Angiotensin-converting enzyme (ACE) inhibitors have been found to attenuate progression of nephropathy in both types of diabetes in hypertensive [9-12] and normotensive patients [13-15] with microalbuminuria. They were also found to lower urinary albumin excretion in normotensive and normoalbuminuric patients with type 1 diabetes [16]. The relation between albuminuria and later progression of nephropathy in these patients has not been established, possibly because of short follow-up periods. No data are available on the effect of early introduction of ACE inhibitors in normotensive and normoalbuminuric patients with type 2 diabetes mellitus.

We designed a randomized, double-blind, placebo-controlled trial of the effect of ACE inhibition on the course of nephropathy in 156 patients with type 2 diabetes. These patients had normal blood pressure and normal urinary albumin excretion at baseline.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

Potential candidates were identified through the computerized records of the central regional laboratory for the northern part of the greater Tel-Aviv area. Persons with hyperglycemia and normal urinalysis results were located through their family physicians. Consent was sought once eligibility was established. Inclusion criteria were age younger than 60 years; age 40 years or older at diagnosis; duration of diabetes mellitus less than 10 years with no clinical evidence of malignant, autoimmune, hepatic, cardiovascular, or renal disease; body mass index less than 30 kg/m²; normal blood pressure on at least two consecutive visits (systolic pressure 140 mm Hg and diastolic pressure 90 mm Hg; mean pressure 107 mm Hg); serum creatinine concentration of 123 µmol/L or less; and urinary albumin excretion of 30 mg/24 h or less. All baseline data were obtained twice during the run-in prerandomization period. Patients were eligible only if values within the predetermined range were found on both examinations. The average of the values was used as the baseline value.

A total of 255 patients who had type 2 diabetes according to World Health Organization criteria [17] and attended one of eight clinics in the greater Tel-Aviv area were found to be eligible and were contacted during 1990 and 1991. Of these patients, 214 gave informed consent to participate. Twenty patients were excluded during the observation period: Six had blood pressure values above normal, 5 had microalbuminuria, 3 had serum creatinine concentrations above the trial criterion, 1 patient developed unstable angina pectoris, and 5 withdrew consent.

Of the 194 patients included in the study, 102 were women and 92 were men (mean age ±SD, 54.9 ± 3.2 years [range, 37 to 59 years]). The known duration of diabetes was 0 to 9 years (mean duration, 5.75 ± 2.8 years). Patients were instructed to use the standard isocaloric diet recommended by the Israeli Diabetic Association, and 69 study patients used diet alone to control their hyperglycemia. Pharmacologic therapy for diabetes was insulin in 34 patients and oral hypoglycemic agents in 91 patients.

Protocol

The protocol was approved by the hospital review board. After a 2-month observation period, patients were randomly assigned in a double-blind manner to receive enalapril (Assia-Riezel Ltd., Ramat-Gan, Israel), 10 mg/d, or placebo. Ninety-seven patients were assigned to receive enalapril, and 97 were assigned to receive placebo. Randomization was done centrally by telephone with a random number table [18]. Patient allocation to placebo or enalapril was recorded and kept by one of the authors. The placebo tablets were similar in appearance to the enalapril tablets. The medications, which came in sealed, numbered packages, were centrally prepared and were given to the patients at each visit by one nurse who was otherwise not involved in the study.

Patients were followed by their family physicians, who were unaware of allocation. Two semiannual prescheduled visits took place each year, and interim visits were scheduled as clinically indicated. At the semiannual visits, hemoglobin A_1c values, serum creatinine concentrations, serum electrolyte levels, and 24-hour albumin excretion and urinary creatinine concentrations were measured. Blood pressure was measured by the physicians twice at each visit by using mercury sphygmomanometers with the patients seated after a 5-minute rest; physicians recorded the average of the two values. The diastolic pressure was determined at Korotkoff phase V. If a systolic blood pressure of 145 mm Hg or more or a diastolic blood pressure of 95 mm Hg or more was found, measurements were repeated weekly. If elevated values persisted on two consecutive visits, a long-acting calcium-channel blocker (diltiazem or verapamil), hydrochlorothiazide, or both were administered. If systolic blood pressure values of 100 mm Hg or less were repeatedly found, the enalapril dosage was reduced to 5 mg/d (half of a 10-mg enalapril tablet or half of a placebo tablet). Fundoscopy was done yearly by an ophthalmologist, and the presence of retinopathy was recorded. For each patient, follow-up was terminated 6 years after randomization.

