Blood pressure targets in patients with chronic kidney disease: A critical evaluation of clinical‐trial evidence and guideline recommendations

1. INTRODUCTION

Hypertension among people with chronic kidney disease (CKD) is an established risk factor for acceleration of atherosclerotic cardiovascular disease and progression of kidney injury to end‐stage renal disease (ESRD). ¹ , ² Randomized trials have demonstrated that blood pressure (BP)‐lowering is translated into a profound benefit on cardiovascular and kidney outcomes. ¹ However, the optimal levels at which BP should be targeted remains an area surrounded by substantial controversy that is reflected by the discrepancy in BP target recommendations for patients with CKD across guidelines released during 2012‐2018 (Table 1).

TABLE 1.

Recommended in guidelines BP targets for management of hypertension in CKD

	2018 ESH/ESC 3	2017 ACC/AHA 4	2014 JNC8 5	2012 KDIGO 2
Stage 3‐5 CKD without albuminuria	130‐139/80‐89 mmHg	<130/80 mmHg	<140/90 mmHg	<140/90 mmHg
Stage 1‐5 CKD with albuminuria a	130‐139/80‐89 mmHg	<130/80 mmHg	<140/90 mmHg	<130/80 mmHg

Abbreviations: ACC/AHA, American College of Cardiology/American Heart Association; ESH/ESC, European Society of Hypertension/European Society of Cardiology; JNC, Joint National Committee on Detection, Evaluation, and Treatment of High BP report; KDIGO, Kidney Disease Improving Global Outcomes.

^aAlbuminuria is defined as albumin excretion rate ≥30 mg/24‐h, approximating an albumin‐to‐creatinine ratio ≥30 mg/g.

The 2012 Kidney Disease Improving Global Outcomes (KDIGO) guideline recommended a lower BP target of <130/80 mmHg only for CKD patients with persistent albuminuria [defined as albuminuria ≥30 mg/24‐h or urinary‐albumin‐to‐creatinine‐ratio (UACR) ≥30 mg/g in a random urine specimen]. ² In contrast, the 2014 guideline released by the eighth Joint National Committee (JNC8) did not differentiate the BP targets by the level of albuminuria. There was a uniform recommendation to lower BP to levels <140/90 mmHg regardless of the level of kidney function or UACR. ⁵

The controversy was further magnified in more recent guidelines released after the publication of the Systolic Blood Pressure Intervention Trial (SPRINT). ⁶ This landmark trial was prematurely terminated because of an impressive cardiovascular benefit with a lower (<120 mmHg) vs a standard (<140 mmHg) systolic BP (SBP) target in non‐diabetic patients with clinic BP ≥ 130/80 mmHg and a high cardiovascular risk profile. Based on the results of SPRINT, the 2017 American Heart Association/American College of Cardiology (AHA/ACC) guideline reappraised the definition of hypertension and recommended a lower BP target of <130/80 mmHg for all patients at high cardiovascular risk, including those with CKD. ⁴ The European Society of Hypertension/European Society of Cardiology (ESH/ESC) did not follow the same approach. In their 2018 guideline, the recommendation for patients with CKD was to lower SBP within the range of 130‐139 mmHg and to individualize the intensity of treatment based on its tolerability. ³

In this article, we explore the above‐described heterogeneity across recent guidelines, providing a critical evaluation of randomized trials that compared a lower vs a standard BP target among patients with CKD. We raise the issue of research‐grade BP monitoring method that was implemented in SPRINT and we ask the question whether the impressive results of this trial can be translated into benefit for our patients when intensification of antihypertensive therapy is guided by routine clinic BP recordings.

