Failure to identify FH results in worse survival compared with people with an FH diagnosis
Why diagnosis of familial hypercholesterolemia matters: Life expectancy is 16 years lower in undiagnosed people
Presented at EAS 2019 in Maastricht, The Netherlands, by Kausik K. Ray (Imperial College London, UK)
Introduction and methods
About one in 200-250 people in Europe are thought to be affected by familial hypercholesterolemia (FH), which puts individuals at very high risk of premature CV events, due to exposure to high levels of LDL-c from birth. If these people are identified early and treated to lower their LDL-C levels, they gain in terms of life quality and quantity. The problem is that lack of awareness among clinicians and the general population hampers diagnosis, as underlined by a worldwide survey of 63 centres in the European Atherosclerosis Society FH Studies Collaboration registry. It is unknown to date whether outcomes differ between individuals with a confirmed diagnosis of FH and individuals with potential FH who are as yet undiagnosed. There is no consistency globally as to whether FH should be screened for.
This collaborative study including the Imperial Centre for Cardiovascular Disease Prevention, Imperial College London, the University of Leicester, Royal Bournemouth Hospital and Amgen Center for Observational Research, retrospectively analyzed the patient records of 1,729,046 individuals in the UK Clinical Practice Research Datalink database. Individuals were initially categorised according to whether or not they had a recorded diagnosis of FH (coded FH: n=6843). For those without a coded diagnosis, patients with available LDL-C levels and other characteristics who could be considered definite or probable FH were classified as potential FH (n=13.459), and the rest were classified as unlikely FH (n=1.707.744), using the Dutch Lipid Clinic Network Criteria. The risk of premature cardiovascular events and death was compared across the three groups.
- The people with unlikely FH were older (mean age 62.2 (SD: 15.8) than those with coded FH (mean: 50.9, SD: 13.7) or potential FH (mean: 60.8, SD: 11.9). People with coded FH had LDL-c measured at a younger age than those with potential FH.
- Mean LDL-c was 4.8 (SD: 1.5), 5.6 (SD: 1.3) and 3.0 (SD: 1.0) in the coded, potential and unlikely FH groups, respectively. Statin prescription was 33.3%, 67% and 23.9%, respectively, and high-statin was taken by 11.3%, 20.5% and 2.8% in these groups.
- The groups with coded (HR: 1.97, 95%CI: 1.82-2.12) and potential FH (HR: 1.89, 95%CI: 1.79-1.99) had a higher risk of premature CV events than those with unlikely FH (both P<0.001).
- People with potential FH but who were not diagnosed yet, had a higher risk of early death (HR: 6.69, 95%CI: 4.88-9.16), compared with those already diagnosed.
- The absence of a diagnosis was associated with worse survival (potential vs coded FH). At age 18, this translated to a loss of 16 years of life compared with diagnosed individuals.
- Among individuals with elevated LDL-c, more than half (68%) of all premature deaths were in individuals without a coded FH diagnosis (56% when excluding smokers and those with hypertension and/or diabetes).
This analysis suggests that a considerable number of individuals in the UK have FH, without having been diagnosed. Failure to identify FH results in a loss of 16 years of life compared with people with an FH diagnosis, suggests a UK study involving over 1.7 million health records in the general population.
Professor Kausik Ray commented: ‘Using a ‘big data’ approach, the results from this study of over 1.7 million people strengthen the case for early detection of FH. Not having an FH diagnosis despite having high LDL cholesterol levels suggestive of FH, increases the likelihood of heart attack and early death. These findings underline the need for early screening for FH, ideally in children, to identify and treat much earlier so that these people can lead longer and healthier lives.’
Our reporting is based on the information provided at the EAS 2019 congress