Caution with Celiac Genetic Testing Plus Two

Conventional wisdom and previous studies have indicated that negative testing for HLA-DQ8 and HLD-DQ2.5 alleles makes a diagnosis of celiac disease (CD), now or in the future, very unlikely.  While ~60% of the population has one of these alleles, testing negative for these alleles has been regarded as having a high negative predictive value (>99%) and can be valuable in cases of equivocal diagnosis.

The authors of recent report (F Fernandez-Banares et al. Clin Gastroenterol Hepatol 2017; 15: 594-96) challenged this wisdom, noting that there is expected to geographical variation in the presence of these alleles. The goal of their study was to assess the prevalence of HLA-DQ2.5/8 among CD patients in Spain by reviewing previous studies; 12 studies were included. To be included, patients had to have villous atrophy, positive serology and available genotyping.

Key finding:

  • Among 2963 Spanish CD patients, 3% “might be negative for HLA-DQ2.5/8.”

This is a brief report.  It is expected that limitations would relate to the accuracy of genotyping and of excluding other causes of villous atrophy.

My take: (from the authors) “This information highlights the need to be cautious when ruling out CD only on the basis of genetics.”

Related blog posts:

Briefly noted:

L Kivela et al. J Pediatr 2017; 183: 115-21.  This study divided children with CD into those identified via screening (n=145) and those identified due to clinical symptoms (n=359). Key findings:

  • There were no differences in serology or histology between the two groups
  • More than half (51.8%) of screen-detected patients had symptoms at diagnosis, but typically these were milder than in the clinically-detected group.
  • Anemia was more common in the ‘clinical group’ 22.9% vs 7.1% (screen group) as was poor growth (36.9% vs. 15.7%).

AJ Irvine et al. Am J Gastroenterol 2017; 112: 65-76. (Thanks to Ben Gold for this reference) In this systemic review with 15,256 individuals (& 9,275 with irritable bowel), “prevalence of positive celiac serology and biopsy-proven CD was significantly higher in subjects with symptoms suggestive of IBS vs. healthy controls.”  The odds ratio for serology-positive and/or biopsy-proven CD ranged from 2.75 to 4.48, though there was no significant increase in these ORs for North American studies.

Palace of Versailles


Why the Genetics of Inflammatory Bowel Diseases Matter Now

A terrific update on the genetics of inflammatory bowel diseases (DPB McGovern, S Kugathasan, JH Cho. Gastroenterol 2015; 149: 1163-76) explains why and how this information matters right now.  The article is a little difficult to read due to its review of highly technical material.

Here’s what I think were the key points:

  • Big advances in understanding the genetics started with the first genome-wide association studies (GWAS) using genome-wide single nucleotide polymorphisms (SNP) chips in 2005.  “The conceptual basis of GWAS is that most complex (ie, not single-gene Mendelian) genetic disorders are polygenic, being driven by multiple common genetic polymorphisms.”
  • “Early GWAS identified the most significant loci.”  Now, more than 200 loci associated with IBD have been identified with GWAS and Immunochip data. Table 1 lists these loci over 4 pages.  About 2/3rds of these are associated with Crohn’s disease (CD) and ulcerative colitis (UC) whereas the remaining 1/3rd are unique to either CD or UC.
  • These loci provide insight into disease mechanisms. NOD2 mutations result in “impaired activation of NF-κB” This supported “the general concept that deficiencies of innate immune cell function represent a central factor in Crohn’s disease, distinguishing it from ulcerative colitis.”
  • ATG16L1 gene mutation “establish the fact that the CD risk allele is correlated with impaired autophagy.”  This is leading directly into treatment efforts.
  • IL23R.  “The most significant association is Arg381Gln…confers a 2- to 3-fold protection against development of IBD.”  The protective effect is thought to be due to “decreased numbers of interleukin (IL)-23 dependent CD4+ Th17 and CD8+ Tc17 cells…decreasing IL-23 signaling, such as through monoclonal antibody blockade of anti-p40 or anit-p19 may be beneficial.”
  • FUT2 mutations.  These mutations affect the mucus layer in Crohn’s disease.
  • Studies in non-Caucasians highlight other susceptibility regions.
  • “Currently, sequencing of the whole exome has become not only a practical method but also a cost-effective option to identify functionally relevant variants in the protein encoding regions of the genome.”

Very Early Onset IBD:

  • Whole exome sequencing (WES) identified XIAP (X-linked inhibitor of apoptosis) in a case of boy with very early onset (VEO) IBD.  XIAP is a positive regulator of NOD2 function.  WES has also identified FOXP3, and IL10RB genes.
  • “The VEO group experiences a more severe disease course and more frequently shows a positive family history for IBD in support of higher genetic load.”  Table 2 lists ~40 genes associated with VEO.  These genes are involved in epithelial barrier function, neutropenia/defects in phagocyte function, hype-and autoinflammation, and  regulatory T cells and immune regulation.

