The amazing powers of exome sequencing – a disclaimer. We have recently blogged frequently on the power of exome sequencing in monogenic disorders. Dixon-Salazar now describe in “Exome Sequencing Can Improve Diagnosis and Alter Patient Management” the usefulness of exome sequencing in disease identification in autosomal recessive disorders. Their overall yield is a novel gene discovery in 22/118 probands and a different diagnosis than the initial in 10/118 patients. While title and abstract suggest that exome sequencing is a cure-all improving patient diagnosis and altering patient management, it should be pointed out that this manuscript exclusively deals with autosomal recessive disorders. Only two novel genes out of 20 are described, leaving the reader with little chance to investigate their claim. Many of their families were selected from countries with a high consanguinity including Morocco, where state-of-the-art diagnostic facilities are difficult to access for some patients. The only change in patient management resulting from the altered diagnosis was stopping supplementary Vitamin E in a family with a SPG11 mutation previously thought to have ataxia with vitamin E deficiency. What the altered direction of therapy in a family with a newly identified a-mannosidase type 1 entails, is left for the reader to imagine. The corresponding reference refers to a paper on stem cell transplant as a definitive treatment option, which will probably not be a treatment option for this family from Islamabad, Pakistan. The paper rather shows that exome sequencing is of use in autosomal recessive disorders and might yield surprises.
The burden of neurodevelopmental disorders. Dixon-Salazar and coworkers provide a strong, motivated plea for research in neurodevelopmental disorders in the introduction of their paper. Neurodevelopmental disorders including childhood epilepsies, autism and intellectual disabilities account for up to 10% of the health care expenditures in the US. Diagnostic procedures are usually low yield and easily exceed $10,000 per patient. Analysis of copy number variations are basically the only investigation in this patient cohort which has a reasonable a priori chance of identifying an underlying cause unless a specific diagnosis is suspected.
Study design. The authors used a straightforward study design, which is shown in Figure 1. In total, their cohort comprised 188 families, of which 40 had mutations in known genes. In addition, 30 families had a single linkage peak, which were studied separately. Neither the genes identified nor the linkage peaks are mentioned. In total 118 families were included, linked either through separate homozygosity mapping or through mapping using exome data. In 22/118 families, novel genes were identified (not mentioned) and in 10/118 families, known OMIM genes were identified for genes different than those associated with the suspected disease.
A diagnosis that wasn’t… The 10 families with altered diagnosis span a range of very rare neurogenetic disorders that were previously diagnosed by the referring physicians with various degrees of certainty. In these families, genes for known OMIM disorders were identified through exome sequencing and except for one mutation, all mutations had not been described previously. In each case, the authors tried to include additional clinical data in order to get more information about the phenotypes. For example, family 890 from Turkey was initially diagnosed with pontocerebellar hypoplasia. Pontocerebellar hypoplasia (PCH) is a very severe congenital disease, in which the cerebellum and the pons, a part of the brain stem, are not properly developed. Affected children have various symptoms ranging from muscular hypotonia (floppiness), severe developmental delay, swallowing difficulties and epilepsy. Most cases of PCH are due to mutation in TSEN54, which was excluded in this family. However, a mutation in VLDLR was found in this family. Autosomal recessive mutations in VLDLR are known to cause VLDLR-associated cerebellar hypoplasia, a disease so for only described in the Hutterites. However, re-examination of the MRI showed findings compatible with VLDLR-associated cerebellar hypoplasia, even though the clinical features were different from previously published cases.
Exome data to identify known genes in epileptic encephalopathies. Even though none of the families published by Dixon-Salazar is a classical epilepsy family, this study provides us with an estimate of the burden of atypical or “hidden” genetic disorders in neurodevelopmental disease. It can be hypothesized that a similar fraction might be true for severe epilepsies or epileptic encephalopathies, suggesting that large sequencing projects including EuroEPINOMICS and Epi4K will identify known disorders with atypical presentation in up to 10% of cases, either through exome sequencing or through candidate gene panels. Consequently, a certain fraction of epileptic encephalopathies may be unidentified known autosomal recessive disorders.