Effects of Losartan and Amlodipine on Urinary Albumin Excretion and Ambulatory Blood Pressure in Hypertensive Type 2 Diabetic Patients With Overt Nephropathy

Hypertensive patients with type 2 diabetes and nephropathy between 31 and 80 years of age were enrolled in the study. The diagnostic criteria of diabetes were serum glucose levels >7.0 mmol/l at fasting and ≥11.1 mmol/l in a spot sample or 2 h after a 75-g glucose load. Overt diabetic nephropathy was diagnosed if the patient had a history of diabetes and persistent macroalbuminuria with urinary albumin excretion >300 mg and <3,000 mg/day in at least two out of three consecutive visits. Patients were eligible for the study if they had a creatinine clearance of >30 ml · min⁻¹ · 1.73 m⁻² and an office systolic blood pressure of 140–200 mmHg and diastolic blood pressure of 90–110 mmHg on at least three separate occasions after an observation period of 2 weeks or longer without antihypertensive medication. Patients who had a history of cardiac, hematologic, or hepatic disease; hormone-induced hypertension; cerebral infarction or hemorrhage; primary glomerulonephritis; or other major diseases were excluded. Informed consent to participate in the study was obtained from each subject. The protocol was approved by the Human Subject Committee of Yokohama City University Hospital in 2001. The study period was from January 2002 to December 2003.

The study, which was designed as an open-label, parallel-prospective, randomized study, consisted of a 2- to 4-week observation period, a 12-week titration phase, and a maintenance phase for a maximum of 12 weeks, comparing the effects of losartan and amlodipine. Antihypertensive agents, if used, were discontinued before the observation period. All subjects underwent ambulatory blood pressure monitoring (ABPM) before the titration period. Patients were then randomly allocated to receive 25 mg/day losartan or 2.5 mg/day amlodipine once a day, titrated to 100 or 10 mg at 2- to 4-week intervals (stepwise increase in increments of 25 or 2.5 mg, respectively) based on office blood pressure. Office blood pressure was estimated by a mercury sphygmomanometer every 2–4 weeks, with the subject seated after resting for 5 min. When patients met the therapeutic goal of systolic blood pressure ≤140 mmHg and diastolic blood pressure ≤90 mmHg or an absolute decrease in systolic blood pressure ≥20 mmHg and diastolic blood pressure ≥10 mmHg, as determined from the mean of two measurements while taking the same dose, they continued the same dosage until they underwent the second ABPM. Patients whose diastolic blood pressure was >110 mmHg or systolic blood pressure was >200 mmHg while taking the maximum doses of losartan or amlodipine or those who had unacceptable adverse reactions were withdrawn from the study. All patients, regardless of whether they met the therapeutic goal, underwent the second ABPM at the end of the titration phase or any time during the maintenance phase. The patients were asked to maintain a diet that contained no more than 5 g NaCl per day.

ABPM was performed using noninvasive multitasking blood pressure recorders (TM2425; A & D, Tokyo, Japan). The device monitors autonomic nervous function simultaneously while measuring blood pressure. The reliability of the devices has been evaluated (9). Blood pressure was measured by the oscillometric method using a cuff. Before the study, the calibration of all the machines was checked against simultaneously determined auscultatory readings obtained from a mercury sphygmomanometer. Blood pressure was measured twice for conditioning and thereafter recorded at 15-min intervals from 0700 to 2100 and at 30-min intervals from 2100 to 0700. The participants were asked to maintain their usual daily activities and to relax their arms during blood pressure measurement. They were also instructed to fill out a diary. Mean systolic and diastolic blood pressure and mean heart rates during the daytime and nighttime were calculated for each participant. ABPM was repeated in patients who had >20% missing values out of the expected number of readings, >30% error rate for the total readings, or missing values for >2 consecutive h. The following readings were omitted as technical artifacts: systolic blood pressure >250 mmHg or <70 mmHg, diastolic blood pressure >130 mmHg or <30 mmHg, pulse pressure >160 mmHg or <20 mmHg, or systolic differences >60 mmHg or diastolic differences >30 mmHg compared with the immediately preceding or successive values.

The method of power spectral analysis using the multitasking recorder has been described in detail elsewhere (9). Briefly, an electrocardiogram was obtained with a precordial lead to analyze the variation in RR interval length. Every 30 min, 512 RR intervals were recorded at a resolution of 7.8 ms and were analyzed after Fast Fourier transformation using appropriate software (A & D). The power spectral densities of the oscillations were divided into two ranges: a low frequency (LF) component in the range of 0.05–0.15 Hz and a high frequency (HF) component in the range of 0.15–0.4 Hz as an index of parasympathetic nervous activity. The LF-to-HF ratio was calculated as an index of the sympathovagal balance.

Blood samples were collected for hematological and biochemical examinations. Twenty-four–hour urine specimens were collected during the observation period and at the completion of treatment when the second ABPM was performed to measure 24-h urinary albumin excretion. The overnight urinary albumin-to-creatinine excretion ratio was measured in a first- morning urine specimen every 4 weeks during the study period. Urinary albumin concentration was determined by a turbidimetric immunoassay kit (Mitsubishi Chemical, Tokyo, Japan).

Based on previous studies (7,10,11), it was estimated that each therapy would induce a difference in systolic and diastolic blood pressure of ≥10 mmHg (7–10%), with an SD of at least 15 mmHg compared with control groups. Then after adjustment for possible withdrawals and noncompliance, the intention-to-treat effect was calculated to be a 5–10% risk reduction. A minimum sample size of 39 case subjects in each group was required to provide 80% power with an α type error of 5%. Skewed distribution of 24-h urinary albumin excretion and urinary albumin-to-creatinine ratio was normalized by logarithmic transformation before statistical analyses and expressed as geometric means with 95% CI. All other normally distributed parameters are presented as means ± SD. Differences between pre- and posttreatment were analyzed using the Wilcoxon’s signed-rank test. To assess the differences in baseline parameters between two groups, the nonparametric Mann-Whitney U test or Student’s t test and two-tailed P values were used, where appropriate. Overnight urinary albumin-to-creatinine ratio was analyzed by an ANOVA with repeated measurement and Dunnett’s test. For correlation studies, Pearson’s correlation coefficients were determined. Values of P < 0.05 were considered statistically significant.

From a total of 122 patients screened for the study, 108 patients met the inclusion criteria. Among these patients, 96 patients were enrolled in the study. The nonparticipants did not differ significantly from participating patients with respect to clinical characteristics. The 96 patients were randomized equally to amlodipine and losartan. Twenty-nine of 48 patients in the losartan group and 30 of 48 patients in the amlodipine group fulfilled the therapeutic goal. The proportion of patients who failed therapy did not differ between the two groups (P = 0.700). Four patients in the losartan group and five in the amlodipine group discontinued the study due to protocol infringements or noncompliance. Finally, the results from 87 patients were assembled for analysis. The two groups were comparable and did not differ in age, sex ratio, and BMI (Table 1). The final titrated doses were 71 ± 25 mg/day for losartan and 5.9 ± 2.2 mg/day for amlodipine. Twenty-eight patients who did not meet the target blood pressure received ABPM at the end of the titration phase at week 12, and 59 patients who met the target had it done between weeks 12 and 24. During treatment, 13 patients in the losartan group and 10 patients in the amlodipine group reported at least one drug-related or possibly drug-related adverse effect. These adverse effects were mild to moderate in severity and were generally tolerated or resolved with continued therapy. There was no adverse event that necessitated dose reduction, suspension, or discontinuation of treatment. No significant changes in hematological or biochemical parameters were detected during therapy.

The two groups did not differ significantly in baseline systolic and diastolic blood pressure at the office and during a 24-h period. Losartan significantly decreased office and 24-h systolic and diastolic blood pressure. Systolic blood pressure decreased more markedly during daytime than during nighttime, resulting in an elevation of the nighttime-to-daytime systolic blood pressure ratios (Table 2). In the amlodipine group, office and 24-h systolic and diastolic blood pressure also decreased during treatment. Both daytime and nocturnal blood pressure decreased to the same degree during treatment (Table 2), showing no differences between pre- and posttreatment nighttime-to-daytime blood pressure ratios.

