Angiotensin-converting enzyme (ACE) inhibitors slow the rate of progression of kidney disease in experimental models of renal insufficiency [11-13]. These agents have also been shown to reduce microalbuminuria in patients with incipient diabetic nephropathy [14, 15] and overt proteinuria in patients with diabetic and nondiabetic renal disease [16-18]. Taken together, these studies are suggestive of a special beneficial effect of ACE inhibitors in patients with renal disease. However, a benefit from reducing proteinuria in terms of improving the outcome of renal disease or in terms of other systemic (nonrenal) effects has not been shown in humans.

One potential benefit of drug use aimed at reducing proteinuria might be the lessening of the lipid abnormalities associated with overt proteinuria. Studies in humans that have shown a reduction in proteinuria with ACE inhibitors did not focus on the effect of this intervention on the lipid profile [14-18]. We report that fosinopril, an ACE inhibitor with dual hepatic and renal elimination [19], had a prompt and sustained beneficial effect on both proteinuria and lipid profile abnormalities in patients with proteinuric renal disease enrolled in a placebo-controlled, double-blind trial of 12 weeks duration followed by an open-label trial of fosinopril of 6 months.

Methods

Top Methods Results Discussion Author & Article Info References

Twenty-eight patients with proteinuric renal disease who entered a double-blind study (fosinopril sodium compared with placebo) are reported. Two patients were withdrawn from the study during the double-blind phase; one was lost to follow-up, and the other had an increase in serum creatinine from 230 to 292 µmol/L. Sixteen of the 26 patients who completed the double-blind phase were subsequently enrolled in an extended trial of fosinopril sodium. All patients except one were men; 17 were black, 8 were white, and 1 was Asian. Their ages ranged from 28 to 70 years. Diabetic nephropathy was diagnosed when overt proteinuria was documented in patients who had type II diabetes and clinical features consistent with diabetic nephropathy. Kidney biopsies were not done in diabetic patients. In nondiabetic patients, the diagnosis of glomerular disease was determined by kidney biopsy except in five patients who declined the procedure. The cause of their renal diseases were as follows: type II diabetic nephropathy in 15 patients, focal glomerular sclerosis in 3, membranous nephropathy in 2, and mesangio-capillary glomerulonephritis in 1. The cause of the renal disease in the remaining five nondiabetic patients was not determined.

For a patient to be included in the study, his or her proteinuria had to exceed 2.0 g/d and serum creatinine concentration had to be 265 µmol/L or less. Of the 26 patients, 21 had nephrotic-range proteinuria, defined as protein excretion exceeding 3.0 g/d [20]. Exclusion criteria were 1) treatment with converting enzyme inhibitors within 4 weeks of the baseline phase; 2) ongoing therapy with medications such as corticosteroids or immunosuppressives or with medications that could influence protein excretion such as nonsteroidal anti-inflammatory agents; 3) recent history of serum potassium concentrations exceeding 5.5 mmol/L; 4) poorly controlled diabetes mellitus; 5) uncontrolled hypertension defined as a diastolic blood pressure of 115 mm Hg or higher; 6) congestive heart failure; 7) noncompliance with medications; and 8) failure to collect 24-hour urine specimens or to keep clinic appointments. All participants gave informed written consent to participate in the study according to a protocol approved by the institutional review boards of both hospitals.

Intervention

Before patients entered the study, all antihypertensive medications were withdrawn. All patients received one placebo tablet daily for 3 weeks (baseline phase). After the baseline phase was completed, patients were randomly assigned to either fosinopril (10 mg once every morning orally, n = 17) or matching placebo (one tablet every morning orally, n = 9) in a double-blind manner with a 2-to-1 allocation ratio for randomization. In both groups, verapamil slow-release (SR), 240 mg per day, was added after week 2 of the double-blind phase if diastolic blood pressure exceeded 95 mm Hg (nine patients in the fosinopril group and four patients in the placebo group).

After 8 weeks, the dose of the blinded medications was doubled if protein excretion had not decreased by at least 30% or if the patient's diastolic blood pressure exceeded 95 mm Hg. Of the 26 patients, the dose was doubled in 19 patients (12 in the fosinopril group and 7 in the placebo group). During the baseline phase and throughout the study, oral furosemide, at a dose ranging from 20 to 160 mg daily (but constant for each patient), was allowed if edema was pronounced or to help control blood pressure (eight patients in the fosinopril group and six patients in the placebo group). Four patients were receiving medications known to affect serum lipid levels: lovastatin (two patients), colestipol, and an estrogen preparation. Except when indicated, the data from these patients were excluded from analysis of the effect of fosinopril on serum lipids to avoid confounding variables.