Measurements

All blood and urine samples were examined by a central laboratory. Assays were not changed during the study period. Glycosylated hemoglobin values were measured by affinity chromatography with a commercial kit (Isolab, Biochemical Methodology, Akron, Ohio). The normal range of this assay is a hemoglobin A_1c value of 3.5% to 5.6% and an intra-assay and interassay coefficient of variability of less than 3%. Urinary albumin concentration was measured twice in 24-hour urine samples by an automated immunoturbidimetric method [19]. This procedure has intra-assay and interassay coefficients of variability of 5.9% and 7.6%, respectively. Creatinine concentrations were determined by using the automated method of Bartels and colleagues [20]. Creatinine clearance, normalized for 1.73 m² of body surface area, was calculated for each visit by using the standard formula (urine creatinine x urine volume/plasma creatinine). The mean blood pressure (defined as the diastolic pressure plus one third of the pulse pressure) was calculated at each visit.

Statistical Analysis

Data are expressed as the mean (±SD) with ranges. A P value less than 0.05 was considered significant. On the basis of the assumptions that 15% of normotensive, normoalbuminuric patients with type 2 diabetes will develop microalbuminuria during 6 years and that treatment with enalapril will reduce the risk for microalbuminuria by 12%, we calculated that 69 patients were required in each group for a type 1 error of 0.05 and a power of 0.80 [21]. To test for adequate randomization and to compare the patients who completed the trial with those who did not complete the trial, the enalapril and placebo groups and the 38 patients who dropped out were compared for mean age; mean duration of diabetes; and mean baseline albumin excretion, creatinine clearance, glycosylated hemoglobin value, and blood pressure by using pooled-variance Student t-tests for independent groups and one-way analysis of variance. To compare the annual means of the various measurements between the two groups and within each group, one between-group factor and one repeated-measures factor were used in analysis of variance. For variables shown to be different by analysis of variance, unpaired t-tests were used for between-group parallel annual means and paired t-tests were used for comparison of intragroup sequential annual means. The rate of decrease of creatinine clearance and the rate of increase of albumin excretion were computed by doing linear regression analysis with all of the semiannual values included in the equation. Urinary albumin values were logarithmically transformed before analysis. The degree of albuminuria at baseline was used as a covariate.

The funding source had no involvement in the design, conduct, or reporting of the trial.

Results

Top Methods Results Discussion Author & Article Info References

Figure 1 shows the flow of participants during the trial. Thirty-eight patients did not complete the trial. Five patients died (3 in the enalapril group and 2 in the placebo group); death was related to coronary heart disease in 3 patients, stroke in 1 patient, and ovarian carcinoma in 1 patient. Six patients violated the protocol (2 patients in the enalapril group stopped taking their medication for more than 6 months, and 4 patients in the placebo group took an ACE inhibitor prescribed by consultant physicians for more than 6 months). Ten patients were lost to follow-up (6 in the enalapril group and 4 in the placebo group). The trial medication was discontinued in 12 patients: Six developed a disturbing cough (4 in the enalapril group and 2 in the placebo group), 4 had an allergic skin reaction (2 in the enalapril group and 2 in the placebo group), 1 patient in the enalapril group developed leukopenia, and 1 patient in the placebo group developed hyperkalemia. Finally, 5 patients developed severe urinary tract infections that had a detectable influence on renal function (2 in the enalapril group and 3 in the placebo group). A total of 156 patients completed the trial, of whom 77 received enalapril and 79 received placebo.

View larger version (22K): [in this window] [in a new window]   Figure 1. Flow of participants through the trial.

Baseline data for the two groups and for patients who did not complete the trial are shown in Table 1. The baseline characteristics of patients in the study groups and those who dropped out did not differ significantly. A modest but steady decrease in hemoglobin A_1c values was seen in the enalapril and the placebo groups and may reflect the change in attitude toward glucose control among family physicians in the early 1990s. However, the two study groups did not differ for this variable. In both groups, body mass index was marginally reduced during the first 2 years of the study; initial values were gradually regained during the subsequent period. Mean arterial blood pressure was successfully kept below 107 mm Hg in all patients throughout the study period. In total, 31 patients in the placebo group and 14 in the enalapril group required anti-hypertensive medications. The mean annual albumin excretion and creatinine clearance of these patients did not differ from their respective group means. The mean blood pressure in the enalapril group was significantly lower than that in the placebo group only during the fifth year of the study.