The two groups did not differ significantly in baseline LF and HF components and LF-to-HF ratios during the 24-h period (Table 3). During losartan treatment, both LF and HF components showed no changes during daytime and nighttime, resulting in no alteration in the nighttime-to-daytime ratio for the LF and HF components. Thus, the LF-to-HF ratios did not differ between before and during losartan treatment. In the amlodipine group, the LF and HF components were unchanged during daytime and nighttime, as observed in the losartan group. Consequently, the LF-to-HF ratio for 24 h did not differ during amlodipine treatment. The nighttime-to-daytime ratio for the LF-to-HF ratio did not differ during treatment in both groups (Table 3), showing no changes in the diurnal sympathetic and parasympathetic nervous rhythm.

The 24-h urinary albumin excretion in the losartan group was 810 mg/day (95% CI 780–1,140) during the observation period and decreased significantly (P < 0.001) to 570 mg/day (510–910) at the completion of treatment when the patient underwent the second ABPM. The amlodipine group showed no difference (P = 0.893) in 24-h urinary albumin excretion during the observation period (790 mg/day [780–1,170]) and at the completion of treatment during the maintenance period (790 mg/day [710–1,260]). In the losartan group, the mean overnight urinary albumin-to-creatinine excretion ratio was 82 mg/mmol (79–112) during the observation period; the ratio decreased significantly (P = 0.015) at the 8th week of study (67 mg/mmol [60–108]) and remained low until the completion of study (P < 0.001) (67 mg/mmol [63–89]). The amlodipine group showed no significant changes (P > 0.897) in overnight urinary albumin-to-creatinine excretion ratio between the observation period (85 mg/mmol [81–121]) and throughout the treatment period (from 80 mg/mmol [75–112] to 90 [86–127]).

Absolute 24-h blood pressure reduction induced by both agents did not correlate with the changes in 24-h urinary albumin excretion (r = 0.12, P = 0.450 for systolic blood pressure and r = 0.02, P = 0.921 for diastolic blood pressure in the losartan group; r = 0.06, P = 0.697 for systolic blood pressure and r = 0.13, P = 0.395 for diastolic blood pressure in the amlodipine group). Absolute nocturnal blood pressure reduction induced by both agents also did not correlate with the changes in overnight urinary albumin-to-creatinine excretion ratio (r = 0.26, P = 0.079 for systolic blood pressure and r = 0.27, P = 0.076 for diastolic blood pressure in the losartan group; r = 0.12, P = 0.467 for systolic blood pressure and r = 0.03, P = 0.879 for diastolic blood pressure in the amlodipine group). There was no correlation between absolute nighttime changes in the LF-to-HF ratio and the changes in overnight urinary albumin-to-creatinine excretion ratio induced by both agents (r = 0.21, P = 0.168 in the losartan group; r = 0.15, P = 0.326 in the amlodipine group).

The principal finding in the present study is that both losartan and amlodipine decreased daytime and nighttime blood pressure to the same degree; however, only losartan exhibited an antiproteinuric effect. This finding indicates that in our patients with overt type 2 diabetic nephropathy, merely decreasing systemic blood pressure during daytime as well as nighttime does not reduce urinary albumin excretion.

In the majority of type 2 diabetic patients, elevated blood pressure is recognized at the time of diagnosis of diabetes or when diabetic nephropathy is first detected (12,13). This leads to the hypothesis that preceding hypertension contributes to the development of diabetic nephropathy, increasing urinary albumin excretion. There is increasing evidence that the circadian blood pressure pattern is more closely related to albuminuria than is casual blood pressure measured at an office (14,15). Several studies have reported that a rise in nocturnal systolic (16) or diastolic (10) blood pressure or both (4) is associated with microalbuminuria in type 2 diabetic patients with incipient nephropathy. Subsequently, the treatment effects of losartan provide proof for a correlation between fall in 24-h blood pressure and reduction of microalbuminuria (17,18). In hypertensive patients with advanced diabetic nephropathy on the other hand, few trials have been performed to evaluate the effect of angiotensin II receptor blockers on the 24-h blood pressure regulation profile and its relation to macroalbuminuria (19). Our results showed that losartan decreased 24-h blood pressure and exhibited an antiproteinuric effect. However, there was no significant correlation between losartan-induced decline in 24-h blood pressure and reduction in 24-h urinary albumin excretion. Absolute changes in blood pressure during nighttime also did not correlate with the changes in the overnight albumin-to-creatinine excretion ratio. Amlodipine also showed similar 24-h blood pressure reduction but had no effect on albumin excretion. We speculate that the antiproteinuric effect of losartan is attributed to mechanisms independent of the regulation of systemic 24-h blood pressure in overt diabetic nephropathy.

