Pharmacological blood-pressure lowering for the prevention of cardiovascular disease and death across the full spectrum of chronic kidney disease severity: an individual-participant data meta-analysis.

Abstract

BACKGROUND Individuals with chronic kidney disease (CKD), particularly those at more advanced stages, have been systematically under-represented in randomised controlled trials (RCTs) of blood-pressure-lowering treatment due to safety concerns, leading to a persistent paucity of evidence for cardiovascular risk management in this high-risk group. We investigated the effect of blood-pressure-lowering treatment on the risk of major cardiovascular disease and death across the full spectrum of CKD stages and by key clinical subgroups. METHODS We conducted a one-stage meta-analysis of individual-participant data from RCTs in which participants were randomly assigned to a blood-pressure-lowering therapy versus a comparator. We used RCTs collated in the Blood Pressure Lowering Treatment Trialists' Collaboration dataset, published at any time in any language, which were eligible for inclusion if they had at least 1000 person-years of follow-up per arm, baseline blood-pressure and creatinine measurements, and time-to-event outcomes; those with unclear randomisation procedures or restricted to heart failure or acute care settings were excluded. Participants with a documented history of heart failure or extreme creatinine values were excluded. No age criteria were applied. The primary outcome was major cardiovascular events, defined as a composite of fatal or non-fatal stroke, ischaemic heart disease, or hospitalisation for, or death from, heart failure. Relative treatment effects were estimated with a stratified Cox proportional hazards model. Heterogeneity of treatment effects was evaluated across prespecified subgroups defined by CKD status, CKD stage (1-5), diabetes, proteinuria, and baseline blood pressure. A stratified network meta-analysis was performed to examine whether treatment effects differed by defined subgroups within each of five principal antihypertensive drug classes. The systematic review was registered in PROSPERO (CRD42018099283). FINDINGS From 52 RCTs (363 684 participants), a total of 285 124 participants from 46 randomised trials met the eligibility criteria; 116 145 (40·7%) were women, 168 979 (59·3%) were men, 59 185 (20·7%) had CKD at baseline, and 86 067 (30·2%) had type 2 diabetes. During a median follow-up of 4·4 years (IQR 3·2-5·1), a 5 mm Hg reduction in systolic blood pressure reduced the risk of major cardiovascular disease in individuals with CKD (hazard ratio [HR] 0·91 [95% CI 0·87-0·94]) and without CKD (0·90 [0·88-0·93]; pinteraction>0·99). Furthermore, these observed relative risk reductions were consistent across all CKD stages, including severe stages 4-5 (pinteraction>0·99). Similar treatment effects were observed by proteinuria status and across blood-pressure categories, down to <120/70 mm Hg. However, the relative treatment effect in individuals with CKD was notably attenuated among those with coexisting diabetes (HR 0·96 [95% CI 0·90-1·02]) compared with those without (0·88 [0·84-0·93]; pinteraction=0·044). The stratified analysis within each drug class showed that the class-specific effects of antihypertensive agents versus placebo on cardiovascular disease risk remained unchanged across the investigated subgroups. INTERPRETATION In the context of cardiovascular risk reduction, the relative benefit of blood-pressure lowering in patients with CKD is similar to that in individuals without CKD, with consistent efficacy across all CKD stages, blood-pressure thresholds, and proteinuria status. However, notably, this relative benefit is attenuated in patients with CKD and concomitant diabetes, underscoring the requirement for adapted therapeutic strategies in this high-risk subgroup. Moreover, the class-specific effects of principal antihypertensives in CKD mirror those observed in the broader population, independent of CKD stage or proteinuria status. FUNDING British Heart Foundation.