#NephJC Chat
Tuesday December, 3rd, 2024 at 9 pm Eastern (AEST = December, 3rd, 1pm)
Wednesday December, 4th, 2024, at 9 pm by Indian Standard Time and 3:30 pm GMT (AEST = December, 4th, 2:30 am)
N Engl J Med 2024 Nov 16. doi: 10.1056/NEJMoa2412006. Online ahead of print.
Intensive Blood-Pressure Control in Patients with Type 2 Diabetes
Y. Bi, M. Li, Y. Liu, T. Li, J. Lu, P. Duan, F. Xu, Q. Dong, Ailiang Wang, T. Wang, R. Zheng, Y. Chen, M. Xu, X. Wang, Xinhuan Zhang, Y. Niu, Z. Kang, C. Lu, Jing Wang, X. Qiu, An Wang, S. Wu, J. Niu, Jingya Wang, Z. Zhao, H. Pan, X. Yang, X. Niu, S. Pang, Xiaoliang Zhang, Y. Dai, Q. Wan, S. Chen, Q. Zheng, S. Dai, J. Deng, L. Liu, G. Wang, H. Zhu, W. Tang, H. Liu, Z. Guo, G. Ning, J. He, Y. Xu, and W. Wang, for the BPROAD Research Group*
PMID: 39555827
Introduction
“The blood jet is poetry and there is no stopping it”- Sylvia Plath.
It is universally known that patients with diabetes, in particular type 2 diabetes (T2D), are at a higher risk of adverse cardiovascular outcomes. Maintaining blood pressure (BP) within a “target range” becomes imperative to decreasing CV morbidity and mortality.
The ACCORD trial (Cushman et al, NEJM 2010) compared intensive BP lowering (<120 mmHg) with the standard target (<140 mmHg) in patients with T2D with a HbA1c of 6% in an intensive control arm and 7-7.9% in a standard glycemic arm. Thus, it was a double factorial RCT. Although it reported a reduced stroke risk in the intensive BP arm, the overall primary outcome of major adverse cardiovascular events (HR 0.88, 95% CI, 0.73 to 1.06; p=0.20) was not significantly lower. Rubbing salt in the wound, serious adverse outcomes (SAE) were higher with intensive BP lowering, and all cause mortality was 7% higher (HR 1.07; 95% 0.85 - 1.35, p = 0.55). This put into question intensive BP lowering in T2D, despite attempts from post hoc analyses to demonstrate that there was an effect modification based upon glycemic control (Beddhu et al, JAHA 2018 and Tsujimoto et al, Hypertension 2018). In fact, in the standard glycemic arm, intensive BP lowering was beneficial. Nevertheless, these post hoc analyses failed to temper the belief that intensive BP control was detrimental to patients with T2D.
The SPRINT trial (Wright et al NEJM 2014 | NephJC Summary) examined similar blood pressure targets, but in patients without T2D, and showed a significant decrease in major cardiovascular events (MACE) in the intensive control BP group. However, the lower BP target was associated with increased risk of hypotension and electrolyte abnormalities. Lastly, we had the ESPRIT trial (Liu et al, NEJM 2024 |NephJC Summary) which also demonstrated a very similar reduction in MACE with intensive BP, but included ~ 38% of patients with T2D in the trial population.
With these conflicting results, evidence for lower BP targets has remained inconclusive in patients with T2D. Management of T2D also has changed since ACCORD was conducted. Glycemic targets are now less intensive (particularly in the elderly), and non-insulin, non-sulfonylurea agents (i.e. GLP1-RAs, SGLT2i) are being widely used, which have beneficial effects on CV outcomes. The current BPROAD trial compared intensive BP control (<120 mmHg) with standard BP targets in T2D without consideration for glycemic control.
The Study
Methods
Trial design and monitoring
BPROAD was a parallel-design, randomized clinical trial conducted at 145 clinical sites across China. Recruitment period was 2 years and total trial duration was 5 years. Ethical approval was obtained from each participating site. Patient safety, data collection, and coordination were managed by separate personnel throughout the trial.
