Ashish Upadhyay, MD; Amy Earley, BS; Shana M. Haynes, DHSc; Katrin Uhlig, MD, MS
Note: A draft of the KDIGO blood pressure guideline will be available for public review in the next few months. If interested in participating in the review, please register at www.kidney.org/professionals/kdigo/guidelinesignup.cfm.
Disclaimer: The judgments and interpretations in this article are those of the authors and are not those of the KDIGO guideline Work Group.
Acknowledgment: The authors thank Andrew Levey, MD, for valuable suggestions and critical review of a prior version of the manuscript.
Financial Support: The authors are supported by KDIGO to conduct systematic reviews and provide methods support for developing KDIGO guidelines.
Potential Conflicts of Interest: Drs. Upadhyay and Uhlig: Grant (money to institution): KDIGO. Ms. Earley: Employment (money to institution): KDIGO. Dr. Haynes: Employment (money to institution): Tufts Medical Center. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M10-2790.
Requests for Single Reprints: Katrin Uhlig, MD, MS, Tufts Center for Kidney Disease Guideline Development and Implementation, William B. Schwartz Division of Nephrology, 800 Washington Street, #391, Boston, MA 02111; e-mail, firstname.lastname@example.org.
Current Author Addresses: Drs. Upadhyay, Haynes, and Uhlig and Ms. Earley: Tufts Center for Kidney Disease Guideline Development and Implementation, William B. Schwartz Division of Nephrology, 800 Washington Street, #391, Boston, MA 02111.
Author Contributions: Conception and design: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Analysis and interpretation of the data: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Drafting of the article: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Critical revision of the article for important intellectual content: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Final approval of the article: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Obtaining of funding: K. Uhlig.
Administrative, technical, or logistic support: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Collection and assembly of data: A. Upadhyay, A. Earley, S.M. Haynes, K. Uhlig.
Upadhyay A., Earley A., Haynes S., Uhlig K.; Systematic Review: Blood Pressure Target in Chronic Kidney Disease and Proteinuria as an Effect Modifier. Ann Intern Med. 2011;154:541-548. doi: 10.7326/0003-4819-154-8-201104190-00335
Download citation file:
Published: Ann Intern Med. 2011;154(8):541-548.
The optimal blood pressure target in patients with chronic kidney disease (CKD) is unclear.
To summarize trials comparing lower versus higher blood pressure targets in adult patients with CKD and focus on proteinuria as an effect modifier.
MEDLINE and the Cochrane Central Register of Controlled Trials (July 2001 through January 2011) were searched for reports from randomized, controlled trials with no language restriction.
Authors screened abstracts to identify reports from trials comparing blood pressure targets in adults with CKD that had more than 50 participants per group; at least 1-year follow-up; and outcomes of death, kidney failure, cardiovascular events, change in kidney function, number of antihypertensive agents, and adverse events.
Reviewers extracted data on study design, methods, sample characteristics, interventions, comparators, outcomes, number of medications, and adverse events and rated study quality and quality of analyses for proteinuria subgroups.
Three trials with a total of 2272 participants were included. Overall, trials did not show that a blood pressure target of less than 125/75 to 130/80 mm Hg is more beneficial than a target of less than 140/90 mm Hg. Lower-quality evidence suggests that a low target may be beneficial in subgroups with proteinuria greater than 300 to 1000 mg/d. Participants in the low target groups needed more antihypertensive medications and had a slightly higher rate of adverse events.
No study included patients with diabetes. Trial duration may have been too short to detect differences in clinically important outcomes, such as death and kidney failure. Ascertainment and reporting of adverse events was not uniform.
Available evidence is inconclusive but does not prove that a blood pressure target of less than 130/80 mm Hg improves clinical outcomes more than a target of less than 140/90 mm Hg in adults with CKD. Whether a lower target benefits patients with proteinuria greater than 300 to 1000 mg/d requires further study.
Kidney Disease: Improving Global Outcomes (KDIGO).
The optimal blood pressure target for patients with hypertension and chronic kidney disease (CKD) is unclear. Recent clinical practice guidelines have recommended a lower blood pressure goal of 130/80 mm Hg or less for patients with CKD compared with 140/90 mm Hg or less for those without CKD, because of a higher risk for cardiovascular disease and kidney failure in CKD (1–3). Because proteinuria amplifies both cardiac and renal risks, some guidelines recommend an even lower target for patients with CKD and proteinuria (4–7).
