Performance diagnostique en situation réelle du score prédictif «Allogenomics» dans une cohorte de 185 patients suspects de néphropathie héréditaire.

Nephrogenetics provides an innovative etiological approach to chronic kidney disease (CKD). However, genetic testing remains complex, costly, and generates a large number of data, requiring targeted use. In 2023, Doreille A. et coll. proposed the Allogenomics Score (SAG) predictive for a positive molecular diagnosis, based on age, CKD stage, family history, and nosological category. A SAG below -2.230916 was associated with a negative predictive value (NPV) of 94 %. We test the SAG in a monocentric study (2023-2025), including 185 patients who underwent exome sequencing filtered on a 304-gene panel. Variants were ACMG-classified, and the individual SAG was calculated. Concordance between the a priori clinical hypothesis and the a posteriori molecular diagnosis was assessed using κ Cohen's coefficient. The median SAG was -1.32 (-2.14; -0.35), corresponding to a predicted probability of positive test of 21 %, with an observed diagnostic yield of 28 %. Clinico-molecular concordance was excellent (92.3 %, κ = 0.86). The probability of a positive genetic diagnosis was significantly associated with SAG (p = 0.042), but no significant association was observed with the -2.230916 threshold (which showed NPV of 84 % in our cohort). These findings indicate excellent clinico-molecular concordance but suboptimal performance of the proposed SAG threshold in clinical routine.