2. CLINICAL‐TRIAL EVIDENCE

The effect of intensive BP‐lowering on kidney outcomes was explored in three separate randomized trials that enrolled non‐diabetic patients with CKD (Table 2). The Modification of Diet in Renal Disease (MDRD) followed a 2 × 2 factorial design and randomized 840 patients with GFR 13‐55 mL/min/1.73 m² to two different levels of dietary protein intake and to achieve a lower mean BP (MBP) of <92 mmHg (approximately <125/75 mmHg) vs a standard MBP of 102‐107 mmHg (approximately 135/85‐140/90 mmHg). ⁷ Over a 2.2‐year‐long follow‐up, the rate of GFR decline did not differ between the lower‐ and standard‐MBP arms. Intensive BP‐lowering did not improve the combined outcome of ESRD or death [relative risk (RR): 0.85; 95% confidence interval (CI): 0.60‐1.22]. ⁷ In the African American Study of Kidney Disease and Hypertension (AASK), 1094 African Americans with hypertensive nephrosclerosis were randomized to a lower vs a standard MBP target (<92 mmHg vs 102‐107 mmHg) and to initiate antihypertensive therapy with metoprolol (50‐200 mg/d), ramipril (2.5‐10 mg/d), or amlodipine (5‐10 mg/d) in a 2 × 3 factorial design. ⁸ Over a follow‐up of 3‐6.4 years, compared with the standard MBP target, intensive BP‐lowering had no benefit on the combined outcome of ≥50% decline in GFR, ESRD, or death (risk reduction: 2%; 95% CI: −22% to 21%). ⁸ In the Ramipril Efficacy in Nephropathy trial 2 (REIN‐2), 338 patients with non‐diabetic, proteinuric CKD already treated with ramipril (2.5‐5 mg/d) were randomized to intensified BP‐lowering (goal BP <130/80 mmHg) or to standard BP control (diastolic BP <90 mmHg). ⁹ Add‐on therapy with felodipine (5‐10 mg/d) was administered to achieve the assigned BP targets. REIN‐2 was prematurely terminated at a median follow‐up of 19 months due to futility. The risk of ESRD was similar in the intensive‐ and standard‐BP groups (HR: 1.00; 95% CI: 0.61‐1.64). ⁹ Thus, during their randomized phase, none of these three trials showed that intensive BP‐lowering is an effective strategy to halt the progression of kidney injury to ESRD.

TABLE 2.

Randomized trials testing a lower vs a standard‐BP target among patients with CKD

Trial	Year	N	Participant characteristics	BP targets: lower vs usual	Primary outcome	Follow‐up	Primary result
MDRD 7	1994	840	Non‐diabetic CKD	MBP <92 vs <107 mmHg	Rate of change in GFR	2.2 y	No between‐group difference
AASK 8	2002	1094	African Americans with hypertensive CKD	MBP <92 vs 102‐107 mmHg	Composite of > 50% reduction in eGFR, ESRD or death	3‐6.4 y	No between‐group difference
REIN‐2 9	2005	338	Non‐diabetic proteinuric CKD	BP <130/80 vs diastolic BP <90 mmHg	Incident ESRD	19 mo	No between‐group difference
SPRINT (CKD sub‐study) 10	2017	2646	Non‐diabetic patients with CKD and systolic BP 130‐180 mmHg	Systolic BP <120 vs <140 mmHg	Myocardial infarction, other acute coronary syndromes, stroke, heart failure, or cardiovascular death	3.26 y	Non‐significant reduction by 19% in the occurrence of the primary cardiovascular outcome

Abbreviations: AASK, African American Study of Kidney Disease and Hypertension; BP, blood pressure; CKD, chronic kidney disease; ESRD, end‐stage renal disease; GFR, glomerular‐filtration rate; MBP, mean blood pressure; MDRD, Modification of Diet in Renal Disease; REIN‐2, Ramipril Efficacy in Nephropathy trial 2; SPRINT, Systolic Blood Pressure Intervention Trial.

Post hoc analyses of the MDRD and AASK trials, however, support the notion that the level of proteinuria may act as treatment effect modifier. When MDRD participants were stratified into subgroups by the level of proteinuria at baseline, intensive BP‐lowering was associated with a slower rate of GFR decline during follow‐up in those with baseline proteinuria of 1‐3 g/d or >3g/d. In the proteinuria stratum of <1 g/d, the rate of GFR decline did not differ between the lower‐ and standard‐MBP arms. ⁷ After the completion of the trial phase of AASK, participants were inserted into a cohort phase during which their BP was targeted to levels <130/80 mmHg. Over the entire follow‐up of 9.1 years (trial plus cohort phase), the composite outcome of doubling of serum creatinine, ESRD, or death did not differ between those initially randomized to the lower vs those initially randomized to the standard MBP target (HR: 0.91; 95% CI: 0.77‐1.08). ¹¹ However, there was a significant interaction between the randomized arm and level of baseline proteinuria for the composite kidney outcome (P = .02 for the interaction). In those with proteinuria >0.22 g/d, initial randomization to the lower BP target was associated with 27% reduction in the composite kidney outcome (HR: 0.73; 95% CI: 0.58‐0.93). This benefit was not evident in those with baseline proteinuria ≤0.22 g/d (HR: 1.18; 95% CI: 0.93‐1.50). ¹¹ Another post hoc analysis incorporating data from 840 MDRD participants with an extended follow‐up of 9.2 years (trial plus cohort phase) showed that compared with the standard BP target, those initially randomized to intensive BP‐lowering had 32% reduced risk of developing ESRD (HR: 0.68; 95% CI: 0.57‐0.82). ¹²