Participants
Patients with T2D (on average > 10 years) who were 50 years of age or older, had an elevated systolic blood pressure, and high/very high risk of cardiovascular disease (based upon SCORE2_diabetes calculator) were enrolled. Patients with an eGFR >30 ml/min/1.72m2 (and a serum creatinine <2 mg/dL) were included, however those with > 1 gram of proteinuria or >600 mg albuminuria on 24-hour urine were excluded. Patients with glomerulonephritis, type 1 diabetes or polycystic kidney disease were also excluded.
Randomization
Patients were randomly assigned to receive either intensive (targeted systolic blood pressure <120 mmHg) or standard blood pressure treatment (targeted SBP<140 mmHg). Block randomization was performed within clinical sites, and stratified by regions. Patients and trial physicians were aware of treatment assignments; outcome assessors, adjudicators, and statisticians were blinded.
Interventions
Blood pressures, glucose and lipid levels were managed based on 2018 Chinese Guidelines for the Prevention and Management of Hypertension (Joint Committee for Guideline Revision, J Geriatr Cardiol. 2019), and physician discretion.
In the intensive treatment group, the goal was to achieve SBP <120 mmHg (based on prior trials ACCORD and SPRINT). If SBP <120 was not achieved, a new class of medication was introduced unless there was a valid reason to not change medical therapy. A “milepost exemption form” was filled if no new drug was added, explaining the logic and reasoning.
Treatment Algorithms for Intensive and Standard Treatment Groups, from Bi et al. NEJM, 2024 (protocol).
Initially, in the standard group, if BP ≥160 mmHg at one visit or ≥140 mmHg at two visits, medication dose adjustments were made or additional drugs were added. If BP fell below 130 mmHg (single visit) or 135 mmHg (at two visits), medication could be down-titrated (similar to ACCORD and SPRINT). Monthly visits were conducted for the initial 3 months and then every 3 months if the BP was at target. If blood pressure targets were not achieved, monthly visits continued.
Trial Measurements
Electronic data-capture was used according to standardised procedure.
Each visit: Antihypertensive and concomitant medications, adverse events were recorded.
Clinical-outcome data: Started 3 months after randomization. Continued 3 monthly thereafter.
COVID-19 era: Data was collected via phone calls and blood pressure was recorded by automated devices provided by the trial coordinating center.
Safety monitoring: Query for severe adverse events (SAE) were carried out every 3 months. Participants were also asked to report any SAE in the interim via phone calls or email. The safety committee followed the patient till the resolution of the event.
Trial Outcomes
The primary outcome was a composite of nonfatal stroke, nonfatal myocardial infarction, treatment or hospitalization for heart failure, and death from cardiovascular causes.
Secondary outcomes were represented by the individual components of the primary outcome, plus fatal stroke, fatal myocardial infarction, and death from any cause. CKD outcomes were represented by CKD progression, CKD development, and incident albuminuria.
The authors mention other outcomes in the protocol and the rationale of the study (BPROAD study group, J Diabetes. 2023), without discussing the results, which may be for future consideration.
Primary and secondary outcomes. BPROAD study group, J Diabetes. 2023
Statistical Analysis
A sample size of 12,702 patients was planned to detect a primary CVD event of 2% per year in the standard group and a 20% reduction of these events in the intensive group (Hazard Ratio 0.80). For the primary outcome, analyses were done on an intention-to-treat principle, using time-to-event via Kaplan-Meier, log-rank test, and Cox regression (stratified by region). Two interim analyses were planned at 50 and 75% of the total anticipated number of primary endpoint events. Stopping rules were established using the Lan-DeMets method with O'Brien-Fleming alpha spending. Primary analysis was conducted when 472 of adjudicated events were accrued. Almost three months before the scheduled end of the trial, 667 (77%) of adjudicated events were accrued, while 180 events were in the process of adjudication, therefore trial chair and statistician recommended not performing the second interim analysis. Two-sided nominal significance was updated to 0.045 for the primary outcome after one interim analysis (54.5% primary events).
For secondary outcomes proportional hazard models were used, effects being reported as hazard ratio with 95% confidence intervals, without multiple comparison adjustments. The Fine Gray competing risk model was used for cardiovascular death. For sensitivity analyses, multiple imputations were used for missing data.
