All articles

Reclassifying a VUS is a process.

A variant of uncertain significance is an open question with a list of things that would answer it. Most are filed as answers and never worked on again.

A woman of thirty-four with a strong family history of early breast cancer has a variant of uncertain significance in a high-penetrance gene. The report explains what the term means and advises that management be based on family history. It does not say what would resolve the uncertainty: that her affected mother and aunt could be tested for the variant, that a functional assay exists for this domain, or that the laboratory will look again when the population data are next released.

Five years on, the variant is still uncertain in the laboratory's records. Nobody tested the mother before she died, and the aunt has since moved abroad. The functional assay was published two years ago. The woman has had two surgical consultations in which the variant was described as a result nobody could interpret.

What the field knows

The uncertain classification is a statement about evidence, and the framework that produces it lists the evidence that would move it. Under the ACMG and AMP guidelines (Richards et al., 2015) a variant is uncertain when the criteria met for and against pathogenicity do not reach either threshold, and each unmet criterion names a kind of evidence: segregation in affected relatives, a de novo observation, a functional result, population frequency, additional reported cases. A variant of uncertain significance therefore comes with an implicit work list, and the guidelines expect laboratories to revisit it as evidence accumulates.

Most uncertain variants are eventually resolved towards benign, and the ones that go the other way matter most. Mersch and colleagues (2018) found that the great majority of reclassifications in a hereditary cancer series were downgrades, with a small number of upgrades that changed management. The work that produces either outcome is specific: Harrison and Rehm (2019) note that much of the reclassification activity in the classification archive follows new population data and shared case-level evidence between laboratories, and the ClinGen expert panels resolve uncertain variants by applying gene-specific rules and by gathering segregation and functional data the original laboratory did not have.

Families are part of the evidence. Segregation testing of affected relatives is one of the strongest and cheapest ways to move a variant, and it is time-limited: an affected parent or aunt who could have been tested this year may not be available next year. Functional assays are increasingly available at scale (Findlay et al., 2018, for one high-penetrance gene) and a laboratory that knows one exists can use it. The reanalysis guidance (Deignan et al., 2019) places the laboratory's re-evaluation of its own uncertain variants within the laboratory's responsibilities, with the proviso that it be systematic.

The uncertain variant also has a human cost that the term hides. Studies of patients receiving such results describe confusion, anxiety and, in some cases, management decisions made as if the variant were pathogenic. A report that gives the uncertainty a shape, what it would take to resolve it and who can help, is a different experience from one that gives it a name.

Where practice breaks

The uncertain result is reported as a terminus. The report explains the term, advises management on other grounds, and stops. The reader, clinician or family, is left with a result that has no next step, because none was written. The term itself is honest. The silence after it, with no owner and no date, is where the process fails, and it fails most often for the families least able to chase it.

The work list is never made. The scientist who classified the variant knew, at the moment of classification, which criteria were unmet and what evidence would meet them. That knowledge is not recorded, and the next person to look at the variant has to reconstruct it from the classification worksheet, if there is one.

The time-limited evidence expires. Affected relatives die, emigrate or lose touch. A segregation study that was possible at the time of reporting is impossible five years later, and the uncertain variant becomes permanently uncertain because the one piece of evidence that would have resolved it was not asked for in time.

Re-evaluation is unsystematic. Uncertain variants are revisited when someone asks, when a family returns or when a scientist happens to notice a new paper. A laboratory with thousands of them cannot review each against each new release by memory, and does not. The variants that are revisited are the ones with the most persistent families, which is a poor way to allocate the work.

How Tulparus handles it

An uncertain classification in the workbench carries its work list as structured items. From the criteria met and unmet, the case records what evidence would move the variant in each direction: segregation in named affected relatives from the pedigree, a functional assay if the knowledge store has one for the gene or domain, additional cases through the matching networks, the next population data release. Each item has an owner, a route and, where it applies, a time limit.

The report, rendered from the case, says what would resolve the uncertainty. For the woman above it would have named her affected mother and aunt as the people whose testing would be most informative and advised that it be arranged while it could be. A counsellor reading that report has something to do and a reason to do it now.

Re-evaluation is driven by releases. When population data, the classification archive or the functional assay literature is updated, the workbench matches the change against every uncertain variant on every case and lists the ones with new evidence, with the item on the work list that the evidence addresses. The functional assay published two years ago would have reached the woman's case the week it entered the knowledge store.

Each re-evaluation is a new version of the classification with its evidence, and the earlier versions are kept. A variant that stays uncertain through three releases shows what was checked each time, which is a different thing from a variant nobody looked at, and the report can say which it is.

What you can do tomorrow

The work list can be written by hand for every uncertain variant reported from now on.

  • When you classify a variant as uncertain, write down the two or three pieces of evidence that would move it, and put them in the report.
  • Name the relatives whose testing would be informative and say that it is time-limited. The counsellor can act on that; 'segregation data may be helpful' they cannot.
  • Keep a register of uncertain variants by gene, and check it against each population and classification release, or at a fixed interval if that is all you can manage.
  • When a functional assay is published for a gene you report, search your register for uncertain variants in it the same week.

Sources

  1. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genetics in Medicine, 2015.
  2. Mersch J, Brown N, Pirzadeh-Miller S, et al. Prevalence of variant reclassification following hereditary cancer genetic testing. JAMA, 2018.
  3. Harrison SM, Rehm HL. Is 'likely pathogenic' really 90% likely? Reclassification data in ClinVar. Genome Medicine, 2019.
  4. Findlay GM, Daza RM, Martin B, et al. Accurate classification of BRCA1 variants with saturation genome editing. Nature, 2018.
  5. Deignan JL, Chung WK, Kearney HM, Monaghan KG, Rehder CW, Chao EC. Points to consider in the reevaluation and reanalysis of genomic test results: a statement of the American College of Medical Genetics and Genomics (ACMG). Genetics in Medicine, 2019.

One uncertain variant, three ways to settle it.See the work list on the case.

A demo takes 20 minutes, screen shared, on synthetic cases. Bring the question from one of your own.