During the first 2 years of treatment, daily albumin excretion decreased in the enalapril group from an initial mean of 11.6 ± 7 mg/24 h to an annual mean of 10.3 mg/24 h and 9.7 mg/24 h for the first and second years, respectively. Thereafter, a small but steady increase in albumin excretion occurred in these patients to a sixth-year mean of 15.8 mg/24 h (P = 0.042). In 5 patients, albumin excretion exceeded 30 mg/24 h. In the placebo group, daily albumin excretion increased gradually and steadily from 10.8 mg/24 h to a sixth-year mean of 26.5 mg/24 h (P = 0.001). Values greater than 30 mg/24 h were found in 15 patients. The degree of albuminuria differed significantly between the enalapril group and the placebo group (P = 0.001). Transition to microalbuminuria at 6 years (the primary end point) occurred in 15 of 79 patients (19%) in the placebo group and 5 of 77 patients (6.5%) in the enalapril group. Therefore, enalapril treatment resulted in an absolute risk reduction of 12.5% (95% CI, 2% to 23%; P = 0.042) for crossing the threshold to microalbuminuria over a 6-year period. Eight normotensive patients with type 2 diabetes and normal albuminuria would need to take enalapril for 6 years to prevent the transition of 1 patient to microalbuminuria.

In the enalapril group, creatinine clearance decreased during the first year from an initial mean of 1.78 mL/s to a first-year mean of 1.73 (mean decrease, 0.05 ± 0.04; P = 0.025). Thereafter, creatinine clearance decreased at a rate of 0.017 mL/s per year; therefore, the mean annual decrease in creatinine clearance in this group was 0.025 mL/s. In the placebo group, creatinine clearance decreased at a fairly steady rate of 0.04 mL/s per year from an initial group mean of 1.82 mL/s to a sixth-year mean of 1.57 mL/s (P = 0.001). The decreases in creatinine clearance in the enalapril group and the placebo group differed significantly (P = 0.040). The follow-up data for mean blood pressure, albumin excretion, and creatinine clearance are given in Figure 2; the annual means of the main variables in the two groups during follow-up are listed in Table 2.

View larger version (20K): [in this window] [in a new window]   Figure 2. Creatinine clearance, albumin excretion, and mean blood pressure in normotensive, normoalbuminuric patients with type 2 diabetes who received enalapril (dashed lines) or placebo (solid lines). Each value represents the annual group mean based on two to three measurements per patient. The decrease in creatinine clearance was more pronounced in the placebo group (P = 0.047 for the fifth year and 0.042 for the sixth year). Patients who received enalapril had a lesser degree of albuminuria (P = 0.001).

In the placebo group, the baseline mean blood pressure and albumin excretion had positive predictive power for the rate of decrease of creatinine clearance during the follow-up period (for mean blood pressure, r = 0.47 and P = 0.024; for albumin excretion, r = 0.54 and P = 0.015). The only baseline variable associated with subsequent progression of albuminuria was baseline albumin excretion. In the enalapril group, baseline albumin excretion predicted both the initial decrease in creatinine clearance (for difference between baseline and lowest first-year values, r = 0.59; P = 0.015) and the change in this variable during the follow-up period (r = 0.46; P = 0.030). The degree of reduction in albuminuria induced by ACE inhibition, calculated as [(baseline value – second-year mean value ÷ baseline value) x 100], correlated inversely with the decrease in creatinine clearance during the study period (r = 0.51; P = 0.02). However, no correlation was seen between the reduction in albuminuria and the initial decrease in creatinine clearance that followed the introduction of enalapril therapy (r = –0.12;P = 0.16).

In the placebo group, univariate analysis showed a significant correlation between the rate of decrease in creatinine clearance and mean blood pressure (r = 0.53; P = 0.016), mean albumin excretion (r = 0.57; P = 0.008), mean plasma total cholesterol level (r = 0.58; P = 0.006), and hemoglobin A_1c values (r = 0.46; P = 0.024). In the enalapril group, a significant correlation was found between the decrease in creatinine clearance and mean blood pressure (r = 0.41; P = 0.05), mean total plasma cholesterol level (r = 0.42; P = 0.046), and mean albumin excretion (r = 0.42; P = 0.044), but creatinine clearance was not correlated with hemoglobin A_1c values (r = 0.22; P = 0.16).

Initially, background retinopathy was recorded in five and six patients in the placebo and enalapril groups, respectively. Annual fundoscopy revealed 15 new cases of retinopathy in the placebo group (19%) and 6 new cases in the enalapril group (7.8%). There was an absolute risk reduction of 11.2 percentage points (CI, 1 to 22 percentage points) for development of retinopathy over 6 years in the enalapril group (P = 0.047). Proliferative retinopathy developed in only 3 patients (1 in the placebo group and 2 in the enalapril group), all of whom were male smokers.

Discussion

Top Methods Results Discussion Author & Article Info References

Our 6-year, double-blind, prospective study shows the existence of a modest renal protective effect of enalapril in normotensive, normoalbuminuric patients with type 2 diabetes. The mean annual decrease in creatinine clearance was 0.04 mL/s in the placebo group compared with 0.025 mL/s in the enalapril group. This result has two components: the initial decrease of 0.05 mL/s at the start of enalapril treatment and a subsequent annual decrease of less than 0.017 mL/s, as is expected for a healthy population 40 to 60 years of age. The difference between the mean group values for creatinine clearance in the enalapril and the placebo groups became statistically significant only for the fifth and sixth years. A substantial decrease in glomerular filtration rate at the start of antihypertensive therapy was described in hypertensive patients with type 1 and type 2 diabetes and nephropathy [8, 10, 22]. We did not observe this phenomenon in normotensive patients with type 2 diabetes and microalbuminuria [23], possibly because the method used to assess kidney function was not sufficiently sensitive.

