Effective Antihypertensive Strategies for High-Risk Patients With Diabetic Nephropathy

Abstract

Aim

Clinical guidelines recommend blood pressure (BP) lowering and renin-angiotensin-aldosterone system inhibition to slow kidney disease progression in patients with diabetic nephropathy. This study’s purpose was to determine whether an antihypertensive regimen including a maximally dosed angiotensin-converting enzyme inhibitor could safely achieve target BP in indigent, predominantly minority patients with this disease.

Methods

We studied 81 hypertensive adults (52% Hispanic and 31% African American) with nephropathy attributed to type 1 or 2 diabetes during the run-in period of a randomized controlled trial. The subjects received lisinopril titrated to 80 mg daily and additional anti-hypertensives to target a systolic BP (SBP) lower than 130 mm Hg. Blood pressure and serum potassium level were measured weekly, and a 4-gram sodium diet was prescribed. The primary outcome variable was SBP change from screening to randomization. Success in achieving SBP goal, change in urine albumin–creatinine ratio, hyperkalemia (serum potassium ≥5.5 mmol/L) and hypotension (SBP < 100 mm Hg) were also analyzed.

Results

The median SBP decreased from 144 to 133 mm Hg (median change, −9.6%.) Fifty-eight (71%) achieved goal SBP during run-in. The median UACR decreased from 206.8 to 112.7 mg/mmol (median change, −42.7%). The UACR reduction correlated with SBP reduction. Seventeen subjects experienced hyperkalemia responsive to dietary/medical management. Two subjects experienced hypotension responsive to medication adjustments.

Conclusion

A regimen using a maximally dosed angiotensin-converting enzyme inhibitor is safe and effective for achieving BP goal in high-risk, predominantly minority patients with diabetic nephropathy. Implementing this regimen necessitates close monitoring of serum potassium level.

Diabetic nephropathy accounts for nearly 50% of new cases of end-stage renal disease (ESRD) in the United States.¹ The incidence and prevalence of ESRD are higher in certain minority populations including Hispanics and African Americans.² As kidney disease progresses in patients with type 1 or type 2 diabetes, most exhibit urine albumin-to-creatinine ratio (UACR) of 33.9 mg/mmol or higher.¹ Hypertension in patients with diabetic nephropathy is highly prevalent, is difficult to control, and contributes to renal disease progression and cardiovascular endpoints. Pharmacologic blood pressure (BP) lowering is associated with albuminuria reduction and kidney function preservation in patients with diabetic nephropathy, and renin-angiotensin-aldosterone system (RAAS) inhibition enhances these effects.^3–10 Lowering systolic BP (SBP) below 140 mm Hg is associated with improved renal outcomes in these patients.¹¹ Current guidelines recommend angiotensin-converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB) dose maximization or combining drugs from these classes together with lowering BP of less than 130/80 mm Hg.^1,12–14 The recommended BP targets stem from the results of retrospective analyses of randomized clinical trials using RAAS-blocking drugs in diabetic nephropathy.^11,15 In these trials, most subjects did not achieve the currently recommended BP target, and it remains uncertain how attainable these goals are.

In practice, maximizing the dose of ACEi or ARB can be challenging given adverse effects such as serum creatinine elevations, hyperkalemia, or cough. Subsequently, adequate BP control with minimal adverse effects in patients with diabetic nephropathy remains difficult. In addition, the efficacy of anti-hypertensive regimens in inner city minority populations with diabetes and nephropathy has not been carefully evaluated.¹⁶ The purpose of the present study was to analyze the BP effects of a maximally dosed ACEi-based antihypertensive regimen in 81 predominantly minority subjects with diabetic nephropathy during the run-in phase of a completed clinical trial and to test the feasibility of the recommended BP goals in this patient population using this regimen.