This systematic review summarizes trials in adults with CKD that compare the effects of treatment with a lower versus higher blood pressure target on clinically important outcomes and focuses particularly on proteinuria as an effect modifier.
We developed and followed a standard protocol for this review that built on the evidence review conducted for the Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guideline for management of blood pressure in CKD.
We searched MEDLINE and the Cochrane Central Register of Controlled Trials for articles indexed from 1 July 2001 to 4 January 2011 without language restriction. In addition, we identified articles from the systematic search conducted for the 2004 Kidney Disease Outcomes Quality Initiative clinical practice guideline on hypertension and antihypertensive agents in CKD (3), which included articles indexed in MEDLINE from 1966 to July 2002. Appendix Table 1 provides the search strategy.
Appendix Table 1.
All authors screened abstracts to identify randomized, controlled trials (RCTs) and observational follow-up reports of RCTs comparing blood pressure targets in adults with non–dialysis-dependent CKD. Trials were eligible if they included participants with CKD (glomerular filtration rate [GFR] <60 mL/min per 1.73 m2, elevated urinary albumin level [>30 mg/d, or urinary albumin–creatinine ratio >0.03 g/g or dipstick-positive albuminuria], or elevated urinary protein level [>300 mg/d, or urinary protein–creatinine ratio >0.2 g/g]) (8). Studies had to report at least 1 of the following outcomes: death, kidney failure, clinical cardiovascular events, categorical change in kidney function, rate of change in GFR (GFR slope), and number of antihypertensive agents needed to achieve blood pressure targets. Adverse events of interest were patient withdrawal because of adverse events, decrease in drug dose or discontinuation of therapy, and symptoms related to hypotension. We excluded studies with fewer than 50 participants per group or follow-up shorter than 1 year.
One of the authors extracted data from each article, and another author confirmed the data extraction. We extracted data on study design, methods, sample characteristics, interventions, comparators, outcomes, number of medications, and adverse events.
We graded overall study quality as good, fair, or poor (9–11). Studies were graded by one of the authors, which was confirmed by another author and finalized in a meeting of all authors. Two authors perused the inclusion and exclusion criteria of the trials to assess applicability.
We assessed the quality of the subgroup analyses by baseline proteinuria on the basis of recently proposed criteria for reporting and interpreting subgroup analyses (12, 13). Although we examined published articles and articles on study designs, we did not contact investigators for unpublished answers to our quality questions.
Data were summarized in tables and synthesized in narrative form. Given the limited number of trials and the clinical heterogeneity for types of CKD, measures of proteinuria, and outcome definitions, we did not perform quantitative meta-analysis.
The authors are supported by KDIGO to conduct systematic reviews and provide methods support for developing KDIGO guidelines, including the ongoing guideline on management of blood pressure in CKD. The findings were presented to the KDIGO guideline Work Group, and the manuscript was confidentially shared with the Work Group Chairs, who approved submission but did not provide feedback or comments on the manuscript. The funding source did not participate in the design, conduct, or reporting of the study.
The Figure summarizes the search yield. Eight reports from 3 RCTs with a total of 2272 patients were included in this review (14–21). No articles examined blood pressure targets exclusively in patients with CKD and diabetes.
BP = blood pressure; KDIGO = Kidney Disease: Improving Global Outcomes; KDOQI = Kidney Disease Outcomes Quality Initiative.
The 3 RCTs included in this review are the MDRD (Modification of Diet in Renal Disease) Study (14–16), the AASK (African American Study of Kidney Disease and Hypertension) Trial (17, 18), and the REIN-2 (Ramipril Efficacy in Nephropathy 2) trial (19). The MDRD Study and AASK Trial also have reports on posttrial observational follow-up (20, 21). All trials provide subgroup analyses by baseline proteinuria levels. Trial characteristics are summarized in Table 1.
The MDRD Study A included 585 patients with GFRs of 25 to 55 mL/min per 1.73 m2 and Study B included 255 patients with GFRs of 13 to 24 mL/min per 1.73 m2. Patients with diabetes who required insulin were excluded from the study, and only 3% of patients were classified as having diabetic nephropathy. The trial phase had a 2 × 2 factorial design, and patients were randomly assigned to a low or usual blood pressure target and 1 of 2 types of diet. The use of all antihypertensive agents was allowed, but angiotensin-converting enzyme (ACE) inhibitors, with or without a diuretic, were encouraged as the first-choice agents and calcium-channel blockers, with or without a diuretic, were encouraged as the second-choice agents. In the posttrial follow-up observation lasting a mean of 6 years, no specific blood pressure target or antihypertensive agent was recommended.
