Classification: assessing variant severity

The American College of Medical Genetics and Genomics (ACMG) has developed standards and guidelines for assessing the likelihood that a variant in a particular gene is cause of a person’s illness.  Recently, revised guidelines have been published (Richards et al, 2015). Clinical geneticists call this variant assessment process "classification". 

 

The RIKEE curators have worked together to adapt ACMG guidelines for assessing KCNQ2 and KCNQ3 variants.  Very important questions we have tried to address are:

 

(1) How likely it is that a KCNQ2/3 variant causes illness?

(2) If the KCNQ2/3 variant is likely to cause illness, how likely is it that a person with that variant will be affected by persistent developmental challenge or disability?

 

We have used the terms “self-limiting/BFNE” to describe outcomes where seizures are mostly confined to the neonatal period, and developmental outcome is within typical limits.  We have used the term “epileptic encephalopathy” to describe outcomes where epilepsy after the neonatal period AND varying degrees of developmental impairment (such as difficulties in language, other social interaction and cognition, and body control) are expected.  These terms are broadly used, but we know they are not perfect - in a small proportion of the variants, there may be an overlap in some families or some unrelated individuals.  For variants causing epileptic encephalopathy, we have not attempted to distinguish between mild, moderate, and more severe degrees of developmental impairment. We have used the following terms based on ACMG guidelines:

• non-pathogenic

• likely non-pathogenic

• uncertain pathogenicity

• likely pathogenic

• pathogenic

 

Steps in assessing each variant involve considering many factors, including:

• population genetics (has the variant been found in surveys of large groups of people who are asymptomatic?)

• family history (is the variant found in two or more symptomatic people in the family? in any people without symptoms?)

• molecular type of variant (missense variants are more likely to cause epileptic encephalopathy, other variant types almost always cause BFNE)

• location of the variant in the predicted 3-dimensional structure of the channel protein

• studies of the function of the variant performed on cultured cells or animals

• whether the symptoms of the patient are similar to other KCNQ2/3 patients

 

The actual process we have developed is explicit and very detailed.  This means we can continue to assess whether our conclusions are correct, and modify specific steps to improve accuracy and usefulness. 

 

To see a fully detailed view, click here: