Titel: Why we should re-analyse – Insights from 152 cases with developmental disorders
ID: P-ClinG-038
Art: Postertalk
Redezeit: 2 min
Session: Poster Session
ClinG 2

Referent: Tobias Bartolomaeus (Leipzg/DE)

Abstract - Text


Clinical guidelines recommend iterative re-analysis in undiagnosed cases. However, re-interpretation and reporting results considering novel data is not well investigated. Thus, we re-evaluated the results of a cohort of 152 consanguineous families with children with developmental disorders that we have reported five years ago.

In 2017, we reported 62 variants in 58 established genes in 61 families, as well as 53 variants in 53 candidate genes in 49 families. The remaining 43 families were negative. We re-evaluated all previously reported variants according to updated classification guidelines for genetic diagnostic or according to our internal candidate gene scoring system for research aspects. All sequencing data was re-processed using up-to-date tools, references, and databases for case-level re-analysis.

In 28/152 (18%) families, re-evaluation and re-analysis led to a clinically relevant change: In 13 families, previously reported (likely) pathogenic variants were re-classified as VUS or benign. Previously reported (likely) pathogenic variants in genes TSEN15, NAPB and FAR1 in three families were re-classified since information on gene-disease validity is limited. In 12 families, we identified 12 disease causing variants that were previously missed. we recommend filtering for low frequency variants that are a) already reported in HGMD or ClinVar, b) truncating, or c) affecting the same amino acid position as a known missense pathogenic variant. With this approach 10/12 (83%) of the previously missed disease causing variants in the here presented cohort can be identified. Two previously reported variants were missed by the updated computational pipeline due to alignment (old data) or reference (ambiguous region in hg38) issues. We submitted all relevant variants to public databases and revised their previous classification.

Our results support the need to re-evaluate research screening studies, not only of negative cases, but also of those that are supposedly solved. We highlight potential benefits and pitfalls of computational re-analysis. Its complexity for old data should be weighed against the decreasing re-testing costs. Since extensive re-analysis, where all variants of an individual are evaluated, is beyond the resources of most institutions, we present our procedure of few easy and automated steps that would re-classify most variants correctly and would identify the majority (10 of 12) of missed variants.