Comparison between different diagnostic scores for the diagnosis of familial hypercholesterolemia: assessment of their diagnostic accuracy in comparison with genetic testing

Abstract

Abstract Background The Dutch Lipid Clinic Network Score (DLCN) is a useful score to guide the physician to the diagnosis of familial hypercholesterolemia (FH). However, some concerns have been pointed out about its diagnostic accuracy in pediatric population; so, other scores such as Simon Broome (SB) and US MedPed Program have been tested and currently used. Aim To test the different diagnostic and predictive accuracy between those three scores (DLCN, SM and MedPed) in relation to positivity of genetic test. Methods We enrolled 25 pediatric patients aged <18 yo referred to our Chieti and Rome lipid clinics for the clinical suspicion of FH. We performed the genetic analysis for searching pathogenetic mutations for FH (NGS by Lipid in Code, LIPIGEN study). Differently from the others, DLCN provides a numerical data as a diagnostic score; thus, to uniform the three scores, we chose to define subjects with a DLCN score ≥6 as likely FH, while subjects with a lower value were defined as unlikely FH. Furthermore, we calculated the SB and the MedPed score, as previously demonstrated, defining the subjects as likely or unlikely FH. Therefore, we analyzed the diagnostic accuracy of those scores in relation to FH mutation data (positive or negative). Results We recruited 25 patients, 13 males (52%), mean age of 11.8 years. Sixteen patients (64%) resulted positive for the genetic mutation The DLCN showed a sensitivity of 43%, specificity of 66%, a positive predictive value (PPV) of 0.7 and a negative predictive value (NPV) of 0.4. SB showed a sensitivity of 68%, specificity of 55%, a PPV of 0.73 and NPV of 0.5. The MedPed score showed a sensitivity of 68%, specificity of 44%, a PPV of 0.68 and NPV of 0.44. Furthermore, we elaborated a ROC curve of low-density lipoprotein cholesterol (LDL-C) values with respect to the positivity of the FH genetic test: the value that showed the best diagnostic accuracy was 190 mg / dL (sensitivity 68.8% and specificity 75%), while the value of 116.5 mg/dL demonstrated a sensitivity of 100% but with a specificity of 11%. Conclusions Our work demonstrates that none of the three scores tested has an optimal diagnostic accuracy with respect to the diagnosis of FH in the pediatric population. By comparison, SB demonstrates the best sensitivity and the best specificity. In addition, the LDL-C value of 190 mg / dL provides the best diagnostic accuracy against the diagnosis of FH in the pediatric population. Further data and studies will be needed to improve the diagnostic performance of these scores, in order to refer pediatric patients to the genetic test, for an early confirmation in order to refer pediatric patients to the genetic test, for an early FH diagnosis confirmation. Funding Acknowledgement Type of funding source: None