MATERIALS AND METHODS

Study Subjects

Subjects were enrolled in a single-center double blind, randomized, placebo-controlled trial designed to test the hypothesis that in patients with diabetes and a UACR of 33.9 mg/mmol or greater despite treatment with a maximally dosed ACEi, adding either an ARB or mineralocorticoid receptor antagonist reduces albuminuria by 30%.¹⁷ Subjects entered the run-in phase if they were adults with type 1 or 2 diabetes; had seated SBP higher than 130 mm Hg or treated SBP lower than 130 mm Hg with previously documented hypertension; and a UACR of 33.9 mg/mmol or greater while taking ACEi for 3 months or more (Fig. 1). Major exclusion criteria included serum creatinine level greater than 265 μmol/L in women and greater than 354 μmol/L in men; secondary cause of hypertension, serum potassium level greater than 5.5 mmol/L, hemoglobin A1c (HbA1c) more than 11%, stroke or myocardial infarction 1 year before randomization, coronary revascularization 6 months earlier, heart failure, and anticipated need for dialysis within 12 months. The subjects provided informed written consent, and the Institutional Review Board of the University of Texas Southwestern Medical Center Dallas approved the protocol.

FIGURE 1.

Study design. After completing the screening visit, eligible subjects entered the run-in period during which lisinopril was titrated to 80 mg daily and final eligibility was determined by demonstrating that 24-hour UACR was 300 mg/g or greater on 2 successive occasions. Subjects were then admitted to the clinical translational research center for final 24-hour urine, blood pressure measurements, and randomization.

Study Design

A screening visit including BP measurement; blood collection for serum chemistry, complete blood cell count, lipid panel, HbA1c; and a spot urine sample collection for measuring albumin and creatinine was conducted on 235 subjects, with 128 eligible for run-in.

At the first run-in visit, all subjects were administered lisinopril ranging from 20 to 80 mg daily. Subsequently, the dose was titrated to 80 mg daily in all subjects. Additional non-ACEi, non-ARB, non–calcium channel blocker, and non–mineralocorticoid receptor antagonist antihypertensive agents were added to achieve an SBP lower than 130 mm Hg in the sequence: thiazide or loop diuretic, β-blocker, α-blocker, central α-2 agonist, and vasodilator. The subjects were prescribed and educated by a research dietitian to adhere to a 4-g sodium, 0.8-mmol/kg potassium per day, and 0.8-g/kg protein diet. The subjects with persistent hyperkalemia greater than 5.5 mmol/L despite dietary and diuretic dose adjustments, regression of albuminuria to normoalbuminuria, or hypotension (SBP < 100 mm Hg) unresponsive to medication adjustment were ineligible for randomization. The subjects with persistent albuminuria after completing the run-in were randomized. This report focuses on the effects of this regimen on SBP and UACR in the 81 subjects who completed the run-in and were randomized.

Study Procedures

Blood Pressure Management

The run-in goals were achievement of SBP lower than 130 mm Hg and maximization of lisinopril before randomization. The subjects received lisinopril titrated to 80 mg daily. Calcium channel blockers were discontinued to obviate confounding effects on albuminuria for the trial. The subjects were seen weekly by coordinators for run-in visits to measure BP, review medical regimens, reinforce dietary restrictions, and measure serum creatinine and potassium levels. Nephrologists added other antihypertensives as necessary.

Blood Pressure Measurement

Blood pressure was measured by staff trained and certified to measure BP according to the American Heart Association.¹⁸ Seated measurements were made after a 5-minute rest and repeated 3 times at 1-minute intervals. The mean of the last 2 measurements was recorded and used to determine medication adjustments.

Data Acquisition and Entry

Trained clinical research personnel experienced in BP measurement, phlebotomy, and participant education performed the visits. Data were entered in an encrypted Microsoft SQL server database with an Access front-end by trained study personnel.

24-Hour Urine Collection

Two outpatient 24-hour urine samples were collected during run-in to establish final eligibility for randomization: UACR 33.9 mg/mmol or greater despite treatment with 80 mg of lisinopril daily. An additional 24-hour urine was collected at the randomization visit. Collection adequacy was based on the total creatinine in the sample, on age, and on sex.