Based on this evidence, the 2012 KDIGO guideline recommended a lower BP target of <130/80 mmHg for CKD patients with persistent albuminuria. ² It has to be noted, however, that the strength of this recommendation was labeled as level 2 and the quality of evidence supporting this guidance was graded as level C. This guidance was based mainly on low‐quality evidence from the aforementioned post hoc analyses of the MDRD and AASK trials. The observational nature of this data cannot demonstrate direct cause‐and‐effect associations. This scientific basis was considered insufficient/weak to mandate a reappraisal of BP targets for CKD patients in JNC8. ⁵ Thus, the discrepancy between these two guidelines is not so large. In our interpretation, a level 2C recommendation by the 2012 KDIGO guideline is a clear recognition of the gap in the existing evidence and a call for future research.

The effect of intensive BP‐lowering on cardiovascular morbidity and all‐cause mortality was evaluated in a prespecified post hoc analysis of 2,646 SPRINT participants with eGFR of 20‐59 mL/min/1.73 m². ¹⁰ The separation between the intensive‐ and standard‐arm was an average difference of −12.3 mmHg in SBP levels. Intensively treated participants required on average ~1 additional antihypertensive medication to achieve the assigned SBP target. This post hoc analysis showed the absence of interaction between the randomized arm and the level of eGFR at baseline for the primary cardiovascular outcome of SPRINT (P ≥ .30 for the interaction). The occurrence of the primary cardiovascular outcome did not differ between the intensive‐ and standard‐arm (HR: 0.81; 95% CI: 0.63‐1.05), ¹⁰ possibly because this post hoc analysis was underpowered to detect a significant between‐arm difference in the subgroup of SPRINT participants with eGFR <60 mL/min/1.73 m². However, intensive BP‐lowering provoked a significant 28% reduction in the risk of all‐cause death (HR: 0.72; 95% CI: 0.53‐0.99). ¹⁰

Intensive BP‐lowering in this post hoc analysis of SPRINT had no benefit on the composite kidney outcome of ≥50% decline in eGFR or ESRD (HR: 0.90; 95% CI: 0.44‐1.83). ¹⁰ Notably, this composite kidney outcome occurred in only 15 participants in the intensive‐arm and in 16 participants in the standard‐arm. This low incidence rate is not surprising, since patients with proteinuria >1 g/d who would presumably carry a higher risk of kidney injury progression were excluded from SPRINT. Although this trial was not originally designed to assess the efficacy of intensive BP‐lowering on kidney outcomes like MDRD and AASK, the results of SPRINT were interpreted as a proof of the cardioprotective benefit of this strategy and provided the scientific basis to recommend a lower BP target of <130/80 mmHg for patients with CKD in the 2017 AHA/ACC guideline. ⁴ The same clinical‐trial evidence was interpreted differentially by the 2018 ESH/ESC guideline ³ and the question that arises is whether the more conservative approach of European hypertension specialists is reasonable and evidence‐based. Of note, the 2018 ESH/ESC guideline ³ provided a more conservative recommendation on BP targets particularly for patients with CKD and not for other patient populations. Whether this differentiation is attributable to issues related to the tolerability of intensive BP‐lowering in the CKD setting (ie, higher prosperity of these patients to adverse events, higher risk of acute kidney injury episodes, etch) or to other reasons remains another area of uncertainty.

3. RESEARCH‐GRADE VS ROUTINE CLINIC BP

The diagnostic accuracy, reproducibility, and predictive value of BP recordings taken at the environment of clinic should be interpreted within the context of the actual methodology implemented (such as, type of BP monitor, the seated rest period before BP measurement, the presence of observer, the number of BP recordings). ¹³ In SPRINT, clinic BP was measured with a fully automated oscillometric device that obtained triplicate recordings after a 5‐minute seated rest and often without the presence of observer. ⁶ This research‐grade BP monitoring technique differs substantially from routine clinic BP recordings. A meta‐analysis of 31 diagnostic‐test studies (incorporating data from 9279 participants) showed that fully automated clinic SBP was similar with the reference‐standard ambulatory daytime SBP [mean difference (MD): 0.3 mmHg; 95% CI: −1.1 to 1.7 mmHg]. ¹⁴ In contrast, routine clinic SBP overestimated ambulatory daytime SBP by 14.5 mmHg (MD: 14.5 mmHg; 95% CI: 11.8‐17.2 mmHg). ¹⁴ The importance of standardization in BP measurement methodology was highlighted by another meta‐analysis of 10 diagnostic‐test studies showing that unattended recordings were similar with attended clinic BP recordings, when the same device and measurement methodology was implemented. ¹⁵ These meta‐analyses, however, quantified only the average differences between BP monitoring techniques and did not explore their actual levels of agreement.

