Health Technology Assessment (Mar 2012)
Elucigene FH20 and LIPOchip for the diagnosis of familial hypercholesterolaemia: a systematic review and economic evaluation
Abstract
Background: Familial hypercholesterolemia (FH) is an autosomal dominant genetic condition causing a high risk of coronary heart disease. The prevalence of this disease is about 1 in 500 in the UK, affecting about 120,000 people across the whole of the UK. Current guidelines recommend DNA testing, however, these guidelines are poorly implemented, therefore 102,000 or 85% of this group remain undiagnosed. Objectives: To assess the diagnostic accuracy, effect on patient outcomes and cost-effectiveness of Elucigene FH20 and LIPOchip for the diagnosis of FH. Data sources: Electronic databases including MEDLINE, MEDLINE In-Process & Other Non-Indexed Citations, EMBASE, BIOSIS, Science Citation Index, Conference Proceedings Citation Index – Science and Cochrane Controlled Trials Register were searched until January 2011. Review methods: A systematic review of the literature on diagnostic accuracy was carried out according to standard methods. An economic model was constructed to assess the cost-effectiveness of alternative diagnostic strategies for the confirmation of clinically diagnosed FH in index cases and for the identification and subsequent testing of first-, second- and possibly third-degree biological relatives of the index case. Twelve strategies were evaluated linking diagnostic accuracy to treatment outcomes and hence quality-adjusted life-years (QALYs). Deterministic and probabilistic sensitivity analyses were undertaken to investigate model and parameter uncertainty. Results: Fifteen studies were included for diagnostic accuracy; three reported Elucigene FH20, five reported LIPOchip, four reported low-density lipoprotein cholesterol (LDL-C) tests and three reported an age- and gender-specific LDL-C test against a reference standard of comprehensive genetic analysis (CGA). Sensitivity ranged from 44% to 52% for Elucigene FH20 and from 33.3% to 94.5% for various versions of LIPOchip in detecting FH-causing mutations in patients with a clinical diagnosis of FH. For LIPOchip version 10 (designed to detect 189 UK specific mutations), sensitivity would be 78.5% (based on single-centre data – Progenika, personal communication). For all other Elucigene FH20 or LIPOchip studies (apart from one LIPOchip study), specificity could not be calculated as no false-positive results could be derived from the given data. The LDL-C test was generally reported to be highly sensitive but with low specificity. For age- and gender-specific LDL-C cut-offs for cascade testing, sensitivity ranged from 68% to 96%. One UK-based study reported sensitivity of 91% and specificity of 93%. For the cost-effectiveness review, only one study reporting cost-effectiveness of any one of the comparators for this assessment was identified. Pre-screen strategies such as Elucigene FH20 followed by CGA were not cost-effective and were dominated by the single more comprehensive tests (e.g. CGA). Of the non-dominated strategies, Elucigene FH20, LIPOchip platform (Spain) and CGA were all cost-effective with associated incremental cost-effectiveness ratios (ICERs) relative to LDL-C of dominance (test is less costly and more effective), £871 and £1030 per QALY gained respectively. CGA generates the greatest QALY gain and, although other tests have lower ICERs relative to LDL-C, this is at the expense of QALY loss compared with the CGA test. Probabilistic sensitivity analysis shows that CGA is associated with an almost 100% probability of cost-effectiveness at the conventional value of willingness to pay of £20,000 per QALY gain. Limitations: There was much uncertainty regarding the diagnostic accuracy of the included tests, with wide variation in sensitivity across reported studies. A lack of published information for the most recent version of LIPOchip created additional uncertainty, especially in relation to the chip’s ability to detect copy number changes. For the economic modelling, we aimed to choose the best studies for the base-case sensitivity of the tests; however, a number of informed choices based on clinical expert opinion had to be made in the absence of published studies for a number of other parameters in the modelling. This adds some uncertainty to our results, although it is unlikely that these would be sufficient in magnitude to alter our main results and conclusions. Conclusions: As targeted tests designed to detect a limited number of genetic mutations, Elucigene FH20 and LIPOchip cannot detect all cases of FH, in contrast with CGA. CGA is therefore the most effective test in terms of sensitivity and QALY gain, and is also highly cost-effective with an associated ICER of £1030 per QALY gain relative to current practice (LDL-C). Other tests such as Elucigene FH20 and LIPOchip are also cost-effective; however, because of inferior sensitivity compared with CGA, these tests offer cost savings but at the expense of large QALY losses compared with CGA. Further prospective multicentred studies are required to evaluate the diagnostic accuracy of new and emerging tests for FH with the LDL-C test in patients with a clinical diagnosis based on the Simon Broome criteria. Such studies should verify both test-positive and -negative results against a reference standard of CGA and should include a full economic evaluation. Funding: The National Institute for Health Research Health Technology Assessment programme.
Keywords
