A few months ago the American College of Medical Genetics (ACMG) and the Clinical Genome Resource (ClinGen) published new technical standards for the classification of copy number variations (CNVs). These guidelines aim to assist clinical laboratories in the interpretation and reporting of CNVs, but are highly detailed and pose serious computational challenges. Meeting the many guideline requirements is almost impossible without a computational assistant.
In order to meet these new needs, we at Genoox have upgraded , our cloud-based software platform for genetic analysis, to align with the new standards, so that Franklin now fully integrates the new standards for CNV classification into each and every analysis.
Since this is a comprehensive topic, which has been covered by many articles and press releases over the past several months, we decided to address it in a series of posts. In this first post, we start by providing an overview of the published paper, highlighting the points that we found most important. In the next two posts, we will discuss the technical aspects of performing analyses according to the new CNV classification scheme, and later we will provide an in-depth review of the integration efforts and upgrades to our platform that ensure that Franklin is consistent with these new standards.
It’s time for new guidelines…
The last classification guidelines for CNVs were released almost a decade ago in 2011. Since then, new developments, resources and technologies have transformed the field of genetics and variant classification, generating an urgent need for new standards.
Today CNV testing is recommended as a first-tier approach for the postnatal evaluation of individuals with neurological disorders such as congenital anomalies, developmental delay, autism spectrum disorder and more, as well as for prenatal evaluation of fetuses with structural anomalies observed by physicians. While chromosomal microarray (CMA) is still the most commonly used technology for CNV detection around the world, in recent years there has been a considerable increase in the use of next-generation sequencing (NGS) as well.
A major problem, however, with CNV testing is the sheer complexity in classifying its outcomes. This is rooted in various challenges that are intrinsic to CNVs – including the uniqueness of each CNV and the involvement of multiple genes, which, in turn, lead to great discordance in variant classification, as is revealed in public databases such as ClinVar and DECIPHER.
The ACMG and ClinGen’s recently published new technical standards for CNV classification are an attempt to remedy these discordance problems. By assisting clinical laboratories in the interpretation and reporting of CNVs, these guidelines represent a major step forward – toward a global standardization in CNV classification.
The paper focuses on the following key points:
- The importance of a global classification scheme for CNVs that is uncoupled from the phenotype, sex, and zygosity of any individual case.
- A CNV classification scheme based on a quantitative and evidence-based scoring framework, with five categories that are widely used in sequence variant classification: pathogenic, likely pathogenic, uncertain significance, likely benign, and benign.
- The CNV classification scheme should be implemented irrespective of the sequencing technology used.
-png.png?width=1285&name=Classification%20table%20Copy%20(1)-png.png)
The table above demonstrates a rule of thumb for the classification of a CNV, in cases where the only assessed parameters for the region are:
- Level of evidence of ClinGen Dosage sensitivity of the assessed region
- CNV results in Gain or Loss of function
Additional details regarding the analyzed case such as observed type of inheritance are also considered during the evaluation process, and we would drill down to the classification process in our second post of this series.
Classification vs. Clinical Significance & Reporting
The current method of CNV classification presents multiple obstacles to building an objective evidence-based classification system.
Prior to these new standards, most clinical labs for genetic analysis had discretion over the classification of CNVs, and their decisions were often tightly bound to the presentation of clinical symptoms in individual cases. For example, the same CNV in two different patients could be classified as benign in one patient who shows no clinical symptoms, but as pathogenic in the other who does show clinical symptoms. This system resulted in great discordance in variant classification and hindered the development of a global standard.
The new guidelines are an attempt to concrete \ standardize the classification process in order to improve the consensus between involved parties (labs)
In contrast, the authors of this paper stress the importance of classifying CNVs regardless of the analyzed phenotype, sex, zygosity, and age of the individual case. That is, there should be a clear distinguishing line between the clinical significance of a variant in terms of patient diagnosis, and the classification of the variant, which should be global and case agnostic.
For example, when a variant that has already been classified as pathogenic (according to prior large-scale, evidence-based determinations) is found in a patient who shows no clinical symptoms related to that variant, the new guidelines insist that this variant should be classified as pathogenic here as well, regardless of the lack of correlation to phenotype in the individual case at hand.
Below you can see some reporting examples for a pathogenic variant
Pathogenic causal variant
Example for incidental findings
Pathogenic Carrier variant
Examples when reporting of variants differs in accordance to the case content:
- Pathogenic variant in a recessive gene while zygosity of case is heterozygous
- X-linked conditions in female carriers
- Pathogenic variant for a disease that occurred or will occur in a different time point than time of testing
- Secondary findings such as case with a developmental delay, while a CNV causes to other conditions such as predisposition for cancer was detected
While the classification of CNVs should be uncoupled from a patient’s clinical symptoms, this is not the case for reporting, where current and potential clinical symptoms do play a large role.
In general, each lab has the discretion to decide whether or not to report carrier alleles or variants that are not related to the primary reason that the patient sought testing. If the lab does decide to report these secondary (or incidental) variant findings, they should be presented as clearly separate from the primary findings and reported in a separate section. The paper’s authors provide guidance regarding when they do recommend reporting carrier status, such as, for example, in cases where a patient’s carrier status may not have clinical consequences for the patient themselves, but may affect future offspring.
Next up…
We hope that this overview and update was useful. In our next two posts, we will explore some of the major technical challenges involved in classifying CNVs in accordance with the new guidelines, and the implications of the integration of these new standards for the upgraded Franklin platform.