In a diagnostic‐test study that included 275 patients with stage 3‐4 CKD and clinic BP <140/90 mmHg, participants underwent clinic BP monitoring with the research‐grade technique that was implemented in SPRINT. On the same day, participants had their clinic BP recorded without specification of a 5‐minute seated rest. ¹⁶ Research‐grade clinic SBP was by 12.7 mmHg lower than routine clinic SBP (MD: −12.7 mmHg; 95% CI: −14.7 to −10.7 mmHg) and the 95% levels of agreement between these two techniques were ranging from −46.1 to 20.7 mmHg. Research‐grade clinic SBP underestimated the reference‐standard ambulatory daytime SBP by 7.9 mmHg (MD: −7.9 mmHg; 95% CI: −9.4 to −6.4 mmHg), whereas the 95% levels of agreement were once again wide and were ranging from −33.2 to 17.4 mmHg. Whereas fully automated clinic SBP and ambulatory daytime SBP were significantly associated with echocardiographically documented left ventricular hypertrophy, routine clinic SBP could not detect evidence of target‐organ damage. ¹⁶

Taken together, the above evidence from diagnostic‐test studies suggests that the differentiation in recommended BP targets between the American and European guidelines may be simply the tip of the iceberg. Guideline groups have taken into consideration the above‐described variability between research‐grade and routine clinic BP recordings. Recognizing that the adoption of a research‐grade BP monitoring technique like that of SPRINT in daily clinical practice would probably be problematic, the recommended BP targets were adjusted to a higher level from that of the achieved clinic SBP of 121.2 mmHg in the intensive‐arm of SPRINT. In our interpretation, a lower or a higher algebraic adjustment for the average bias inserted by routine clinic BP recordings remains an oversimplification. ¹⁷ An algebraic adjustment for the average differences provides little to no reflection of the actual variability between research‐grade and routine clinic BP at the level of individual patients, given the wide 95% levels of agreement between these two techniques in diagnostic‐test studies. ¹⁶ Thus, implementation of intensive BP targets in daily clinical practice necessitates the optimization of our BP measurement methodology. In other words, if antihypertensive therapy continues to be guided by what is already done in routine clinical practice, then intensive BP‐lowering may not be beneficial.

4. CONCLUSION

Evidence to support that intensive BP‐lowering to levels <130/80 mmHg is an effective strategy to delay the progression of kidney injury to ESKD is weak. All randomized trials that compared a lower vs a standard BP target in patients with CKD failed to show a benefit of intensive BP‐lowering on kidney outcomes. ⁷ , ⁸ , ⁹ The notion that intensive BP‐lowering is beneficial for those with proteinuric CKD is based on low‐quality evidence from post hoc analysis of the MDRD and AASK with an extended observational follow‐up after the completion of the randomized phase of these two trials. ¹¹ , ¹² In accordance with the results of MDRD, AASK, and REIN‐2, in a post hoc analysis of 2646 SPRINT participants with eGFR <60 mL/min/1.73 m², intensive BP‐lowering did not improve the composite kidney outcome. ¹⁰ This post hoc analysis, however, provided some evidence to support a potential benefit of intensive BP‐lowering on survival and cardiovascular outcomes. ¹⁰ Even if we interpret this evidence as a proof of cardioprotection, we believe that this benefit may not be generalizable to the majority of our patients if antihypertensive therapy continues to be guided by BP recordings taken under routine clinical practice conditions. In our view, individualization of the intensity of therapy based on its tolerability and optimization of our BP measurement methodology represent important steps to improve BP control and clinical outcomes in this high‐risk population. The wider adoption of home or ambulatory BP monitoring is another important step to improve the management of hypertension among patients with CKD.

CONFLICT OF INTEREST

The authors have no conflicts of interest to disclose.

AUTHOR'S CONTRIBUTION

Literature search: PIG, VV; Drafting the manuscript: PIG; Revisions on the initial draft: PEZ, VL; Approval of the final paper: VV, VL, PEZ.

Georgianos PI, Vaios V, Zebekakis PE, Liakopoulos V. Blood pressure targets in patients with chronic kidney disease: A critical evaluation of clinical‐trial evidence and guideline recommendations. J Clin Hypertens. 2020;22:924–928. 10.1111/jch.13859