Reducing coronary risk by raising HDL-cholesterol: risk modelling the addition of nicotinic acid to existing therapy

ABSTRACT

Background and objectives: Reduction in total cholesterol (TC) and LDL-cholesterol (LDL-C) forms one of the principal objectives of most cardiovascular secondary prevention strategies. Many patients being treated with statins, however, have significant residual dyslipidaemia, with many having suboptimal HDL-cholesterol (HDL-C) levels. The addition of nicotinic acid to a statin has been shown to improve this profile, although clinical outcome evidence is currently lacking. This study set out to model the impact of nicotinic acid therapy on cardiovascular risk in these patients, based on Framingham risk assessments on a cohort of patients drawn from UK general practitioner records.

Methods: Cardiovascular risk profiles were extracted from a research database of 602 222 patients from 98 UK general practices. 23 262 statin-treated patients with established cardiovascular disease or diabetes were identified and their 4-year Framingham risk was estimated. Patients who had either TC or HDL-C outside the desirable range then had their lipid profile adjusted in accordance with the likely performance of nicotinic acid, and the Framingham risk was then re-assessed.

Results: Baseline 4-year coronary risk in the group as a whole was 11.5% (95%CI: 11.4–11.6). After adjustment of the lipid profile, this was reduced to 9.7% (95%CI: 9.6–9.8), a reduction in risk of 15.9% (95%CI: 15.1–16.6). When modelling was limited to those with diabetes or an abnormal treated lipid profile, the magnitude of change was increased to 23–29% depending on sex and subgroup.

Conclusions: Risk factor modelling suggests that raising HDL-C levels using nicotinic acid in statin-treated patients is likely to yield significant incremental clinical benefits. The results of clinical trials currently under way are awaited with interest.

Key Words::

• Cardiovascular disease
• Cholesterol
• Coronary heart disease
• Coronary risk
• Disease modelling
• HDL-C

Acknowledgements

Declaration of interest: This work was supported by an unrestricted educational grant from Merck Sharp & Dohme Limited. The views expressed in this publication are those of the authors, and not necessarily those of the publisher or sponsor.

The PCDQ (Primary Care Data Quality) programme is based within Primary Care Informatics, Division of Community Health Sciences, St. George's – University of London and received funding from Merck Sharp & Dohme Ltd to collect the original data between October 2005 and January 2006.

J.B. is a director of JB Medical Ltd, a medical education consultancy, and an honorary research assistant at St. George's – University of London. JB Medical Ltd has been paid by Merck Sharp & Dohme Ltd to write educational programmes and clinical papers in the area of cardiology. J.B. has also received an unrestricted grant from Merck Sharp & Dohme Ltd for his participation in writing this paper. S.deL., J.v.V., T.C. and N.H. did not receive any financial support for writing this paper, and have not declared any conflict of interest.

S.deL., J.v.V., T.C. and N.H. conceived the PCDQ project, recruited practices and collected data. J.B. initiated this analysis, designed the research question, carried out the data analysis and wrote the first draft of the paper. S.deL., J.v.V., T.C. and N.H. commented on the first draft and JB managed the incorporation of comments to produce a final draft manuscript for submission.