Genetic risk profile predicts early CD surgery

May 13, 2016

Genome-wide association studies, which test single nucleotide polymorphisms (SNPs) in large populations for association with a specific disease, have revolutionized the search for genetic influences on complex conditions. This is especially true for Crohn's disease (CD) and ulcerative colitis (UC), which have been at the forefront of complex disease genetics for the last decade.

The first genome-wide association study of CD was published in Human Molecular Genetics in 2005. It evaluated 80,592 gene-centric SNPs in a small cohort of Japanese patients and controls.

Since then, completion of the HapMap, advances in technology and the rapidly declining cost of complete genomic sequencing have led to a growing number of genome-wide association studies and subsequent meta-analyses. These studies have identified multiple CD and UC susceptibility genes, some of which are common to both diseases and to other immune-mediated conditions.

One of the largest genetic analyses of inflammatory bowel disease (IBD) appeared in Nature in 2012. After evaluating 1.23 million SNPs in more than 75,000 cases and controls, researchers identified 71 new associations, increasing the number of confirmed IBD loci to 163. Most were found to contribute to both CD and UC, 70 percent (113 out of 163) were shared with other complex diseases and many overlapped with susceptibility loci for mycobacterial infections.

This study and others demonstrate the ability of genome-wide association studies to implicate previously unsuspected pathways in the pathogenesis of disease. The value of these findings in clinical practice is still unclear, but efforts are underway to convert them into clinical applications, says Ming-Hsi Wang, M.D., Ph.D., a gastroenterologist specializing in the epidemiology and genetics of IBD at Mayo Clinic's campus in Jacksonville, Florida.

One idea, he says, is to use genetic profiles to help identify which patients with newly diagnosed Crohn's disease might benefit most from aggressive initial therapy. "Right now, clinicians have limited ability to identify Crohn's disease patients who are at high risk of rapid disease progression and early surgery. So we set out to develop a model that would use both clinical and genetic factors to identify those patients," he explains.

To derive the model, Dr. Wang and colleagues used a large multicenter IBD cohort. They first analyzed clinical factors such as gender, age at diagnosis, family and smoking history, and location and behavior of the disease. Participants were then genotyped using the Illumina gene chip, and the researchers identified 437 SNPs associated with early surgery. They next searched for genetic risk profiles using logical regression, a machine learning algorithm, ultimately identifying three different profiles.

Of 582 patients with Crohn's disease, 246 — 42 percent — had early surgery. Early age at diagnosis, ileal disease, complicated CD, and one or more of the three genetic risk profiles were associated with surgery in these patients. The mean follow-up was 6.2 years.

"The predictability of clinical factors alone had an AUC of 0.72, which increased to 0.84 after we integrated the three genetic risk profiles," Dr. Wang says. "We then constructed a risk score for early Crohn's disease surgery based on the combined clinical and genetic predictors. The score had a sensitivity and specificity of 72 and 82 percent, respectively."

The study results were presented at the 2015 annual meeting of the American College of Gastroenterology and published in the October 2015 American Journal of Gastroenterology supplement.

"The old paradigm for treating IBD was a step-up strategy — starting with a milder agent and increasing to a stronger one, if necessary. But it became apparent that the longer inflammation was uncontrolled, the greater the likelihood of scar tissue formation, blockage and subsequent bowel resection. So the new approach is to be more aggressive at the very beginning of the disease course — the step-down strategy — to prevent bad consequences down the road.

"On the other hand, some patients have an indolent disease course and won't need an aggressive approach that carries with it a number of risks. This model can help guide treatment by determining which patients will need more-aggressive treatment sooner," Dr. Wang explains.

He believes the predictive profile can become clinically practicable once independently validated. "The cost of genetic sequencing has been tremendously reduced in the last few years; I would expect the profile to be clinically feasible and affordable for patient care," he says.

For more information

Yamazaki K, et al. Single nucleotide polymorphisms in TNFSF15 confer susceptibility to Crohn's disease. Human Molecular Genetics. 2005;14:3499.

Jostins L, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature. 2012;491:119.

Wang MH, et al. Integrating genetic risk profiles with clinical predictors can predict need for early surgery in Crohn's disease. The American Journal of Gastroenterology. 2015;110(suppl):S766.