Another principal finding is that neither of the agents affected the autonomic nervous activity during the day and night. It is well known that impaired diurnal autonomic nervous rhythm and albuminuria are closely related to the blunted decline in nocturnal blood pressure in diabetic patients; however, the association between diabetic neuropathy and nephropathy still remains to be explored. Several studies (3,20) have suggested that diabetic neuropathy could have a primary pathogenic role in the development of diabetic nephropathy and that blunted diurnal rhythm of the autonomic nervous activity with higher blood pressure during nighttime would affect renal circulation, increasing nocturnal urine output and albuminuria. In the present findings, both agents decreased blood pressure during nighttime as well as daytime without affecting the 24-h autonomic nervous activity. However, only losartan decreased both 24-h urinary albumin excretion and overnight albumin excretion. Nevertheless, there was no correlation between the losartan-induced absolute reduction of overnight albumin-to-creatinine excretion ratio and the change in the nocturnal LF-to-HF ratio, an index of the sympathovagal balance. This indicates that the antiproteinuric action of losartan during nighttime as well as daytime is not related to change in the autonomic nervous activity in type 2 diabetic patients who have already developed macroalbuminuria.

At the initial stage of renal dysfunction when microalbuminuria is detected, systemic blood pressure regulation for 24 h by amlodipine effectively reduces urinary albumin excretion to restore renal function (21,22), suggesting that microalbuminuria might be a consequence of hypertension. In patients with overt nephropathy, however, marked albumin excretion is observed regardless of blood pressure levels, indicating that proteinuria is more likely to be relevant to glomerular lesion than increased systemic blood pressure. In these subjects, the transglomerular passage of albumin may be stimulated by changes in the properties of the glomerular transcapillary permeability rather than intraglomerular hemodynamic changes. Smith et al. (23) indicated that dihydropyridine calcium channel blockers have no direct effect on membrane permeability of albumin. Praga et al. (24) reported that amlodipine increases urinary transforming growth factor-β levels, which cause mesangial matrix synthesis and fibrosis. Thus, in patients with advanced diabetic nephropathy, it is suggested that dihydropyridine calcium channel blockers no longer have antiproteinuric effect, although they reduce 24-h blood pressure.

Meanwhile angiotensin II is one of the major regulators of intraglomerular hemodynamics and basement membrane permeability, directly or through the production of profibrogenic cytokines (25), and is involved in the remodeling of diabetic nephropathy. In animal studies using diabetic and obese rats, ACE inhibitors blocked angiotensin II formation to reduce proteinuria and structural glomerular damage (26,27) without changing intraglomerular capillary pressure (28). Therefore, we suspect that in diabetic patients with advanced renal disease, losartan may improve glomerular protein leakage primarily by competitively inhibiting the effects of angiotensin II on glomerular basement membrane permeability and mesangial cell and matrix proliferation. However, to prove this hypothesis, rigorous evaluation of the angiotensin II receptor blocker effects on the intraglomerular mechanism in a large scale study is required.

In conclusion, although both losartan and amlodipine decrease daytime and nighttime blood pressure equally without affecting the autonomic nervous activity, only losartan decreases urinary albumin excretion in type 2 diabetes with macroalbuminuria. These findings suggest that 24-h blood pressure regulation alone is not enough to improve overt nephropathy, and additional beneficial effects of losartan on intraglomerular mechanism are crucial for an antiproteinuric effect.