Blood pressure variability: a new therapeutic target on the horizon

A growing amount of evidence supports the view that variability of cardiovascular risk factors is associated with cardiovascular fatal and nonfatal events independently from absolute value [1]. This association has been consistently reported for blood pressure (BP), BMI and metabolic paramaters, such as glycemia, glycosylated hemoglobin and low-density lipoprotein (LDL) cholesterol. In particular, variability and instability of BP, independently of average BP, has emerged as an important predictor of cardiovascular morbidity and mortality. In healthy individuals, BP variability (BPV) is regarded as a dynamic phenomenon including short-term, mid-term, and long-term fluctuations resulting from an interaction of physical and psychic stimuli (i.e. physical activity, mental stress, duration/quality of sleep, smoking, alcohol consumption, seasonal variations in temperature), modulated by variable individual reflex responses. The dynamics of these fluctuations are obviously more complex in the hypertensive setting in relation to the severity of hypertension, use of antihypertensive medications, comorbitidies, compliance to treatment and the presence of hypertension mediated organ damage (HMOD) [2]. In spite of the fact that for over half a century arterial BP has been known for its high variability over time this key aspect has long been neglected in assessing cardiovascular risk in hypertensive patients, as well as in evaluating the efficacy of antihypertensive treatment. In a pioneering study published in the late 1980s, Mancia's group provided the first evidence of the impact of BPV on several markers of HMOD [3]. Seventy-three patients with primary systemic hypertension, who underwent 24-h ambulatory BP monitored intra-arterially by the Oxford technique during hospitalization, were re-examined after a median follow-up period of 7 years. In this selected cohort, a significant correlation was found between overall HMOD score at follow-up evaluation (a composite of electrocardiographic and/or echocardiographic left ventricular hypertrophy, retinopathy and renal dysfunction) and BPV assessed at baseline evaluation [i.e. among half-hour standard deviation (SD) of 24-h mean BP] after adjusting for major confounders. Indeed, for any level of 24-h mean BP, patients with low 24-h BPV had a lower incidence and severity of HMOD. Despite the relevance of these observations, a long period of time elapsed before clinical research provided further evidence on the prognostic and therapeutic implications of BPV. A fundamental step in this direction occurred 20 years later as a result of a series papers by Rothwell et al.[4–6] released in 2010 that opened an intense phase of research in this field and in particular on the long-term BPV, assessed as visit-to-visit variability, with the publications of hundreds of articles in different clinical settings and their meta-analyses [7–16]. In a post hoc analysis of the Anglo-Scandinavian Cardiac Outcomes Trial Blood Pressure Lowering Arm (ASCOT-BPLA) and Medical Research Council (MRC), Rothwell et al.[5] investigated whether visit-to-visit BPV might explain disparities on stroke prevention because of different BP-lowering drugs (i.e. calcium channel blocker vs. beta-blocker). In the ASCOT-BPLA study, 19 257 patients with high BP (mean 164/95 mmHg at baseline) were randomly assigned to treatment with the calcium channel blocker amlodipine with or without the ACE inhibitor perindopril, or to treatment with the beta-blocker atenolol with or without the diuretic bendroflumethiazide. During 5.5 years of follow-up, BPV was assessed by comparing multiple BP readings taken at each visit and at several different visits. Lower within-individual visit-to-visit BPV in the amlodipine/perindopril arm of the ASCOT study accounted for the more effective stroke reduction rate compared with atenolol and diuretic-based therapies, despite overall similar BP reduction. Indeed, the lower risk of stroke in the amlodipine group [hazard ratio 0.78, 95% confidence interval (CI) 0.67–0.90] was abolished by adjusting for within-individual SD of clinic SBP (hazard ratio 0.99, 95% CI 0.85–1.16). In the MRC trial, all measures of within-individual visit-to-visit variability in SBP were increased in the atenolol group compared with the diuretic group during initial follow-up and subsequent temporal trends in BPV during follow-up in the atenolol group correlated with trends in stroke risk. Therefore, the hypothesis that erratic BP levels reflected by visit-to-visit variability could be an important factor in determining the risk of HMOD and cardiovascular events was addressed in a variety of clinical settings, such as population-based cohorts [9], elderly individuals [12], patients with hypertension [10], diabetes [17], chronic kidney disease (CKD) [11] and ischemic stroke [18]. In an early systematic review and meta-analysis, based on 37 studies published by mid-2014, an association between visit-to-visit BPV with cardiovascular disease and mortality outcomes was present in some but not all studies [13]. Pooled data obtained in a subsample of the above-mentioned studies showed statistically significant associations, albeit modest in magnitude, between visit-to-visit SBPV and outcomes including stroke, coronary heart disease, cardiovascular mortality, and all-cause mortality. However, inconsistent results exist in this association. A subsequent meta-analysis of 13 prospective studies totaling 77 299 patients followed for an average period of 6.3 years suggested that visit-to-visit SBPV was a strong predictor (independent of age and mean SBP) of all-cause and cardiovascular mortality and stroke [15]. In detail, a 1 mmHg increase in SD was significantly associated with a 3% higher risk of all-cause death, 10% higher risk of cardiovascular mortality and 2% higher risk of stroke. The unfavorable impact of visit-to-visit BPV on incident stroke was strongly highlighted by the meta-analysis of Wang et al.