Comparative Alignment of CFTR2 and CFTR-France Classifications with a Turkish Cystic Fibrosis Referral Cohort
TL;DR
Recurrent non-CF-causing variants should be reviewed case by case; the present data alone do not justify pathogenic reclassification, and phase-resolved and functional studies are required.
Abstract
CFTR variant spectra vary across populations, potentially limiting the transferability of reference classifications. We compared CFTR2 and CFTR-France with a single-centre Turkish cystic fibrosis referral cohort of 418 individuals tested between 2016 and 2022. To avoid incorporation bias, the primary analysis used measured sweat-test status as a genetics-independent anchor; unperformed tests were excluded rather than counted as negative. A CFTR variant was detected in 162/418 individuals (38.8%). Sweat-test results were available for 148 individuals: 85 positive and 63 negative. Variant detection was more frequent among sweat-positive individuals (50/85 vs. 23/63; OR 2.48, 95% CI 1.27–4.86; Fisher’s exact p = 0.008). Excluding exact ties, measured-sweat alignment was 22/28 (78.6%, 95% CI 59.0–91.7%) for CFTR2 and 15/21 (71.4%, 95% CI 47.8–88.7%) for CFTR-France. Direct database agreement was 362/395 (91.6%; Cohen’s κ = 0.74) across three classes and 330/330 (100%) for directional binary labels. CFTR2 and CFTR-France represented 64/75 (85.3%) and 58/75 (77.3%) cohort variants, respectively. Measured-sweat alignment was high for CF-causing but variable for non-CF-causing variants. Therefore, recurrent non-CF-causing variants should be reviewed case by case; the present data alone do not justify pathogenic reclassification, and phase-resolved and functional studies are required.