Laboratory Measurements

Urine albumin was quantified by immunoprecipitation (DiaSorin, Stillwater, MN) in the Clinical Translational Research Center (CTRC) laboratory by published methods. The within-run coefficient of variation (CV) range was 1.7% to 2.8 % and the between-run CV range was 2.4% to 4.2%. Creatinine levels in serum and urine were measured using a CX3 autoanalyzer (modified Jaffe reaction). HbA1c, measured by affinity chromatography in the CTRC, had a within-run CV of 1.4% and a between-run CV of 2.0% in diabetic patients using this method. Serum and urine chemistries including serum urea nitrogen and serum sodium, potassium, chloride, total CO₂, aspartate amino-transferase, alanine aminotransferase, alkaline phosphatase, bilirubin and total protein and urine creatinine sodium, urea, and total protein were performed on a Beckman Model CX9 autoanalyzer of the UT Southwestern CTRC laboratory. After statistical analysis, all laboratory values were transformed to SI units from the originally recorded values for the purpose of reporting.

Safety

Visits during the run-in period occurred approximately once weekly. Medications were adjusted for excessively low BP or symptoms potentially related to hypotension. Serum creatinine and potassium levels were measured every visit. A serum potassium level of 5.0 mmol/L or greater was managed with dietary and/or medication adjustments. Hyperkalemia, defined as a serum potassium level of 5.5 mmol/L or greater, was additionally managed with oral sodium potassium polystyrene sulfonate (Kayexalate). Persistently hyperkalemic subjects were not randomized.

Data Analysis and Statistical Considerations

The primary outcome in this analysis was median SBP change from the screening visit to the randomization visit. Based on our prespecified SBP goal of 130 mm Hg, we compared those subjects who achieved this goal at randomization versus those who did not. We also evaluated those who achieved an SBP of 140 mm Hg or less versus those who achieved an SBP of 140 mm Hg or greater at randomization, an SBP level recommended for hypertensive patients without diabetes and a level known to reduce cardiovascular events in those with type 2 diabetes.¹⁹ The Wilcoxon rank sum test was used to compare characteristics between groups based on whether or not an SBP of 130 or 140 mm Hg was achieved at randomization.

A secondary analysis included the UACR change from the spot urine sample at screening and the 24-hour urine collection at randomization. The UACR and other continuous variables were compared between these visits with the Wilcoxon signed rank test. We chose the randomization visit for comparison to the screening visit because we had full ascertainment of both BP and 24-hour UACR in each subject.

We assessed relationships between change in BP, UACR, and other variables with Spearman rank correlation coefficients. The correlation between urine sodium and UACR change after adjustment for BP was assessed with partial Spearman rank correlations coefficients. Results are reported as median and 25th to 75th percentile unless otherwise indicated. Statistical analysis was performed with SAS version 9.2 (SAS Institute, Cary, NC).

RESULTS

Subject Demographics

Among the 235 screened subjects, 128 entered the run-in phase and 81 were eligible for randomization. Reasons for ineligibility to randomize included regression to micro or normoalbuminuria (n = 16), hyperkalemia (n = 6), low BP (n = 3), failure to meet inclusion criteria (n = 4), serious medical condition (n = 7), withdrawal of consent (n = 7), lost to follow-up (n = 1), noncompliance (n = 1), and study closure before randomization (n = 2). The study population of the 81 subjects who completed run-in was ethnically and racially diverse, with most subjects being Hispanic (52%) or African American (31%; Table 1). Forty-eight percent of the subjects were men. The run-in period ranged from 22 to 101 days (median, 35 days). The duration of diabetes ranged from 1 to 40 years, and 28.4% of the subjects had cardiovascular disease.

TABLE 1.

