(152) My patient has an unresolved CFTR genotype…what next?

Background: Most individuals with cystic fibrosis (CF) have 2 causal variants in separate CFTR alleles identified via genetic testing using a variant panel or sequencing of the coding and flanking intronic regions. Despite improved genetic testing technology and increased understanding of rare CFTR variants, a small subset of those with a clinical diagnosis of CF are left with an ‘incomplete genotype’ – 1 or 0 identified causal variants – and thus have uncertainty regarding prognosis, treatment, and genetic counseling. We sought to consider options for such cases, report on retrospective and prospective in-depth genetic analysis, and suggest next steps for evaluation.

Methods: A randomly selected subset of individuals with sweat [Cl-] >60 mmol/L enrolled in the Cystic Fibrosis Foundation Mutation Analysis Program (MAP; n=4,479 total enrollees), who had undergone exonic CFTR sequencing and deletion/duplication analysis that identified 1 (n=20) or 0 (n=20) causal CFTR variants, underwent full-gene sequencing of CFTR. Prospective analysis included upstream, downstream, and intronic regions. Additionally, 49 individuals with CF from the CF Genome Project (CFGP; 0.97% of the cohort [n=5,058]), who had whole genome sequencing (WGS) with 1 or 0 causal CFTR variants identified, underwent retrospective phenotype review, investigation of non-coding regions of CFTR, and interrogation of other genes. We searched for loss-of-function variants in all protein-coding genes and screened 866 genes implicated in CF-like phenotypes, the CFTR interactome, or CFTR biogenesis and functional pathways.

Results: After full-gene CFTR sequencing, 9 of 20 individuals (45%) from MAP with 1 known causal variant had a second putatively-causal variant identified: 6 intronic (pathogenic: c.1680-877G >T, c.3874-4522A >G; likely pathogenic: c.2909-15T >G, c.3469-1304C >G; VUS: c.1584+689G >A, c.2490+1083C >T) and 1 pathogenic partial exon deletion (c.3325_3367+249del). No putatively-causal variants were found in those with 0 previously-identified causal CFTR variants. Of 49 CFGP subjects considered, 43 were excluded because their CFTR genotypes contained a non CF-causing variant previously considered causal (n=24 subjects), the genotype reported in clinical records was not confirmed by WGS (n=5), or a CF phenotype was not confirmed (n=14). Of the 6 remaining subjects, siblings (n=2 [33%]) with 1 known causal CFTR variant were found to have a putatively-causal deep intronic variant (c.3874-4522A >G); no variants of interest were detected in the other 4 individuals. The screen of other genes revealed no other variants that would explain the phenotype of these individuals.

Conclusions: Individuals with a clinical diagnosis of CF and 1 known causal variant may benefit from full-gene CFTR sequencing, which can detect deep intronic and complex structural variation. Individuals with no causal variants after typical clinical testing are unlikely to benefit from further interrogation of CFTR, though ancestry or consanguinity may impact decision-making regarding further testing. A small fraction of those with clinical features of CF and an ’incomplete genotype’ may be symptomatic CF carriers, epigenetic cases, or phenocopies. Whole genome long-read sequencing may provide further insight. The large number of individuals in the CFGP excluded (n=43 of 49 subjects) suggests that clinicians should frequently review updated interpretations of CFTR variants and consider whether a patient’s phenotype remains consistent with CF.

Acknowledgements:

References: