Physicians' Academy for Cardiovascular Education

A time-dependent model to estimate clinical benefit with lipid-lowering therapies

Time-Dependent Cardiovascular Treatment Benefit Model for Lipid-Lowering Therapies

Literature - Khan I, Peterson ED, Cannon CP et al. - .J Am Heart Assoc. 2020;9:e016506. DOI: 10.1161/JAHA.120.016506

Introduction and methods

Large randomized clinical trials (RCTs) typically have a limited duration of several years, and therefore evidence demonstrating long term benefit of LDL-c lowering is limited. Furthermore, not all populations are included in RCTs, such as diabetes patients in primary setting. For these purposes, it can be useful to construct a model derived from past trial evidence to evaluate different scenarios not tested in RCTs.

A model was developed from data of past RCTs of lipid-lowering trials (LLTs, including statin and nonstatins trials) for a generalizable assessment of impact of LDL-c lowering on CV risk reduction. Time-dependent clinical benefit in RCTs with LLTs was the focus of the study. The model was validated by comparing predictions of the model with trial-reported outcomes. Implications of LDL-c lowering over extended (5-15 years) time in both primary and secondary prevention settings were examined. In addition, baseline risk estimates were calculated using this model.

Published data from RCTs of LLTs (statins, ezetimibe, PCSK9 inhibitors and anacetrapib) with ≥1000 individuals and end points defined as MACE or mortality [1-4] were used to construct the model. Exclusion criteria were open-label design, reported data not suitable for model estimation, trials involving special populations and trials with bococizumab. Also, data from the Mendelian randomization meta-analysis from Ference et al. [5] were used as a single study and informed the model over a time span of 40 years.

Main results


This time-dependent model accurately predicted clinical benefit by incorporating patient profile, timing, duration and type of treatment. This model can help in decision making by facilitating a patient-specific assessment of benefit. Also, this model can be applied for analyses of benefit with LTTs in various patient populations not included in RCTs and various time frames.


Show references

Find this article online at J Am Heart Assoc

Share this page with your colleagues and friends: