Can an ACE inhibitor reduce the risk of 
developing progressive renal disease?

A 50-year-old African American woman presents to the outpatient clinic for her annual checkup. She has chronic essential hypertension that has been well-controlled with hydrochlorothiazide 25 mg for 15 years. She takes no other medications and has no known drug allergies. She reports 6 months of menstrual irregularities; her review of systems is otherwise negative. On physical examination, vital signs include a BP of 128/80 mm Hg and body mass index of 28 kg/m². The remainder of the examination is unremarkable. Routine laboratory testing reveals a fasting glucose level of 80 mg/dL, BUN of 18 mg/dL, serum creatinine of 1.1 mg/dL, and an estimated glomerular filtration rate (eGFR) of 63 mL/min/1.73 m². Urinalysis by dipstick is notable for 1+ protein but no other abnormalities. 


African American patients are reported to have a higher incidence of end-stage renal disease (ESRD)¹ with current research implicating potential genetic factors that lead to progressive disease² or underrecognition of early-stage chronic kidney disease (CKD) in subsets of the African American population by conventional GFR calculation methods.^3,4 The authors sought guidance from the literature on the best therapeutic approach to minimize the likelihood of this patient's renal disease progressing because she demographically has been identified as a patient who may receive a delayed diagnosis of CKD or be at increased risk of progressing to ESRD.


CLINICAL QUESTION


Does adding an ACE inhibitor to the antihypertensive regimen of nondiabetic African American patients at goal BP impede the progression of hypertensive renal disease?


SEARCH CRITERIA AND 
RESULTS 


A search of peer-reviewed literature was conducted through PubMed using the therapy clinical query tool with a narrow filter and the search terms African American, hypertensive, end-stage renal disease, angiotensin-converting enzyme inhibitor, AND diseaseprogression. The search was limited to clinical trials with human subjects that were written in English and had been published within the past 10 years. This search strategy generated eight articles, six of which were excluded: three that studied outcomes irrelevant to the clinical question, one that summarized without uniquely analyzing a randomized controlled trial chosen for review, one that was an interim analysis of that same trial, and one that was a subset analysis of a cohort study chosen for review. This left one randomized controlled trial and the cohort study for review.^5,6


EVALUATING THE EVIDENCE 


The African American Study of Kidney Disease and Hypertension (AASK trial) was a randomized 3×2 factorial trial conducted in the United States from 1995 to 2001.⁵ The trial recruited 1,094 self-identified African American males and females aged 18 to 70 years from 21 clinical centers throughout the United States with a diagnosis of primary hypertension, GFR between 20 and 65 mL/min/1.73 m², and no known diabetes mellitus or other identifiable causes of renal insufficiency except for hypertension. Each participant was randomized to one of two BP goals—either a mean arterial pressure (MAP) of 102 to 107 mm Hg (usual; n = 554) or a MAP less than 92 mm Hg (lower; n = 540)—and to one of three initial drug treatment groups—ramipril (2.5-10 mg/day; n = 436), metoprolol (50-200 mg/day; n = 441), or amlodipine (5-10 mg/day; n = 217). Additional open-label antihypertensives were added sequentially if the goal BP could not be met with the randomized drug. Outcome measures included GFR slope and a clinical composite outcome composed of GFR decline of at least 50% or by 25 mL/min/1.73 m² or more from baseline, ESRD, or death.⁵ 


Through the 4-year follow-up period, there was no significant difference in GFR slope or the rate of clinical composite outcomes between the lower BP goal and the usual BP goal groups. No drug group showed a consistent significant difference in GFR slope, though the amlodipine group showed an initial increase in GFR during the acute phase (first 3 months of follow-up) that was ultimately negated by a more rapid decline in GFR during the 4-year follow-up. The ramipril group demonstrated a significant reduction in all clinical composite outcomes compared to the metoprolol and amlodipine groups, with a risk reduction of 22% (95% confidence interval [CI], 1%-38%; P = .04) and 38% (CI, 14%-56%; P = .004), respectively. The metoprolol group had a significantly lower risk for ESRD alone (59% risk reduction; CI, 36%-74%; P ≤ .001) and for ESRD or death (42% risk reduction; CI, 17%-60%; P = .003) compared to the amlodipine group but failed to show a significant difference in the clinical composite outcome (20% risk reduction; CI, -10%-41%; P = .17)⁵ (Table 1).


The average age and BMI of the AASK study participants was 54.6 years and 30.6 kg/m², respectively.⁵ The majority of participants were male (61.2%).⁵ These characteristic should be considered when attempting to apply the results to specific patients. The study was well-designed as a high-powered, double-blind with respect to drug treatment, randomized controlled trial that examined clinically significant outcomes using intent-to-treat analysis. Eligibility criteria were clearly defined, and the baseline characteristics of the treatment groups were comparable. A 1b level of evidence is appropriate for this trial, allowing clinical decisions based on the results to be made with a high degree of certainty. 


Appel and colleagues conducted a posttrial cohort study from 2002 to 2007 in which they followed 759 participants from the AASK trial who 
had not reached ESRD or death⁶ (Table 2). Although only 647 participants actually enrolled in the cohort study, the authors favored a conservative statistical approach and used n = 759 to account for adverse events that occurred during the transition period. The mean age of enrollees was younger than those not enrolled (mean age at the end of the trial 62.8 years versus 59.7 years in study; P = .009), but all remaining demographics and clinical characteristics (GFR, serum creatinine level, protein­uria) were not significantly different. Participants were switched from randomized drugs to ramipril (maximum 10 mg/day) or to an angiotensin-receptor blocker (ARB) if ramipril was not tolerated. In keeping with results from the AASK trial, the initial BP goal was less than 140/90 mm Hg until new national guidelines released in 2003 reduced the goal to less than 130/80 mm Hg. If the BP goal could not be met with ACE inhibitor or ARB therapy, additional antihypertensive drugs were added. The primary outcome was a clinical composite defined by doubling of the serum creatinine level from AASK trial baseline, ESRD, or death. A secondary outcome was the mean annual change in the eGFR.⁶


Kaplan-Meier curves were used to determine the cumulative probability of study outcomes. A competing risk formulation was then used to construct cumulative incidence curves to determine the accretive probability of study outcomes occurring before a renal event without censoring the renal events or death. Analysis of the secondary outcome was achieved by using two separate mixed-effects models to display the mean eGFR slope of the original 1,094 trial participants and the cohort study enrollees, regardless of whether the participant remained at risk for the composite outcome.⁶