Endocrine Abstracts

Introduction: For the clinical diagnosis of heterozygous familial hypercholesterolemia (HeFH) validated algorithms are used, with the Dutch Lipid Clinic Network (DLCN) Criteria being the most recommended in our environment. One of the items that scores for the system proposed by the DLCN is the value of LDL cholesterol. However, LDLc levels can be distorted by lipoprotein a (Lpa(a)) levels.

Objective: To analyze the change in the validity of the DLCN criteria for the diagnosis of HeFH with pathogenic mutation confirmed when using the LDLc.

Methods: Observational, retrospective, cross-sectional, and analytical study in a cohort of 91 patients with clinical diagnostic criteria for probable or defined familial hypercholesterolemia according to the DLCN criteria, followed in our Lipid Unit at Lozano Blesa University Clinical Hospital, from 1 May 2019 to 31 December 2020. The DLCN criteria have been calculated based on LDLc values, and LDLc corrected for Lp(a) [according to the corrected LDLc formula = LDLc – 0.30*Lp(a)]. In all the statistical test performed, significant values of P<0.05 were considered. Several ROC (Receiver Operating Characteristic) curves have been generated between the initial LDLc and that LDLc corrected by Lp(a) values to predict the presence of genetic mutation. Corresponding cross tables have been made between LDLc (initial and corrected) and genetic mutation; as well as between the DLCN score (obtained with uncorrected LDLc) and genetic mutation.

Results: We included 91 patients (51.65% women, mean age at diagnosis 48.80±12.96 years). 43.95% of them did not present mutation in the genetic study for HeFH. Lp(a) in subjects with no known mutation was 24.70 ± 86.10 mg/dl compared to 22.50 ± 60.3 mg/dl in patients with known mutation. The area under the ROC curve of the LDLc variable and the presence of mutation in the genetic test was higher in the case of LDLc corrected by Lp(a) (AUC=0.639,P=0.038) than in the uncorrected LDLc (AUC=0.623,P=0.065). This improves the specificity of the LDLc variable to predict genetic mutation if used corrected by Lp(a): 70% versus 62.74% if used uncorrected. When using the LDLc corrected by Lp(a) for the calculation of the DLCN score, 66.66% of the patients were reclassified from probable diagnostic group to a possible diagnosis, without presenting 56.26% of them a known genetic mutation.

Conclusions: The use of adjustment of LDL cholesterol levels according to Lp(a) levels in the diagnosis of HeFH facilitates a more specific diagnosis than with the use of unadjusted LDLc.