Funding Source
The primary sponsor was The National Key Research and Development Program of the Ministry of Science and Technology. No author declared any conflict of interest. The authors responsible for drafting the manuscript are employees of the institutions conducting the trial. Data analysis was performed by the trial and data coordinating center. It’s not specified whether the statisticians were independent of trial’s investigators and sponsors.
Results
Study Population
Out of 16,862 eligible patients, 12,821 were randomized from February 2019 to December 2021, with 6,414 assigned to intensive treatment and 6,407 to standard treatment. During a median follow-up of 4.2 years (interquartile range, 2.9 to 4.6 years), 152 patients discontinued the trial intervention, 605 patients were lost to follow-up, and 423 patients withdrew consent but still were included in the analysis.
Figure S5. Trial profile. Bi et al. NEJM, 2024.
The study population was evenly distributed between treatment groups, with an average age of 63.8 years (44.6% were 65 years or older) and 45.3% of participants were female. Both groups shared similar baseline characteristics, including blood pressure (140/76 mmHg), hypertension duration (~12 years), diabetes duration (~10 years), BMI (~26.7), history of clinical cardiovascular disease (22.5%), and HgA1C (7.6%). Both treatment groups had a mean eGFR of just under 88 ml/min/1.73 m², with just over 7% having eGFR <60 ml/min/1.73 m². Approximately 40% in both groups had a urinary ACR >30 mg/g (> 3.4 mg/mmol).
Table 1. Baseline characteristics. Bi et al. NEJM, 2024.
Nearly all participants (99%) were on antihypertensive drugs, with calcium-channel blockers being the most common (59%), followed by RAS inhibitors (58%). The average number of baseline BP medications was 1.7 for both groups.
Table 1. (Continued.) Baseline characteristics. Bi et al. NEJM, 2024.
A similar proportion (98%) were on hypoglycemic medications, primarily metformin (~66%) and insulin (48%), with only ~ 10% on flozins, and ~ 4% on GLP1RAs. All participants were of Asian descent, reflecting the ethnicity of interest but not the global diversity of T2D and hypertension.
Table 1. (Continued.) Baseline characteristics. Bi et al. NEJM, 2024.
At baseline, both groups had similar systolic blood pressure (SBP ~140 mmHg), but after intervention, the intensive-treatment group achieved lower levels (mean SBP 121.6 mmHg and median 118.3 mmHg) ) compared to the standard-treatment group (mean SBP 133.2 mmHg and median 135 mmHg) at 1 year, with about 60% of intensive-treatment patients meeting their target.
Figure 1. Systolic blood pressure in two treatment arms. Bi et al. NEJM, 2024.
There was a steady increase in the number of patients achieving the BP goal in the intensive group, and 70.8 % of the participants had systolic blood pressure less than 120 mmHg during the protocol visit at week 57 (details in table S4). The between-group difference in systolic blood pressure was sustained throughout the trial. By incorporating mortality data directly into the analysis, the joint model provided treatment effect estimates that aren't distorted by patient deaths during the follow-up period (shown in figure S6). Both SBP and DBP measured in the clinic exhibited similar trends.
The intensive-treatment group used more antihypertensive medications (2.1 vs 1.3), with RAS inhibitors being most common and diuretics more frequently used (26.4% vs 10.3%) compared to the standard arm. Other cardiovascular risk factors, including glycated hemoglobin (7.6%), hypoglycemic drugs, statins, BMI, waist circumference, and lipid levels were similar between groups at 48-month visit. Use of flozins doubled from baseline (~ 20%) but GLP1RAs remained the same (~ 4%). Insulin use remained the same (though no dosing data was provided) and metformin use dropped by ~ 3% from baseline. The HbA1c remained the same at baseline and at 48 months.
Table S5. Utilization of antihypertensive and hypoglycemic medications, statin, and aspirin at 48-month visit. Bi et al. NEJM, 2024.
Clinical Outcomes
Over a median follow-up of 4.2 years, primary-outcome events occurred in 393 patients (1.65 per 100 person-years) in the intensive-treatment group and 492 patients (2.09 per 100 person-years) in the standard-treatment group (HR 0.79; 95% CI, 0.69–0.90). Kaplan–Meier curves began to diverge after 1 year of intervention.
