a strong risk factor for CD (OR ~ 18 for homozygotes), it accounts for only a small proportion of cases, even in this selected group of families. Although heterozygotes did not have a significantly elevated OR, they were much more prevalent than homozygotes among cases, contributing to a more than double estimate of attributable fraction.
How did frequencies of these variants compare in the Crohn disease and unaffected control groups? How did frequencies of SNP 13 compare with the results of Ogura et al.?
All three variants were relatively rare but much more common in people with CD—among 936 chromosomes, 29% had at least one variant. SNP13 (=3020insC) was somewhat more common in this population of patients with CD than in Ogura et al. (12% vs. 8.2%) and less common among controls (2% vs. 4%), but overall, the findings were similar.
Were any of the three variants associated with ulcerative colitis? Why is this finding important?
None of these variants appear to be associated with ulcerative colitis (UC), because their frequencies were similar to those in controls. Ogura et al. reported a similar finding, which could help identify potential differences in pathogenetic mechanisms for CD and UC. These disorders have overlapping phenotypes, and drug therapy is mostly undifferentiated.
How were these NOD2 variants associated with Crohn disease?
People with one of these variants were 3 times more likely to have CD. In people with two variants (whether the same or different), risk was increased 40-fold. The authors suggested that this pattern could indicate a “recessive” trait, at least in this sample of cases, of whom half came from multiplex families.
How important are these NOD2 variants as risk factors for Crohn disease at the population level?
These three variants accounted for a sizeable proportion of CD cases in this study population. Nevertheless, most people with CD did not have even one of them. As in the other study, attributable fraction estimates should be regarded with caution. Without more information, we cannot assume that the cases are population-based or that the controls are drawn from the same population.
What is the estimated absolute risk (incidence) of Crohn disease by genotype?Answer 10
Assume that OR ~ RR and total population risk is 0.001.
Let x be absolute risk of CD in people without any of the NOD2 variants (wild type).
Apportion population risk using prevalence and risk data from the table on previous page:
+ (15/103)(3) x
+ (0.22/103)(40) x
+ (0.29/103)(40) x
Solving for x =0.0007, we can estimate an approximate absolute risk for each group:
|Genotype||(0.7)(OR)||absolute risk per 1000|
Considered together, what do the findings of these three studies mean for researchers investigating the genetic basis of Crohn disease? For people with one or more of these NOD2 variants? For people with Crohn disease?Answer 11
Together, these findings represent a step forward in understanding the genetics of CD. However, there is more work to do. Additional population-based studies need to confirm these findings in patients with “sporadic” CD (i.e., not drawn from families with multiple affected members), comparing them with carefully chosen control groups. Both Ogura and Hugot noted that genetic susceptibility to CD is not limited to chromosome 16; at least five other loci have been implicated and await further elucidation. Additional studies should be designed to address the interaction of NOD2 (and other potential susceptibility genes) with environmental exposures in the pathogenesis of CD.
For people with one or more of the implicated NOD2 variants:
These studies show that even people homozygous for these variants have a better-than-even chance of remaining free of CD. For now, there is nothing to be gained by testing family members of affected people for susceptibility to CD.
For people with Crohn disease:
Eventually, better understanding of pathogenetic mechanisms is likely to benefit people with CD by permitting more efficient clinical trials and suggesting more specific therapeutic interventions. Epidemiologic studies of gene-environment interaction may eventually help identify ways to prevent flare-ups in people with CD, and perhaps even prevent disease onset in some susceptible people.