The well-known effect of ACE inhibition on the course of albuminuria was again shown in this study. The absolute difference between the enalapril and the placebo groups is obviously modest compared with those found in similar studies of patients with microalbuminuria [13, 23] or overt nephropathy [22]. However, the trend is identical, and the predictive power of baseline albumin excretion on the subsequent course of kidney function is also evident. Furthermore, as Rossing and colleagues demonstrated [24] in patients with type 1 diabetes, the initial reduction in albuminuria induced by ACE inhibitor predicted the extent of the subsequent renal protective effect.

In the placebo group, blood pressure, total cholesterol level, and hemoglobin A_1c value were associated with renal function. The correlation of cholesterol and renal function occurred in the enalapril group, but the correlation with hemoglobin A_1c value was not significant. The correlation between the decrease in creatinine clearance and the increase in albuminuria persisted in both groups. Financial reasons precluded the use of gold standard methods to assess renal function. However, in a long-term study of a fairly stable group of patients with initially normal or near-normal renal function, repeated measurement of creatinine clearance may provide a reasonable index of the changes in kidney function.

All patients maintained normal blood pressure throughout the study. However, the mean values in the enalapril group were somewhat lower than those in the placebo group; this difference reached statistical significance during the fifth year of the study. Therefore, our study does not show whether the renal protection induced by an ACE inhibitor is caused solely by effective control of blood pressure or involves additional specific changes in glomerular physiology independent of the control of blood pressure [25, 26].

Our patients were normotensive, had normal urinary albumin excretion, had fair control of glucose metabolism, and did not have severe hyperlipidemia. This combination of characteristics defines a low-risk category for the development of diabetic nephropathy [27-29]. Thus, the demonstration of a renal protective effect of ACE inhibition in these patients clearly bears on the potential usefulness of these agents for the preservation of renal function in type 2 diabetes. Our results should not be interpreted as advocating the introduction of ACE inhibitors for the prophylaxis of nephropathy upon diagnosis of diabetes. However, further study to identify low-risk patients who are likely to benefit most from this therapy is warranted. Several tentative methods have been suggested by recent studies, such as determination of urinary excretion of transferrin [30], examination of amiloride-sensitive sodium ion-hydrogen ion exchange in erythrocytes [31], or a therapeutic test with an ACE inhibitor and determination of the degree of decrease in albuminuria [24].

An important limitation of our study is that we were unable to perform an intention-to-treat analysis because we did not have complete follow-up data on the patients who left the study. Only patients for whom a complete set of data was available were included. The baseline data of the 38 patients who underwent randomization but did not complete the study did not differ from those of patients in the placebo group or the enalapril group who completed the study. Those 38 patients were also fairly symmetrically distributed at the start in the two groups.

Studies done mainly in type 1 diabetes have suggested an association between increased blood pressure and the onset and progression of diabetic retinopathy [32, 33]. This may suggest that reducing blood pressure will slow the course of retinopathy. The Eurodiab Controlled Trial of Lisinopril in Insulin-Dependent Diabetes Mellitus [16] reported a significant benefit of the ACE inhibitor lisinopril over placebo in normotensive patients with type 1 diabetes. The risk for progression of retinopathy was halved (12% with lisinopril and 25% with placebo) during 2 years of observation. In our study and a previous study on patients with type 2 diabetes and microalbuminuria, fewer new cases of retinopathy occurred in the enalapril group than in the placebo group. However, retinopathy was not a preplanned end point in either study, and no objective method was used to document retinopathy. These tentative observations should therefore serve only to encourage a prospective investigation into this matter. Also, the differences in retinopathy may indicate differences in the severity of diabetes between the two groups.

In conclusion, we found a modest but statistically significant renal protective effect of ACE inhibition in a low-risk group of normotensive, normoalbuminuric patients with type 2 diabetes mellitus. Long-term studies are needed to find out whether this effect will persist and retard the development of overt diabetic nephropathy.

Dr. Brosh: Department of Cardiology, Soraski Medical Center, Tel-Aviv, Israel.

Drs. Levi, Bar-Dayan, and Rachmani: Department of Medicine, Meir Hospital, Kfar-Sava 44281, Israel.

Dr. D. Ravid: Department of Obstetrics and Gynecology, Meir Hospital, Kfar-Sava 44281, Israel.