Clinical scores fail to sufficiently identify children with familial hypercholesterolaemia
Raphael Schmieder, Johannes Krefting, Sara Ates, Lea D. Schlieben, Stefan Arens, Olga Kordonouri, Michaela Sander, Stefan Holdenrieder, Volker Mall, Thomas Meitinger, Moritz von Scheidt, Wolfgang Köenig, Georg Leipold, Holger Prokisch, Heribert Schunkert, Veronika Sanin
Abstract
AIMS: The study aimed to assess the effectiveness of three clinical diagnostic criteria [Simon Broome (SB), MEDPED (MP), and guideline-derived (GL-EAS)] in identifying children with familial hypercholesterolaemia (FH) compared with genetic testing. The evaluation involved 1337 children with elevated LDL cholesterol (LDL-C) levels, focusing on the sensitivity and specificity of these clinical scores in detecting genetically confirmed FH cases. METHODS AND RESULTS: Clinical data were gathered by a self-reporting questionnaire. Clinical FH was defined in accordance with the tested FH score. Genetically confirmed heterozygous FH (HeFH) was defined by a (likely) pathogenic variant. Of the 1337 children undergoing genetic analysis, 211 showed a pathogenic FH mutation. Applying SB, MP, and GL-EAS criteria resulted in 210/1337, 125/1337, and 112/835 children being categorized to have FH clinically. The sensitivity of the clinical scores ranged from 0.44 to 0.54 with a positive predictive value (PPV) of 0.51-0.79. The specificity was 0.91-0.97 with a negative predictive value (NPV) of 0.89-0.91. Similar results were observed for the three clinical scores regarding sensitivity, specificity, PPV, and NPV in subgroup analyses defined by gender, age (<10 years vs. ≥10 years), or weight [≥90th BMI (body mass index) percentile vs. <90th BMI percentile]. CONCLUSION: Clinical FH scores offer a high degree of specificity for FH diagnosis in children, but at the expense of low sensitivity. Specifically, half of the mutation-positive children in this study would have been missed for early diagnosis and preventive treatment. Given the widespread availability of affordable genetic testing, such analysis should be performed at a lower threshold than that indicated by these clinical scores.