Patients’ Characteristics and Demographic Data

Characteristics (n = 81 Patients)	Median (25th–75th Percentile) or Percent
Age, yrs	52 (46–58)
BMI, kg/m2	31 (28–38)
Race/Sex, %
Male	48
African American	31
Hispanic	52
Non–Hispanic white	15
Cardiovascular History, %
Coronary artery disease	9.8
Myocardial infarction	4.9
Amputation	8.6
Congestive heart failure	6.2
Stroke	7.4
Any above comorbidity	28.4
Years with Diabetes Mellitus	15 (10–22)
Run–in Days	35 (29–46)
Screening Laboratories/Parameters
Serum creatinine, μmol/L	133 (88–186)
Serum urea nitrogen, mmol/L	10.4 (7.50–13.9)
Serum potassium, mmol/L	4.5 (4.2–5)
UACR, mg/mmol	206.8 (107–377.7)
Loge UACR	7.51 (6.85–8.13)
SBP, mm Hg	144 (135–158); mean 148
Diastolic blood pressure, mm Hg	80 (73–88)
Total cholesterol, mmol/L	5.04 (4.14–5.95)
Screening Medication Use, %
ACEi	84
Diuretics	65
β–blockers	41
Calcium channel blockers	6
α–Antagonists	4
Clonidine	4
Insulin	79
Metformin	16
PPAR	13.6
Statin	69
Sulfonylurea	19

BP During Run-In

Fifty-eight (72%) and 71 (88%) of the study subjects achieved an SBP lower than 130 mm Hg and lower than 140 mm Hg, respectively, during the run-in period. The median value for the lowest achieved SBP during the run-in period was 124 mm Hg (118 and 134 mm Hg). Twenty subjects exhibited an SBP increase during the run-in period. Among this group, 13 had an SBP lower than 130 mm Hg and 18 had an SBP lower than 140 mm Hg at a run-in visit.

BP at Randomization

The median SBP decreased significantly from 144 mm Hg (25th and 75th percentile, 135 and 158 mm Hg) at screening to 133 mm Hg (25th and 75th percentile, 123 and 144 mm Hg) at randomization. The median percent change was −9.6% (25th and 75th percentile, −15.8% and 0%). Thirty-four subjects (42%) achieved an SBP lower than 130 mm Hg at randomization (Table 2). Among this subgroup, SBP decreased from 142 mm Hg (25th and 75th percentile, 129 and 153 mm Hg) at screening to 122 mm Hg (25th and 75th percentile, 113 and 125 mm Hg) at randomization. The change in the SBP and diastolic BP (DBP) was −15% (25th and 75th percentile, −23% and −10%) and −14% (25th and 75th percentile, −20% and 0%), respectively. Among 47 subjects in whom the median SBP was 130 mm Hg or greater at randomization, the change in SBP was −3% (25th and 75th percentile, −10% and 4%), and the change in DBP was −4% (25th and 75th percentile, −13% and 4%). The changes in both the SBP and the DBP were significantly greater in those achieving the SBP goal of lower than 130 mm Hg compared with those who did not. There were no detectable differences between these subgroups in age, ethnicity, race, sex, body mass index, renal function, or 24-hour urine sodium excretion at the randomization visit.

TABLE 2.

Characteristics of Patients Based on Whether or Not Target SBP of 130 mm Hg Was Achieved

Variable*	< 130 mm Hg (n = 34)	≥130 mm Hg (n = 47)	P†
	Median (25th, 75th Percentile)	Median (25th, 75th Percentile)
Age, yrs	52.5 (42, 59)	52 (48, 58)	0.79
BMI, kg/m2	29 (27, 38)	32 (28, 39)	0.51
Years with diabetes	15 (11, 23)	15 (9, 22)	0.53
Run in days	37 (29, 46)	34 (29, 46)	0.74
Serum creatinine, μmol/L	133 (88, 194)	141 (97, 177)	0.72
Cholesterol, mmol/L	4.65 (3.90, 5.43)	5.53 (4.37, 6.52)	0.003
LDL cholesterol, mmol/L	2.30 (1.68, 3.15)	3.03 (2.20, 3.67)	0.03
Screen SBP, mm Hg	142 (129, 153)	148 (140, 162)	0.04
Final SBP, mm Hg	122 (113, 125)	143 (138, 151)	<0.0001
SBP, percent change	−14.7 (−23, −10)	−2.8 (−10.3, 4.3)	<0.0001
Screen DBP, mm Hg	80 (73, 86)	82 (73, 89)	0.35
Final DBP, mm Hg	68 (59, 73)	77 (71, 84)	<0.0001
DBP, percent change	−14 (−20, 0)	−3.7 (−13, 4)	0.01
Screen UACR, mg/mmol	156.3 (71.5, 338.2)	212.6 (113.9, 437.6)	0.07
Final UACR, mg/mmol	72.7 (45.1, 188.7)	117.4 (52.7, 254.2)	0.09
UACR percent change	−42.5 (−55.6, −15.9)	−43.7 (−63.5, −7.3)	0.74
Screen HbA1c, %	7.9 (6.5, 9.6)	8.1 (6.7, 8.9)	0.98
Post HbA1c, %	7.2 (6.2, 8.5)	7.6 (6.7, 8.8)	0.31
HbA1c percent change	−3.9 (−8.2, 1.3)	−1.4 (−8.3, 1.7)	0.31

^*The final measurements refer to the randomization visit.

^†Comparisons made with the Wilcoxon rank sum test.

Forty-six subjects (57%) achieved an SBP lower than 140 mm Hg at randomization. Among this subgroup, screening SBP decreased from 142 mm Hg (25th and 75th percentile, 129 and 155 mm Hg) to 124 mm Hg (25th and 75th percentile, 116 and 130 mm Hg). The change in SBP and DBP was −14% (25th and 75th percentile, −22% and −5%) and −10% (25th and 75th percentile, −19% and 0%). Among the 35 subjects with an SBP of 140 mm Hg or greater at randomization, the median SBP decreased from 153 mm Hg (25th and 75th percentile, 142 and 162 mm Hg) to 146 mm Hg (25th and 75th percentile, 142 and 153 mm Hg) at randomization. The change in SBP in this subgroup was −2% (25th and 75th percentile, −9% and 5%), and the DBP change was −3% (25th and 75th percentile, −13% and 4%). Comparing the 2 subgroups, the SBP change was significantly greater in those who achieved an SBP lower than 140 mm Hg (P < 0.0001). There were no detectable differences in these subgroups in age, ethnicity, race, sex, body mass index, renal function, or 24-hour urine sodium excretion at randomization.

Antihypertensive Use

All subjects were administered 80 mg of lisinopril once daily and diuretics and β-blockers were the second and third leading antihypertensive agents used per protocol (Table 3). At randomization, the median number of antihypertensives taken was 3 (range, 1–6).

TABLE 3.

Medication Regimen at Screening and Randomization

	Screening	Randomization
Antihypertensive Medication, n (%)
ACEi	68 (84.0)	81 (100)
Diuretic	53 (65.4)	73 (90.1)
β-Blocker	33 (40.7)	58 (71.6)
α-Blocker	3 (3.7)	24 (29.6)
Central adrenergic agonist	3 (3.7)	8 (9.9)
Vasodilator	1 (1.2)	1 (1.2)
ARB	15 (18.5)	—
CCB	5 (6.2)	—
Spironolactone	1 (1.2)	—
No. Concomitant non-ACEi
Antihypertensive Medications, n (%)
0	20 (24.7)	7 (8.6)
1	23 (28.4)	15 (18.5)
2	26 (32.1)	31 (38.3)
3	10 (12.4)	25 (30.9)
≥4	2 (2.5)	3 (3.7)

Urine Albumin–Creatinine Ratio

Median UACR decreased significantly from 206.8 mg/mmol (25th and 75th percentile, 107.0 and 384.5 mg/mmol) to 112.7 mg/mmol (25th and 75th percentile, 51.9 and 210.5 mg/mmol) from screening to randomization with a median change of −43% (25th and 75th percentile, −61% and −14%). The change in UACR correlated positively with changes in both SBP (r = 0.31, P = 0.006) and DBP (r = 0.26, P = 0.02). Overall, both SBP and UACR tended to decrease during medication titration; however, parallel decreases in BP and UACR were not observed in every instance. There was no correlation between UACR and HgbA1c at screening and randomization.

Results From the Randomized Trial

The complete results of the randomized trial have been published elsewhere.¹⁷ Briefly, the study showed that there was a 34% reduction in albuminuria for spironolactone compared with placebo (P = 0.0007) but only a 16.8% reduction for losartan versus placebo (P = 0.20). All subjects remained on lisinopril at 80 mg daily during this trial. The study also showed that the clinic and ambulatory BP decreased significantly in all 3 randomization groups, but there was no statistical significance between groups. The final mean (SD) clinic SBPs were 126.8 (15.9) mm Hg, 132.3 (21.7) mm Hg, and 121.7 (14.6) mm Hg for placebo, losartan, and spironolactone, but the changes from baseline to repeated measures were not statistically different. Further details on these results may be found in the original manuscript.¹⁷

24-Hour Urine Sodium

The 24-hour urine sodium excretion at randomization did not correlate with screening, final, or change in SBP (r = 0.02, P = 0.89; r = 0.14, P = 0.20; r = 0.18, P = 0.12). A significant positive correlation existed between the 24-hour urine sodium and both randomization DBP and change in DBP (r = 0.31, P = 0.005; r = 0.22, P = 0.049). The 24-hour urine sodium excretion did not correlate with screening or final UACR (r = −0.09, P = 0.41; r = 0.12, P = 0.30), but there was a significant correlation with percent change in UACR (r = 0.28, P = 0.01).

Safety

The mean serum potassium level during run-in (322 measurements in 81 subjects) was 4.56 mmol/L (range, 3.1–6.1 mmol/L). Seventeen subjects had at least 1 episode of hyperkalemia during run-in, and 5 had at least 1 serum potassium level of 6.0 mmol/L or greater. Sixteen subjects experienced a serum potassium level between 5.0 and 5.5 mmol/L. All hyperkalemic episodes responded to dietary and/or medical management. Thirteen episodes of asymptomatic hypotension including an SBP lower than 110 mm Hg (11 subjects) and lower than 100 mm Hg (2 subjects) occurred. Each episode resolved after adjustment of antihypertensive medication dose(s).

DISCUSSION

The principal new finding in this study is that combining a maximally dosed ACEi with diuretics, β-blockers, and α-blockers is safe and effective for reducing BP in a population of indigent patients with diabetic nephropathy. Our regimen achieved an SBP lower than 130 mm Hg in 72% of the subjects during run-in and in more than 40% of the subjects at the randomization. The subjects not achieving this goal at randomization were more hypertensive at screening than those achieving the goal but otherwise had similar baseline characteristics. Furthermore, 88% of the subjects during run-in and 57% at randomization reached an SBP lower than 140 mm Hg, the target SBP of patients in the general population and an SBP associated with improved renal outcomes in patients with diabetic nephropathy.¹¹ Importantly, we achieved these results in a population largely composed of Hispanics and African Americans, groups at high risk of development and progression of diabetic nephropathy but underrepresented in previous clinical trials studying diabetic nephropathy. Thus, our findings demonstrate the feasibility and safety of achieving recommended BP goals in such vulnerable populations. There were notably some subjects whose BP was controlled during run-in but was higher than 130 mm Hg at the randomization visit. Given that the run-in visits involved trained study nurses measuring blood pressure using standardized procedures and validated equipment, we feel that these run-in measurements legitimately reflect success of the regimen. During the randomized trial, BP continued to be well controlled in the placebo arm, further documenting the success of the regimen.

The optimal BP for preventing renal and CV outcomes in patients with diabetic nephropathy has not been established prospectively from a randomized clinical trial. The landmark trials documenting the efficacy of ARB and ACEi in patients with diabetic nephropathy used the primary end points of the composite of ESRD, doubling serum creatinine, or death. The benefit of these drugs was attributed to effects of RAAS inhibition and considered to be independent of BP lowering. After reports of these results, retrospective analyses of these trials have provided information about the relevance of specific BP targets in patients with diabetic nephropathy concerning CV, renal, and mortality outcomes. In the Irbesartan in Diabetic Nephropathy Trial, the lowest quartiles for baseline SBP (< 145 mm Hg) and follow-up SBP (< 134 mm Hg) were associated with the fewest renal endpoints.¹⁵ Only 30% of the subjects achieved the SBP goal of 135 mm Hg in this study. In retrospective analysis of the Reduction in Endpoints in NIDDM with the Angiotensin Antagonist Losartan (RENAAL) trial,¹¹ using an SBP lower than 130 mm Hg as a reference, a baseline SBP higher than 160 mm Hg and a follow-up SBP higher than 140 mm Hg had higher hazard ratios for the primary outcome (composite of doubling serum creatinine, death, and ESRD). An analysis of the relationship between SBP and individual end points was not provided. These studies add to the prior observational evidence that lower SBP slows progression of diabetic nephropathy but uses the availability of data in reference to specific BP targets.

Despite specific recommendations existing for BP control, previous studies have documented how challenging these goals are in clinical practice. The National Health and Nutrition Examination Survey (NHANES) 1999–2000 showed that in a group of subjects with diabetes mellitus, 51.4% had hypertension, 85.2% were on an antihypertensive, but only 35.8% had BP controlled to lower than 130/80 mm Hg.²⁰ Studies in patients with diabetic nephropathy have also demonstrated difficulties, particularly with meeting SBP goal, even with regular visits to specialized clinics.²¹ Many of the original studies documenting the efficacy of ARB and ACEis in diabetic nephropathy had a mean SBP higher than 140 mm Hg in all arms, although the targets were higher than the present goals.^4,7,8 In the Irbesartan in Diabetic Nephropathy Trial, the achieved BP after more than 2 years were reported as 140/77, 141/77, and 144/80 mm Hg for the irbesartan, amlodipine, and placebo groups. In RENAAL, the mean BPs for the losartan and placebo groups were 140/74 and 142/74 mm Hg.⁴ For the Irbesartan in Microalbuminuria (IRMA II) study, a study that investigated the use of irbesartan to prevent diabetic nephropathy in subjects with microalbuminuria, the study’s mean BPs were 141/83, 143/83, and 144/83 mm Hg for irbesartan 300 mg, irbesartan 150 mg, and placebo.⁸ A study using ACEi in patients with type 1 diabetes and albuminuria showed a mean achieved BP of 129/77 for captopril versus 137/84 for placebo, but by design, this study excluded subjects with hypertension.⁹ The Diabetics Exposed to Telmisartan and Enalapril (DETAIL) study demonstrated noninferiority between ACEi and ARB in patients with type 2 diabetes and microalbuminuria using glomerular filtration rate decline as a primary outcome.²² The BP reductions were not significant between groups, suggesting that there should be similar expectations for BP control with both ACEi and ARB. Finally, one study in type 2 diabetic patients with microalbuminuria compared irbesartan at dosages of 300 mg/d with 600 mg or 900 mg using urine albumin excretion as the primary end point.²³ It was demonstrated that there was no significant difference between the dosage of the ARB and the final SBP. Systolic BP was lowered in all groups (mean, 140 mm Hg at baseline, with reductions of 8, 9, and 9mm Hg), but this study did not include subjects with overt albuminuria and did not report the percentage of subjects with an SBP lower than 130 mm Hg.

Most of these studies also had small proportions of African American and Hispanic subjects, making it difficult to make generalizations about the results. Our study used an aggressive multidrug regimen including a maximally dosed ACEi and resulted in more subjects achieving the goal. Importantly, 80mg/d of lisinopril alone was insufficient to achieve optimal BP control or normalize albuminuria. This finding is consistent with prior studies in which multiple antihypertensive agents were necessary to achieve an SBP lower than 140 mm Hg in patients with diabetes and nephropathy.^7,11

Lowering BP with an ACEi was associated with significant reductions in albuminuria in our study. Previous studies show that both SBP and urine albumin excretion predict renal and cardiovascular outcomes in patients with diabetic nephropathy. For example, post hoc analysis from the RENAAL trial demonstrated that within each strata of SBP, lower levels of albuminuria were associated with less risk for adverse renal outcomes.²⁴ Post hoc analysis also noted discordance between an individual’s changes in SBP and albuminuria. Our findings mirror those in RENAAL, and we found a significant correlation with SBP lowering and albuminuria reduction. However, we also observed decreases in SBP accompanied by increases in UACR in some subjects. Further studies are needed to determine why some subjects respond more favorably to RAAS inhibition and BP lowering than others.