Figure 2. Kaplan-Meier curves for the primary outcome. Bi et al. NEJM, 2024.
Secondary outcomes
Fatal or nonfatal stroke occurred in 284 patients in the intensive-treatment group, corresponding to 1.19 events per 100 person-years, compared to 356 patients in the standard-treatment group, with 1.50 events per 100 person-years (HR 0.79; 95% CI, 0.67-0.92). Fatal or nonfatal myocardial infarction, treatment or hospitalization for heart failure, and death from cardiovascular causes were not statistically significant. There was no significant difference between arms in terms of cardiovascular or all-cause death.
Table 2. Primary and secondary outcomes. Bi et al. NEJM, 2024.
The effects of the interventions on the primary outcome were consistent across most prespecified subgroups. However, some of the subgroups were quite small and underpowered to detect any significant effects (i.e. age > 80, patients with previous CKD) and wide confidence intervals.
Figure 3. The effects of the interventions on the primary composite outcome across prespecified subgroups. Bi et al. NEJM, 2024.
At the end of the follow-up, the eGFR was available for ~72% in both arms (table S8). CKD progression and development were similar between the two groups. Incident albuminuria was lower and occurred in 554 patients (11.29 events per 100 person-years) in the intensive-treatment group and 648 patients (13.84 events per 100 person-years) in the standard-treatment group. Sensitivity analyses yielded consistent results across different assumptions for missing outcomes and competing risks (Table S9).
Safety Outcomes
Serious adverse events occurred in 2340 patients in the intensive-treatment group and 2328 patients in the standard-treatment group, with no significant difference between groups. Most conditions of interest were similar (including AKI), though symptomatic hypotension and high serum potassium (>5.5 mmol/L: 2.8% vs. 2%, P=0.003) were more common in the intensive-treatment group.
Table 3. Adverse events. Bi et al. NEJM, 2024.
Discussion
In patients with T2D with elevated systolic blood pressure and cardiovascular risk, intensive treatment significantly reduced major cardiovascular events over a 5 year period (HR 0.79; 95% CI: 0.69–0.90). Symptomatic hypotension and elevated potassium were more common in the intensive treatment arm, but were manageable. The study's main strengths include a large sample size from 145 sites, including hospitals and primary health care institutions, high adherence to intervention, few participants lost to follow-up and duration of follow up.
Limitations
Participants and physicians were not blinded, though outcome assessors were. Diastolic BP differed significantly between groups as expected, potentially confounding the independent effect of systolic BP. Importantly, only 60% of the intensive-treatment group achieved target BP after one year, which again is a common theme in trials of intensive BP lowering. Additionally, during COVID-19 lockdowns, data were collected via telephone with self-reported home BP monitoring in patients. Since this was exclusively a Chinese study, findings may not be generalizable to other populations, though there is no reason to think so. Subgroups of interest to nephrologists including patients with eGFR below 30 ml/min per 1.73 m2 or urine albumin creatinine ratio ≥600 mg/g were excluded.
Evaluating BPROAD: Insights from Past Hypertension Trials
The BPROAD and ACCORD BP trials naturally lend themselves to comparison, as both examined identical systolic blood pressure targets (<120 mmHg vs <140 mmHg) in high cardiovascular risk populations with T2D (Cushman et al. NEJM 2010). Similarly, comparisons with the ESPRIT trial (Liu et al. Lancet, 2024), which included a predominantly Chinese population with a diabetic subgroup, and the SPRINT trial (Wright Jr et al. NEJM 2015), which excluded individuals with diabetes, should also be considered due to their shared focus on the same systolic blood pressure targets.
Trial population and method of BP measurement
BPROAD was the largest trial (n=12,821) examining systolic blood pressure targets (<120 mmHg vs <140 mmHg) in high cardiovascular risk populations with diabetes, approximately triple ACCORD's size. Unlike SPRINT's unattended measurements, BPROAD used attended office BP measurements similar to ESPRIT and ACCORD, though it adapted to home BP monitoring during COVID-19.
