Abstract
In the pharmacologic prevention of vascular events, clinicians need to translate average effects from a clinical trial to the individual patient. Prediction models can contribute to individualized vascular disease prevention by selecting patients for treatment based on estimated risk or expected benefit from treatment. For patients with diabetes or vascular disease, currently no individualized approach for vascular disease prevention is applied, as these patients are generally all considered high-risk. An individualized approach may however provide several opportunities for these patients: the intensity of secondary prevention of major vascular events can be determined based on estimated prognosis and predictions can be used to inform the patient.
In thesis, it was shown that there is wide variation in 10-year absolute treatment effects of moderate-intensity statin therapy in patients with type 2 diabetes, ranging from very low (<2% 10-year absolute risk reduction, or iNNT >50) to very high (>5% 10-year absolute risk reduction, or iNNT <20). Selective prediction-based statin treatment of patients with type 2 diabetes may result in higher net benefit than a strategy in which all patients are treated, in particular if 10-year numbers willing to treat are 50 or lower. Next, it was shown that in patients with clinical vascular disease there is wide variation in the risk of recurrent vascular events with 18% of the patients at <10% 10-year risk and 22% at >30% 10-year risk. Even if all risk factors would be at guideline-recommended target, the 10-year residual risk would be estimated to be <10% for half of the patients with vascular disease, but many patients are still at high and very high residual risk of recurrent vascular events. The SMART risk score for 10-year risk of a recurrent major vascular event was used for estimating 10-year risks. The SMART score was externally validated and showed reasonable performance in external populations of patients with coronary, cerebrovascular and/or peripheral artery disease, apart from overestimation of risk in very high risk patients (10-year risk >40%). Next, we developed and externally validated the internationally applicable REACH-SMART model that can be used to make lifetime predictions for individual patients with vascular disease.
Subsequently, the estimation of individualized treatment effect was taken a step further by showing how preventive treatment effect in terms of disease-free life-years gained in individual patients can be predicted based on randomized clinical trial data. It was shown that highest treatment effect is generally achieved in the younger patients with otherwise high cardiovascular risk factors, but not necessarily high estimated 10-year risk. Such an approach was applied to estimate the expected lifetime benefit from PCSK9-inhibition in high-dose statin-treated patients with coronary artery disease, which was shown to vary from <6 months to ≥18 months free of stroke or MI. Highest benefit was expected if treatment is initiated in younger patients (age 40-60) with relatively high levels of LDL-c and other risk factors.
The findings presented in this thesis support an individualized approach for vascular disease prevention in patients with diabetes or clinical vascular disease, that are currently considered high-risk. Also, this thesis provides a basis for future research on the translation of group-level evidence to the individual patient in the consulting room.
Original language | English |
---|---|
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 1 Dec 2016 |
Publisher | |
Print ISBNs | 978-94-6295-542-4 |
Publication status | Published - 1 Dec 2016 |
Keywords
- cardiovascular disease
- secondary prevention
- individualized
- risk stratification
- distribution
- lifetime prediction