Genetic Testing Will Impact Current Therapies and Help Explain Extraintestinal Manifestations:

  • Currently testing for TPMT variations is recommended prior to use of thiopurines due to concerns of toxicity in individuals with decreased metabolism of these medications.  However, genetic testing can identify other individuals with propensity to leukopenia (eg. NUDT15 polymorphism) and those with increase risk for pancreatitis (eg. HAL-DQA1-HLA-DRB1)
  • Primary Sclerosing Cholangitis (PSC) is associated with numerous genetic loci as well. PSC “genetically is more similar to UC than to CD.” Most other extraintestinal manifestation studies have been underpowered.
  • IBD share more genetic similarity to spondyloarthropathy (SpA) than any other immune-mediated diseases.  “The vast majority of shared susceptibility loci are concordant between IBD and SpA.”
  • With regard to psoriasis, the genetic relationship to IBD is complex.  Anti-TNF agents can cause psoriaform lesions in IBD patients.  In addition, anti-IL17a therapy, “so successful in psoriasis, appears to worsen Crohn’s disease” but not in those with a TNFSF15 variant.  Specific genotyping may help identify which patients with CD are susceptible to psoriaform lesions and those who may improve with therapy typically given for psoriasis.

My take: This article shows how understanding genetics of IBD is providing insight into pathophysiology and more personalized treatment approaches.

Briefly noted: EM Stoffel, CR Boland. Gastroenterol 2015; 149: 1191-1203. Excellent review of the genetics and genetic testing for Hereditary Colorectal Cancer.  The review includes polyposis syndromes and Lynch syndrome.

Atlanta Botanical Gardens

Atlanta Botanical Gardens

Understanding chromosomal microarrays

The advantages and disadvantages of chromosomal microarrays are highlighted in a recent article and a related editorial (NEJM 2012; 367: 2175-84, 2249-51).

Chromosomal microarrays can detect almost all of the chromosomal imbalances detected with conventional cytogenetic analysis.  They are recommended as a first-tier test for postnatal developmental delays, autism spectrum, or with multiple congenital anomalies.  Clinically significant findings occur in 15% of those with normal conventional karyotypes.

Specific advantages of microarrays:

  • Higher resolution
  • Faster turnaround due to automation
  • Does not require dividing cells (useful in fetal death)
  • Eliminates need to culture amniocytes or chorionic villi

Major downside:

  • In a few percent, may detect copy-number variant of uncertain clinical significance
  • Increased cost

The referenced study compared chromosomal microarray to karyotyping and enrolled 4406 women at 29 centers who were undergoing prenatal diagnosis.  Indications for screening included advanced maternal age (47%), abnormal result on Down’s syndrome screening (19%), structural abnormalities on ultrasonography (25%), and other indications (9%) (some had multiple indications).

The microarrays were of two varieties.  The first microarray (71% of cases) consisted of a fourplex array with each array consisting of 44,000 oligonucleotide probes.  The second platform (29% of cases) contained 1.8 million  oligonucleotide probes.

Key Findings:

  • Microarray identified all of the abnormalities on conventional karyotyping except for balanced translocations (no loss of genetic material).
  • There were 94 of 3822 fetal samples with copy-number variants of uncertain clinical significance (see blog reference below).  Subsequently, 30 of these were classified as pathogenic and 8 as benign.
  • In samples with normal karyotypes, microarray identified clinically relevant findings in 6% of those with a structural abnormality and 1.7% of those with advanced maternal age or positive Down’s syndrome screening.

The implications of this study and related studies are that microarray has a role in prenatal evaluation of structural abnormalities.  Ultimately, whole-exome sequencing will likely supersede microarray analysis and further the conundrum of interpreting abnormalities of uncertain clinical significance.

Related blog entry:

“It is never boring to be a physician”

…”because patients are so different.  Each has a story to tell.” This is the beginning of an excellent editorial (NEJM 2012; 367: 1350-52).  The editorial helps translate a study about “Phenotypic Heterogeneity” due to copy-number variants (NEJM 2012; 367: 1321-31).  The study itself involved analyzing 32,587 samples from children with developmental delay.  Ultimately, the study focused on 2312 children with 72 copy-number variants.

Both the study and the editorial try to explain why individuals with the same genetic defect have variability in the severity/expression.  Some of the factors include environment, chance, and modifier genes.  Due to the availability of relatively cheap and quick technology, one can sequence all of the 22,000 human genes.  The techonology for this study used microarray-based comparative genomic hybridization to interogate the entire genome.

Copy-number variation refers to changes in the number of genes compared to normal.  Typically, two gene copies are present.  Thus, a deletion or duplications will result in copy-number variants.

In the aforementioned study, investigators found that large copy-number variants contributed to differences in outcomes.  That is, persons with “the same chromosomal abnormality may have very different clinical outcomes” because they a have a second genetic event (copy-number variant elsewhere) that “makes matters worse for them.”

These copy-number variants contribute to learning disabilities.  Large copy-number variants occur in less than 1% of the unaffected population but in 8.6% of children with learning disabilities.  Males are m much more susceptible to have developmental delay due to these copy-number variants and unaffected women are more likely to transmit these copy-number variants.  These gender differences are thought to be due to the fact that males have only a single X chromosome and are more vulnerable to genetic insults as a consequence.

Besides learning disabilities, copy-number variants have been implicated in well-known diseases such as schizophrenia, autism, cardiac disease, and epilepsy.  In addition, well-defined syndromes, like Smith-Magenis syndrome, Williams-Beuren syndrome, Sotos syndrome, and MAPT (17q21.31) deletion syndrome, are also well-recognized as being associated with de novo copy-number variants.

It really is an exciting time to be a physician.  Widely available genetic tests can finally explain a lot of previously unanswered questions.