Novel Diagnostic Solutions

Medical Research Council (MRC)

Dr Richard Oram was awarded an MRC CiC Award in order to develop and test a simple, cheap, new method for determining the genetic risk of Celiac disease (CD) to be used to help diagnosis of the disease.

The challenge

Celiac disease (CD) is a common disorder of the small intestine triggered by an immune response to gluten. A key to reducing the impact of the disease is early diagnosis and a gluten-free diet. Diagnosis is often delayed because patients start the gluten-free diet before CD is confirmed. Genetic testing is very helpful in situations where diagnosis is proving difficult (especially for children already eating gluten-free) but is currently expensive and so not widely used.

What was done to help

Dr Oram’s group developed and tested a simple, cheap, new method for determining genetic risk of CD that will be used to help diagnosis.

This study builds on a successful type 1 diabetes (T1D) GRS project with Randox. The project developed and validated a Celiac disease genetic risk score (CD-GRS) based on single nucleotide polymorphisms (SNPs) tagging the classical celiac HLA risk, secondary HLA alleles associated with celiac, and non-HLA genetic risk. It was developed using data from published meta-analyses of genetic risk in CD and tested in UK Biobank and Wellcome Trust Case Control Cohort (WTCCC).

Subsequently the group tested the ability of the CD-GRS to discriminate biopsy-proven celiac from non-celiac disease in a specialist paediatric celiac disease clinic, and it tested the combined predictive power of measuring aTTG antibodies and the CD-GRS against biopsy diagnosis in the specialist CD clinic. Finally the group developed a simple in-house SNP assay to measure celiac genetic risk, including direct from blood genotyping assays, and work with the commercial partner (Randox) to develop a biochip to offer the CD-GRS as a diagnostic test.

The results

Dr Oram’s group generated a new Celiac disease genetic risk score (CD-GRS) through designing new primers and re-genotyping samples. Using the large Wellcome Trust Case Control Cohort (WTCCC), the group has trained a CD-GRS that was highly effective at discriminating CD patients from healthy controls. They validated the CD-GRS in 1237 cases in the UK Biobank using the LCG KASP platform, achieving a ROC-AUC of 0.88 [95% CI: 0.87– 0.89]). The CD-GRS showed comparable CD discrimination in the clinical validation cohort (ROC-AUC=0.82 [0.75 – 0.90]). Simulations showed ~50% improvement in efficiency of Celiac disease prediction compared to traditional HLA testing.

Future Impact

CD-GRS is a gluten-independent marker that complements existing methods to discriminate CD in clinical populations, potentially leading to less costly, less invasive medical care. As SNP genotyping is cheaper than HLA typing, the CD-GRS may also improve general population prediction in public health settings to improve ascertainment and enable treatment during the asymptomatic phase of CD. The low cost of a SNP-based GRS in combination with the large potential market (particularly in North America) make the potential for future income from this very high.

Additionally, Exeter have strengthened their collaboration with the Pacific Northwest Research Institute, and refined a dual screening panel using the LGC screening platform that can be performed on neonatal dried blood spots to preform birth screening for CD and T1D genetic risk. There is sufficient scope to develop a combined GRS that could be used for birth cohort screening for both diseases.