Pulse pressure as a predictive marker for cardiovascular events : relation to biomarkers and antihypertensive treatment

Blood pressure (BP), in particularly ambulatory blood pressure (ABP), is a strong predictor for cardiovascular (CV) disease (CVD). Pulse pressure (PP) is related to vascular disease and ambulatory PP (APP) may have a specific value in certain populations. It is unknown whether ABP is a better predictor for CV events compared to office BP in patients with peripheral arterial disease (PAD). NT-proBNP, hs-CRP and cystatin C are biomarkers that are increasingly used for risk prediction but prospective studies on the predictive value of these biomarkers adjusted for ABP are scarce. Although PP may have a clinical value, the relation to outcomes in interventional antihypertensive studies has not been sufficiently studied. The overall aim of this thesis was to study the predictive value of ABP with special reference to PP in relation to the biomarkers NT-proBNP, hs-CRP, and cystatin C and to evaluate whether ABP and these biomarkers improved risk prediction when added to traditional risk factor models. We further aimed to study whether the antihypertensive treatment effect on CV events was dependent on baseline PP.

Material and methods: This thesis was based on studies in patients with PAD, elderly men and high-risk hypertensives. We investigated the relations of ABP with special reference to APP and the biomarkers NT-proBNP, hs-CRP, and cystatin C to CV events during long-term follow-up. We used Cox regression models and C-statistics, net reclassification improvement and integrated discrimination improvement. We studied whether the difference in CV events between two different antihypertensive treatments was dependent on baseline PP.

Results: APP was a better predictor of CV events compared to office BP in PAD patients and a combination of APP, NT-proBNP, and hs-CRP improved discrimination and net reclassification. In elderly male subjects, the substitution of office BP with ABP in a model with traditional risk factors improved discrimination and reclassification. The addition of NTproBNP to the ABP model improved reclassification but not discrimination. However, the addition of ABP to a traditional model that included any of the biomarkers did not improve discrimination or reclassification. In high-risk hypertensive patients, we observed a positive relationship between baseline PP and incident CVD. However, the superior treatment effect of amlodipine as compared to hydrochlorothiazide when combined with benazepril was independent of baseline PP. The absolute treatment effect was higher in the higher tertiles of PP.

Conclusion: Pulse pressure is a predictor for CV events and seems to be most useful in patients with established CVD. NT-proBNP has additive value for risk prediction in patients with CVD as well as in the elderly. Combinations of pulse pressure and NT-proBNP may help to tailor treatment in subjects to prevent incident CVD. The difference in reduction of CV events between two different antihypertensive treatments was not dependent on baseline pulse pressure. That is, there is presently no evidence to support that a subject’s pulse pressure per se should direct the choice of antihypertensive drugs for treatment.

List of scientific papers

I. Ambulatory pulse pressure predicts cardiovascular events in patients with peripheral arterial disease. Per H. Skoglund, Jan Östergren, Per Svensson Blood Pressure, 2012; 21: 227–23
https://doi.org/10.3109/00365599.2012.676755

II. Amino-terminal pro-B-type natriuretic peptide and high-sensitivity C reactive protein but not cystatin C predict cardiovascular events in male patients with peripheral artery disease independently of ambulatory pulse pressure. Per H. Skoglund, Johannes Arpegård, Jan Östergren, Per Svensson. American Journal of Hypertension, 2014;27(3):363-371
https://doi.org/10.1093/ajh/hpt278

III. NT-proBNP but not cystatin C or CRP improve risk prediction for cardiovascular disease beyond ambulatory blood pressure and traditional risk factors in elderly men. Per H. Skoglund, Jonas Höijer, Johan Ärnlöv, Björn Zethelius, Per Svensson. [Manuscript]

IV. Amlodipine+benazepril is superior to hydochlorothiazide+benazepril irrespective of baseline pulse pressure : Subanalysis of the ACCOMPLISH trial. Per H. Skoglund, Per Svensson, Joline Asp, Björn Dahlöf, Sverre E. Kjeldsen, Kenneth A., Jamerson, Michael A. Weber, Yan Jia, Dion H. Zappe, Jan, Östergren, the ACCOMPLISH investigators. Journal of Clinical Hypertension, 2015 Feb;17(2):141-­6.
https://doi.org/10.1111/jch.12460.