Systematic Reclassification of Variants of Uncertain Significance in a Clinical Nephrogenomics Program: Insights from the RAPIDS Collaboration
Tung Lin Lee, Yaochun Zhang, Tina Si Ting Lim, Chee Teck Koh, Kay Yuan Chong, Christina Titin, Hui Zhuan Tan, Irene Mok, Jia Liang Kwek, Cynthia Ciwei Lim, Jason Chon Jun Choo, Jeannette Goh, Sylvia Kam, Breana Cham, Jiin Ying Lim, Ee Shien Tan, Esther Huimin Leow, Yong Hong Ng, Jia Ying Celeste Yap, Sing Ming Chao, Siew Le Chong, Ru Sin Lim, Su Mein Goh, Regina Shaoying Lim, See Cheng Yeo, Ziyin Wang, Chia Wei Lim, Dana Leng Hui Chan, David Lu, Sharon Teo, Perry Y W Lau, Mya Than, Jun Li Ng, Gek Cher Chan, Yi Da, Boon Wee Teo, Hui Kim Yap, Hui-Lin Chin, Kar Hui NgAbstract
Background
A nationwide nephrologist-led genetic testing program for suspected monogenic glomerular disease and unexplained chronic kidney disease was implemented across four tertiary hospitals in Singapore. Initial testing yielded a genetic diagnosis in 11.5% probands, while 77 (18.4%) harboured variants of uncertain significance (VUS) that could explain their kidney disease (VUS-congruent). We evaluated a structured pipeline for reclassifying VUS-congruent to likely pathogenic or pathogenic (LP/P) to improve diagnostic yield and clinical utility.
Methods
Probands with VUS-congruent were reviewed at monthly multidisciplinary genomic board meetings. A stepwise framework incorporated genotype–phenotype correlations (Step 1), then re-appraisal of existing evidence (Step 2) through review of population databases, literature and in-silico prediction. For VUS with high suspicion of pathogenicity (VUS-High), resolution strategies were recommended (Step 3), including familial variant-specific testing, deeper phenotyping, and functional studies. Primary outcome was the proportion of VUS-congruent reclassified to LP/P; secondary outcomes were improvements in diagnostic yield and clinical utility.
Results
Among 418 probands tested over 28 months, 77 (18.4%) had VUS-congruent. Step 2 reclassified 26/77 VUS-congruent to LP/P, with another 22/77 VUS-congruent stratified as VUS-High. Among these 22 recommended for Step 3, completed strategies were functional studies (n=8), segregation analysis (n=1), deeper phenotyping (n=1), phase determination (n=1) and combination strategies (n=6). This enabled VUS-High reclassification to LP/P for 16 additional probands. Overall, 42/77 probands (54.5%) had VUS-congruent reclassified to LP/P, increasing diagnostic yield from 11.5% to 21.5%. Among these, 28 (66.7%) experienced clinically meaningful management changes.
Conclusion
A structured, multidisciplinary VUS resolution workflow enabled considerable VUS reclassification, nearly doubling the diagnostic yield.