The AASK Trial included black adults with hypertensive nephrosclerosis. Patients with diabetes were excluded from the study. The trial phase had a 3 × 2 factorial design, and patients were randomly assigned to a low or usual blood pressure target and therapy with 1 of 3 initial antihypertensive agents (ACE inhibitor, β-blocker, or dihydropyridine calcium-channel blocker). The trial protocol allowed the sequential addition of furosemide, doxazosin, clonidine, hydralazine, and minoxidil to achieve the randomized blood pressure target. In the 8- to 12-year posttrial follow-up observation, all patients were assigned the blood pressure target of less than 130/80 mm Hg and were treated with either an ACE inhibitor or an angiotensin-receptor blocker.
The REIN-2 trial included patients with proteinuria greater than 1000 mg/d for at least 3 months. Patients with proteinuria between 1000 and 3000 mg/d were included if their creatinine clearance was less than 45 mL/min per 1.73 m2, and patients with proteinuria greater than 3000 mg/d were included if their creatinine clearance was less than 70 mL/min per 1.73 m2. Patients with type 1 diabetes mellitus were excluded. All patients were treated with the ACE inhibitor ramipril, 5 mg/d, during the trial. The dihydropyridine calcium-channel blocker felodipine, 5 to 10 mg/d, was an add-on therapy in the group assigned to the low blood pressure target. Antihypertensive agents other than ACE inhibitors, angiotensin-II–receptor blockers, and dihydropyridine calcium-channel blockers were allowed in both groups.
Methodological quality of the 3 trials was graded as good. Quality of the follow-up reports was graded as fair because the original trial interventions were not continued and, in the MDRD Study, the choice of agents and blood pressure targets were not defined.
Table 2 shows the quality assessment of the subgroup analyses by baseline proteinuria levels. Published articles reporting results and study design were reviewed for quality assessment (15–22).
Table 3 summarizes the trial and follow-up results, and Appendix Table 2 summarizes the results from the proteinuria subgroup and interaction analyses.
Appendix Table 2.
The MDRD Study, AASK Trial, and REIN-2 trial failed to show benefit for clinical outcomes from the low versus usual blood pressure targets. Although the point estimates in the MDRD Study and AASK Trial suggested that the low target might reduce kidney failure, the CIs around the estimates were wide and included the possibility of either important benefit or harm. The only statistically significant result was in the MDRD Study follow-up, which showed a 23% reduction (95% CI, 18% to 43%) in the hazard for kidney failure in the group assigned to the low target. Data on deaths and cardiovascular disease outcomes were not informative given the lack of ascertainment or low event rates.
The cut points and measures for proteinuria assessment varied across studies. A total of 11 subgroup results were reported; 7 showed benefits for the low blood pressure target in higher proteinuria subgroups (Appendix Table 2). Benefit was seen in subgroups with a urinary protein–creatinine ratio greater than 0.22 g/g in the AASK Trial (considered to correlate with urinary protein excretion >300 mg/d) and urinary protein excretion greater than 1000 mg/d in the MDRD Study. One subgroup analysis in the follow-up report from the AASK Trial showed that the usual blood pressure target was beneficial in patients with a urinary protein–creatinine ratio of 0.22 g/g or less for the composite outcome of a 50% decrease in GFR or kidney failure.
Inconsistencies between interaction tests and subgroup analyses were found within and across studies (Appendix Table 2). In the MDRD Study, overall results for the rate of decrease in GFR were negative, but the low blood pressure target was beneficial in high proteinuria subgroups (>1000 mg/d in MDRD Study A and >3000 mg/d in MDRD Study B). The interaction tests by baseline proteinuria were also positive, suggesting benefit of the low blood pressure target mainly in people with high proteinuria. In the follow-up report from the MDRD Study, however, the result for kidney failure was favorable for the low blood pressure target. Subgroup analysis showed that the low target was beneficial in patients with proteinuria greater than 1000 mg/d, but the interaction test by baseline proteinuria was not statistically significant, suggesting benefit of the low target in all proteinuria levels. In the AASK Trial, the results for the primary composite clinical outcome were negative, whereas the subgroup analysis showed that the low blood pressure target was beneficial in patients with urinary protein–creatinine ratios greater than 0.22 g/g in both the trial and follow-up. For the other composite outcomes, however, subgroup analyses and corresponding interaction tests favored lower blood pressure targets in the high proteinuria group only during follow-up.
