Author(s):
Stevens, C.A.T. ; Elshorbagy, A. ; Vallejo-Vaz, A.J. ; Dharmayat, K. ; Lyons, A. ; Bourbon, M. ; Chora, J. ; Humphries, S.E. ; Catapano, A.L. ; Hovingh, G. ; Mata, P. ; Santos, R. ; Soran, H. ; Watts, G.F. ; Raal, F. ; Freiberger, T. ; Ray, K.K. ; on behalf of all the EAS FHSC Collaborators
Date: 2024
Persistent ID: http://hdl.handle.net/10400.18/9197
Origin: Repositório Científico do Instituto Nacional de Saúde
Subject(s): Familial Hypercholesterolaemia; Lipid-lowering Medication; Correction Factors; Doenças Cardio e Cérebro-vasculares
Description
Background and Aims: Pretreatment LDL-C measurements aid familial hypercholesterolaemia (FH) diagnosis, and are crucial in epidemiologic studies investigating FH, but are often unavailable because individuals are already on lipid-lowering medication (LLM). Several formulae have been reported to estimate pre-treatment LDL-C in people on LLM by ‘correcting’ their LDL-C concentrations for LLM type and dosage, based on observational or trial evidence of drug efficacy. We compared 4 published correction factors in estimating pre-treatment LDL-C in patients with FH. Methods: Cross-sectional analysis of adults with pathogenic/likely-pathogenic FH variants in the EAS-FH Studies Collaboration (FHSC) Registry. At the time of LDL-C measurement, N=3012 participants were not on LLM (Untreated group), and N=3226 were on LLM monotherapy, with information on LLM type and dosage allowing estimation of pre-treatment LDL-C (Corrected group) based on correction factors by Ruel 2018, Ellis 2016, Haralambos 2015 and Besseling 2014. We compared the groups for clinical characteristics and LDL-C by gene and variant. Results: The Corrected group was older than the Untreated group (median[IQR]: 50[39,63] vs. 38[28,50]y), with similar proportion of women (54.5% vs. 56.8%;p=0.14) but more comorbidities (all p<0.001). In the Corrected group, 3120 were on statins, 106 on ezetimibe, none on PCSK9-inhibitors. The Corrected group had higher LDL-C vs. Untreated group, with the difference greater at upper percentiles, regardless of correction factor. LDL-C was highest in those with LDLR>APOB>PCSK9 gene variants, but Corrected was still higher than Untreated LDL-C within each gene group. The difference in Corrected vs. Untreated LDL-C varied by variant, from +0.6 to +3.5mmol/L (20 commonest variants). The LDL-C differences persisted after adjusting for age, sex and comorbidities. Conclusions: Application of current LDL-C correction factors appears to overestimate pre-treatment LDL-C in epidemiologic settings, or the Untreated and Corrected groups might have inherently different LDL-C profiles. The accuracy of using LDL-C correction factors in FH therefore warrants further investigation.