Intensive versus standard BP control across RCTs
Achieving Blood Pressure Targets
The intensive treatment group in BPROAD reached mean systolic blood pressure of 121.6 mmHg, while ESPRIT achieved 119.1 mmHg and SPRINT reached 121.5 mmHg. After one year, the BPROAD trial demonstrated a clear separation between arms, with a mean systolic blood pressure of 133.2 mmHg in the standard group. A treatment algorithm similar to that used in the SPRINT and ESPRIT trial was applied. Medications in the standard treatment arm were adjusted if systolic blood pressure dropped below 130 mmHg at a single visit or below 135 mmHg at two consecutive visits, following consultation with participants.
Cardiovascular Outcomes
The ACCORD trial failed to show significant cardiovascular benefits of intensive BP control in T2D, as the observed event rate in the standard group (2.09% per year) was half the expected 4%. In contrast, the SPRINT trial demonstrated CV risk reduction in non-diabetic patients, and the ESPRIT trial, reported reduced vascular events with intensive BP control, regardless of diabetes history. The BPROAD trial extends this much more clearly in the diabetic population.
Though BPROAD, unlike ACCORD, included HHF in its composite endpoint, the overall benefit was primarily due to decreased stroke events (284 vs 356; HR 0.79; 95% CI, 0.67-0.92), which were also lower in ACCORD when seen on their own. It seems that in the Chinese population, the main CV complication is not acute myocardial infarction, but stroke with a ratio of nearly 6:1 (Xin D et al. JACC 2019). Unlike SPRINT and ESPRIT, BPROAD showed no difference in all-cause mortality between groups.
Intensive versus standard BP control across RCTs
Role of glycemic control
While ACCORD BP trial initially showed no benefit from intensive BP control, further analysis revealed that intensive BP lowering reduced CV events in ACCORD's standard glycemic group (HbA1c ≤8.0%) versus intensive glycemia arm (HbA1c <6.0%) (Beddhu et al, JAHA 2018). The BPROAD trial's favorable CV outcomes with HbA1c levels (7.6 ± 1.4% at 48 months) align with ACCORD BP's post-hoc findings, reinforcing that good (but not intensive) glycemic control optimizes cardiovascular benefits from intensive BP management.
Optimizing Blood Pressure Management in Diabetic CKD Patients
Managing high blood pressure effectively is crucial for optimizing CV, not renal health in patients with CKD. While SPRINT included patients with eGFR as low as 20 ml/min/1.73m², other trials had more restrictive criteria: BPROAD excluded eGFR below 30 or creatinine above 2 mg/dL, ACCORD excluded creatinine above 1.5 mg/dL, and ESPRIT excluded eGFR below 45 ml/min/1.73m². Both BPROAD and ACCORD demonstrated that intensive blood pressure treatment reduced albuminuria. Although ACCORD showed lower eGFR in the intensive group, neither ACCORD nor BPROAD found differences in major kidney outcomes. All the major nephrology trials in nondiabetic CKD (i.e., MDRD, AASK and REIN 2) showed no renal benefit from intensive BP control. SPRINT's post-hoc analysis revealed no increase in urinary tubular injury biomarkers with intensive treatment, suggesting the initial eGFR decline reflects reversible hemodynamic changes rather than kidney damage (Malhotra et al, AJKD 2019). CV outcomes and CV death predominate in CKD, so intensive BP lowering should be pursued for that reason, not renal benefits per se.
Clinical Implications and Practice Translation
The BPROAD trial provides robust evidence supporting intensive BP management (target <120 mmHg) in T2D patients, with a meaningful effect size (ARR: 0.44 events/100 person-years) that strengthens the case for implementation in public health policy. Standardized BP measurement, incorporating both office and home/ambulatory monitoring, forms the foundation for accurate treatment decisions. Successful implementation requires proper escalation of antihypertensive medications, vigilant monitoring for adverse effects such as hypotension and hyperkalemia, and thoughtful individualization of targets based on patient factors like age, fall risk, and orthostatic hypotension.
Conclusion
The BPROAD trial demonstrated that intensive blood pressure treatment targeting systolic blood pressure below 120 mmHg significantly reduced MACE, primarily driven by reduced stroke events, in patients with T2D and hypertension. More research on this topic is not needed.