In the MDRD Study, the mean number of antihypertensive agents prescribed per patient for the low and usual blood pressure target groups was 1.9 and 1.5, respectively, in Study A and 2.1 and 1.8 in Study B. In the AASK Trial, the mean number of antihypertensive agents was not reported, but the total number of drug classes used during the trial was 3.0 in the low target group and 2.4 in the usual target group.
In the MDRD Study, participants were asked to report on 24 symptoms attributable to low blood pressure during each follow-up visit. Of the 24 symptoms, only the frequency of “feeling faint” differed significantly between the low and usual blood pressure target groups in both studies (Study A and Study B). “Feeling faint” was reported in 15% of visits per patient in the low target group compared with 12% in the usual target group in Study A (P = 0.0012) and 18% of visits per patient in the low target group compared with 12% in the usual target group in Study B (P = 0.009). More patients in the low target group also needed a reduction in antihypertensive medications because of persistent symptoms of hypotension (3.2% vs. 0.7% in the usual target group; P = 0.01). The incidence of conditions requiring patient withdrawal from the study, however, did not differ between the groups.
In the AASK Trial, all participants were asked about shortness of breath, syncope, dizziness, and lightheadedness. In addition, the investigators also reported on the incidence of hyperkalemia, angioedema, edema, cough, and sexual dysfunction. Among all adverse events, only cough was more frequent in the low target group than in the usual target group (55% vs. 47%; P < 0.05). This is not explained by the greater use of ACE inhibitors in the low target group, because the 3 × 2 factorial design ensured similar distribution of drug classes in the 2 groups.
In the REIN-2 trial, 6 of 169 participants in the low target group and 3 of 169 participants in the usual target group had treatment-related adverse events that necessitated withdrawal from the trial. A total of 23% of participants in the low target group and 17% in the usual target group had serious adverse events, but specific events were not reported.
This systematic review of RCTs in adults with CKD did not find conclusive evidence favoring a blood pressure target of less than 125/75 to 130/80 mm Hg rather than a target of less than 140/90 mm Hg. After a mean 2- to 4-year follow-up, the main trial results did not show benefit for clinical outcomes (14–19). Only the posttrial follow-up report from the MDRD Study showed benefit of the lower target for kidney failure after about a 6-year follow-up (21). Subgroup analyses by baseline proteinuria levels in the MDRD Study and AASK Trial, but not in the REIN-2 trial, suggest benefit for the lower target in patients with proteinuria greater than 1000 mg/d and urinary protein–creatinine ratio greater than 0.22 g/g, respectively (15–21). Treatment to a lower target required, on average, 0.3 to 0.6 additional antihypertensive drugs per patient (14, 17, 20). A slightly higher rate of adverse events was suggested in the low target groups (14, 17, 19).
Clinical heterogeneity in study populations and interventions may have resulted in some of the inconsistencies across trials. Trials studied participants with different causes of CKD and used different antihypertensive regimens. Only the AASK Trial had, by design, a balanced use of an ACE inhibitor, a β-blocker, and a calcium-channel blocker in each group. The benefit for kidney failure in the MDRD Study follow-up may have been confounded by the greater use of ACE inhibitors in the low target group than in the usual target group (51% vs. 32%). The lack of any benefit in the higher proteinuria subgroup of the REIN-2 trial was in the setting of all patients receiving an ACE inhibitor and the low target group also receiving the dihydropyridine calcium-channel blocker felodipine.
Trial reports were graded as good quality, whereas the reports from posttrial follow-up and subgroup analyses were graded as fair quality. Confidence in subgroup findings is enhanced if the purpose and categories for analysis are prespecified in the study design (12, 13, 23–25). The REIN-2 trial was the only study that had prespecified subgroup analysis by baseline proteinuria levels, but it did not report on interaction testing. Cut points and measurement techniques used for proteinuria subgroup analyses varied in all 3 studies. Results were inconsistent between interaction tests and subgroup analyses for the same or different outcomes, both within and across studies. Exploratory subgroup analyses, although important in generating hypothesis for future research, are often underpowered, susceptible to spurious results because of multiple testing, and vulnerable to reporting and publication biases (12, 13).