We also examined the safety of our regimen. Hyperkalemia (potassium ≥5.5 mmol/L) was successfully treated with medical management and dietary adjustments. Potassium was checked routinely during medication titration, and we recommended this in the outpatient setting to ensure patient safety. The published National Kidney Foundation guidelines recommend that monitoring of serum potassium levels in patients with chronic kidney disease after initiation or uptitration of ACEi or ARB in the outpatient setting be based on the baseline serum potassium level.²⁵ This frequency is recommended to be less than 2 weeks for subjects considered to be high risk. Thus, our study procedures were in accord with what is considered standard of care in the ambulatory setting for this patient population. Some subjects were excluded from run-in because of baseline hyperkalemia, emphasizing that RAAS inhibitors at any dose must be administered with strict dietary education in patients with renal impairment.

Dietary sodium intake has important effects on UACR. Higher salt intake generally increases UACR, and lower salt intake generally decreases UACR.²⁶ In addition, higher sodium intake antagonizes ACEi’s antiproteinuric effects.²⁷ Our study did not show that final 24-hour urine sodium correlated with the final BP or UACR, eliminating dietary sodium intake as a potential confounder. There was no relation between glycemic control and UACR, which has previously been suggested to exist.²⁸ The 24-hour urine collections (obtained twice during run-in) also allowed us to monitor dietary sodium and potassium intake. Based on the results of these studies, the subjects were further counseled by a dietitian to promote education and adherence to the appropriate diet.

Our study has several limitations. First, there was no control group without lisinopril as a form of RAAS inhibition in its medical regimen. However, the focus was to strictly report observations from the run-in of a randomized study. As BP targets and recommendations have stemmed from retrospective analyses of clinical trials, our interest lie in assessing the feasibility of achieving these goals with the current recommended therapy. Furthermore, we demonstrated these findings in a population of patients that is frequently underrepresented in clinical trials, but is the predominant population of our health care center. The incidence rate of ESRD is more than 3 times higher in African Americans than Caucasians and 1.5 times higher in Hispanics than non-Hispanics,²⁹ and it is important to confirm that African American and Hispanic patients with diabetic nephropathy will respond to maximal RAAS inhibition. Patients who were run in, but not randomized, were not included in this analysis because of incomplete ascertainment of final UACR or corresponding BP. Regardless of the circumstances leading to ineligibility, some subjects had reached the fully titrated lisinopril dosage of 80 mg daily, whereas others had not yet reached this dose. The 81 randomized subjects were all known to be taking 80 mg of lisinopril daily, and we had full ascertainment of their UACR and BP. Another limitation was the initial urine specimen came from a spot urine collection without measurement of urine sodium, so we were unable to determine how changes in urinary sodium excretion correlated with BP and UACR changes. Finally, we compared the first morning–voided urine albumin collected at screening with the 24-hour urine albumin, a limitation noted in other studies.³⁰ Still, we used current practice guidelines for screening for nephropathy using the first-morning UACR, and we recommend morning UACR for monitoring clinic patients.

In conclusion, we demonstrated that ACEi dose maximization combined with other antihypertensives safely and effectively lowered BP to levels associated with proven benefit in high-risk patients with diabetic nephropathy. Furthermore, greater SBP reduction was associated with greater UACR reduction. Guidelines clearly state that patients with diabetic nephropathy should receive some form of RAAS inhibition. Furthermore, increasing the dose of a single RAAS blocking agent or adding on another agent for dual RAAS blockade causes greater short-term albuminuria reduction,^23,31,32 but it is not clear if these interventions consistently achieve SBP targets demonstrated to provide long-term renoprotection. Further long-term studies investigating clinical end points such as mortality are required before the full utility and benefit of this strategy can be determined in diabetic nephropathy. Based on our results, efforts should be made to maximize ACEi and add further antihypertensives as necessary in all patients with diabetic nephropathy, realizing that the recommended goal BP of 130/80 is attainable.