[19] based on a very large pooled population of 307 804 patients from 14 studies. After adjustments for prior stroke history and cardiovascular risk factors, high visit-to-visit SBPV predicted an increased risk of stroke (relative risk = 1.20, 95% CI 1.07–1.35, P = 0.0005), without difference between women and men. More recently, Li et al. performed a meta-analysis of 14 studies including more than 11 million participants (belonging to the general population, patients with hypertension, with type 1 and 2 diabetes) aimed at assessing the relationship between visit-to-visit BPV and the risk of CKD [19]. They found that the risk of CKD [i.e. new onset albuminuria, proteinuria, decline in estimated glomerular filtration rate (e-GFR), and end-stage renal disease] was significantly greater in patients with high baseline SBPV than in patients with low-baseline SBPV (relative risk = 1.69, 95% CI 1.38–2.08). At the same time, it is noteworthy that so far, no meta-analyses have been published on the association of BPV with other relevant markers of HMOD (i.e. left ventricular hypertrophy, arterial stiffness and carotid atherosclerosis). In this scenario, the findings provided by the study of Mancusi et al.[20], carried out in a large representative sample of treated hypertensive patients enrolled in the Campania Salute Network (a real-world observational registry resulting from the interaction between hypertension specialist clinics and general practitioners), provides interesting insights regarding the complex relationship between visit-to-visit BPV, cardiovascular events and HMOD. A total of 3555 adult hypertensive patients (mean age 52 years) without prevalent coronary or valvular heart disease, cerebrovascular disease, advanced CKD (i.e. e-GFR >30 ml/min) and preserved ejection fraction were included in the study. All patients had more than six visits during follow-up, over a minimum of 1 year (average 8 years). Visit-to-visit BPV was calculated as SD of SBP and DBP, across the follow-up visits, excluding values from the initial visit. The primary cardiovascular end-point of the study was a composite constituting fatal and nonfatal myocardial infarction, stroke, sudden cardiac death, heart failure requiring hospitalization, transitory ischemic attack, myocardial or carotid revascularization and atrial fibrillation. Baseline left ventricular (LV) mass index, e-GFR and their changes during the follow-up were taken as markers of HMOD. Patients were categorized in tertiles of SD of SBP during follow-up and high visit-to-visit BPV was identified as the highest tertiles (>14.8 mmHg). As for the clinical characteristics at initial observation, a progressive increase in age, prevalence of women, obesity, diabetes, office SBP/DBP and pulse pressure as well as in severity of HMOD occurred from the first to the third tertiles of SD of SBP. Both increase in LV mass index and decline in eGFR at follow-up were associated with high visit-to-visit SBPV independently of several important confounders, including average SBP and baseline measures of such markers of HMOD. It is interesting to note, however, that the progression of cardiac and renal organ damage was evident in the whole study population starting from the low tertile of SBPV and that the differences of mean baseline-follow-up changes in both LVMI and eGFR between tertiles were quite limited. Overall, these data suggest that high visit-to-visit BPV may be a marker of high risk but not necessarily causally related to HMOD. Furthermore, it is possible that subclinical cardiac and vascular damage may be implicated in the genesis of abnormal BPV. Finally, regarding the primary outcome of the study, a total of 264 cardiovascular events occurred during the follow-up. Patients with high visit-to-visit SBPV exhibited a 51% increase in risk compared with their counterpart with low BPV. Also in this analysis, the excess risk remained significant even after adjustment for conventional risk factors, organ damage and classes of antihypertensive drugs. Classifying patients in relation to achieved BP control (average office BP values during follow-up visits <140/90 mmHg), high visit-to-visit BPV turned out to be an independent predictor of cardiovascular in the subgroup of those who had reached the therapeutic target. This was not the case for the uncontrolled patients. These results contrast with those recently suggested by a post hoc analysis of the Systolic Blood Pressure Intervention Trial (SPRINT) [21]. In that study BPV was not significantly associated with the primary composite endpoint of fatal and nonfatal cardiovascular events, or with heart failure or stroke hospitalizations. It should be mentioned, however, that in the SPRINT study, visit-to-visit BPV was markedly lower than in the cohort by Mancusi et al. (7.8 vs. 13.0 mmHg). In conclusion, despite the limitations recognized by the authors (i.e. differences in office BP measurement methods, time intervals between visits not defined by protocol, lack of information on therapeutic compliance) and the heterogeneity of the components of the primary outcome, the study of Mancusi et al. has the great merit of stimulating the debate on the clinical and prognostic role of long-term BPV and on the need to target the management of patients with high BPV in terms of a more comprehensive evaluation of the BP profile through out-of-office measurements, assessment of HMOD and specific therapeutic approach. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.