A major limitation of the evidence base is that the 3 trials excluded people with type 1 diabetes mellitus and included very few patients with diabetic kidney disease. In addition, trial duration may have been too short to detect differences for clinically important outcomes, such as death and kidney failure. We captured delayed effects by including reports from the posttrial follow-up, but these were limited by the possibility of confounding from lack of randomization. We also cannot discount publication bias or selective reporting of some outcomes or subgroup analyses, because we did not contact investigators for unpublished data. Studies with small sample sizes and short follow-up were excluded, but they probably would not have changed our conclusions. Ascertainment and reporting of adverse events were also inconsistent across studies.
To put our findings in the context of the current literature, we conducted a separate MEDLINE search to identify reviews on this topic published between 1 January 2008 and 31 December 2010 (search terms are provided in Appendix Table 3). We believe ours is the only review specifically in patients with CKD and with a detailed analysis of effect modification by baseline proteinuria. The 2009 Cochrane review by Arguedas and colleagues (26) in patients with hypertension included the CKD trials and also found no benefit for a target lower than 140/90 mm Hg to 160/100 mm Hg for mortality or morbidity. The reviews by Rosendorff and Black (27) and Roy and colleagues (28) highlight limited evidence for a low blood pressure target in patients with cardiovascular disease and raise concern about impaired coronary perfusion from aggressive lowering of diastolic blood pressure in patients with coronary artery disease or left ventricular hypertrophy.
Appendix Table 3.
In addition to these reviews, the ACCORD (Action to Control Cardiovascular Risk in Diabetes) trial provides important supplemental information (29). The ACCORD trial compared systolic blood pressure targets of less than 120 mm Hg and less than 140 mm Hg in more than 4000 patients with diabetes at high risk for cardiovascular outcomes. Most participants had normal kidney function, but approximately 40% had micro- or macroalbuminuria. The main results did not show any difference in the primary composite outcome of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death between the 2 groups, but GFR decreased below 30 mL/min per 1.73 m2 more often and more serious adverse events were attributed to antihypertensive treatment in the low target group. Finally, several recent commentaries and editorials have also highlighted the controversies surrounding the topic of blood pressure targets in CKD (30–35).
Future research needs to address the evidence gap for patients with diabetes and CKD. A subgroup analysis of persons with elevated albuminuria in the ACCORD trial would be informative in this regard. For patients without diabetes, the ongoing SPRINT (Systolic Blood Pressure Intervention Trial) (36) and the HALT-CKD (HALT Progression of Polycystic Kidney Disease) study (37) could provide valuable data. SPRINT compares systolic blood pressure targets of less than 120 mm Hg and less than 140 mm Hg in 9000 patients with hypertension and without diabetes; it will be completed by 2018. It is enrolling participants with CKD, cardiovascular diseases, or risk factors for cardiovascular diseases. Similarly, the HALT-PKD study compares the efficacy of blood pressure targets of 95/60 to 110/75 mm Hg and 120/70 to 130/80 mm Hg in patients with autosomal dominant polycystic kidney disease; it will be completed by 2013. Although both SPRINT and the HALT-PKD study excluded participants with heavy proteinuria (urinary albumin–creatinine ratio >0.6 g/g for SPRINT and >0.5 to 1.0 g/g for HALT-PKD), subgroup analysis in the lower proteinuria range should be considered for these trials to better understand the effect modification by proteinuria. For ACCORD, SPRINT, HALT-PKD, and other future trials, the use of well-established albuminuria, proteinuria, and GFR categories will be helpful to harmonize and compare subgroup results across studies (38).
In summary, evidence does not conclusively show that a currently recommended blood pressure target of less than 130/80 mm Hg improves clinical outcomes more than a conventional target of less than 140/90 mm Hg in adults with CKD. A lower target may be beneficial in persons with proteinuria greater than 300 to 1000 mg/d. We suggest that practitioners use discretion in patients with CKD and proteinuria and base the blood pressure target on individualized risk–benefit assessment and the patient's tolerance and preferences. Treatment to a lower target may require greater vigilance to monitor for and avoid possible symptoms and adverse events from hypotension.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Nephrology, Urological Disorders, Chronic Kidney Disease.
Results provided by:
Copyright © 2016 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use
This PDF is available to Subscribers Only