When the double-blind phase was completed, all patients were given one placebo tablet daily for 6 weeks (washout phase). During this washout phase, oral verapamil SR, 240 mg per day, was permitted only for those patients who had received it during the double-blind phase. After completion of the 6-week washout period, verapamil SR was withdrawn from all patients. A 23-week open-label trial of oral fosinopril, 10 mg once a day, was then started. Beginning at week 7 of this phase, the dose of fosinopril was increased to 20 mg every morning in all patients. Oral furosemide was given at the same dose throughout the study in 7 of the 16 patients.

Dietary and Analytical Evaluations

A dietary history was obtained by a registered dietitian during the baseline phase, at week 8 of the double-blind phase, and at the end of the open-label phase. Total caloric intake and the percentage of the caloric intake composed of fat, protein, and carbohydrates were recorded. All patients were instructed not to change their usual diet during the study. Patients were asked to refrain from strenuous exercise (jogging, weight lifting, vigorous aerobics) on the days when they collected 24-hour urine specimens.

At each visit, the patient's blood pressure was measured with a standard sphygmomanometer. After the patient rested in a seated position for 10 minutes, the blood pressure and heart rate were measured three times at 5-minute intervals; the means of the three readings were recorded. Mean arterial pressure was calculated as the diastolic blood pressure plus one third of the difference between the systolic and diastolic blood pressure. Clinical and laboratory data were recorded at the entry visit and weeks 2 and 3 of the baseline phase; weeks 1, 2, 4, 6, 8, 10, and 12 of the double-blind phase; weeks 2 and 6 of the washout phase; and weeks 1, 3, 7, 11, 15, 19, and 23 of the open-label phase. Venous blood was drawn to determine a complete blood count, fasting blood glucose, electrolytes, creatinine, urea nitrogen, total serum protein and albumin, liver function enzymes, triglycerides, and serum total cholesterol. A 24-hour urine sample was collected to measure total protein, albumin, creatinine, urea nitrogen, and sodium excretion. Urinary creatinine was measured to assess both creatinine clearance and the adequacy of the collection. Urine urea nitrogen and sodium were used as markers of protein and sodium intake, respectively.

Serum total cholesterol and high-density lipoprotein (HDL) cholesterol levels were measured as previously described [21]. Serum low-density lipoprotein (LDL) cholesterol was calculated as (total cholesterol -[HDL cholesterol + triglycerides/5]). This formula can be used only when triglycerides are less than 400 mg/dL. Two patients whose serum triglycerides exceeded 400 mg/dL were excluded from analysis of the effect of fosinopril on serum LDL cholesterol. Plasma lp(a) protein was measured using an assay based on "sandwich" enzyme-linked immunosorbent assay (ELISA) that is insensitive to the presence of plasminogen [22]. Serum ACE activity was measured 24 hours after the previous dose of fosinopril by the spectrophotometric method described by Cushman and Cheung [23].

Statistical Analysis

When paired analysis was used, data are reported as mean values followed by the mean change from baseline with the associated 95% confidence intervals (CI). Where appropriate, data are also reported as mean ± standard error of the mean. Two-way analysis of variance (ANOVA) was used to compare the change in protein excretion and total serum cholesterol from baseline over the course of the randomized trial between the two treatment groups.

Results

Top Methods Results Discussion Author & Article Info References

General Data

When considered as a group, our patients had mild to moderate renal impairment as judged by their elevated serum creatinine concentration (mean, 168 ± 9 µmol/L) and reduced creatinine clearance (0.98 ± 0.10 mL/s). Serum creatinine levels ranged from 80 to 265 µmol/L and creatinine clearance, from 0.38 to 2.22 mL/s. Of the 26 patients, 16 had hypertension, defined as systolic blood pressure of 150 mm Hg or diastolic blood pressure of 95 mm Hg or higher. Systolic blood pressure ranged from 122 to 190 mm Hg (mean, 151 ± 4 mm Hg), and diastolic blood pressure ranged from 74 to 114 mm Hg (mean, 92 ± 2 mm Hg). Protein excretion ranged from 2.2 to 15.7 g/d (mean, 5.4 ± 0.5 g/d). Serum albumin ranged from 23.0 to 45.0 g/L (mean, 37.2 ± 1.1 g/L). Three of the 26 patients had serum albumin levels below 32.0 g/L.

Serum total cholesterol was greater than 5.2 mmol/L in 22 of 26 patients (range, 4.1 to 7.8 mmol/L; mean, 6.5 ± 0.3 mmol/L). Plasma lp(a) protein levels ranged from 0.22 to 10.8 mg/dL (laboratory normal range, <7.00 mg/dL). Only three patients had frankly elevated levels of lp(a) protein (>7 mg/dL). Other laboratories reported much higher levels of lp(a) than those we report, which are based on measurements of lp(a) protein rather than total lp(a). A conversion factor of 4.2 is needed to convert lp(a) protein to total lp(a).

Randomized Trial

Renal Function and Blood Pressure

The general characteristics of the patients randomized to receive fosinopril (n = 17) and placebo (n = 9) are given in Table 1. Baseline measurements of blood pressure, creatinine clearance, and urine urea nitrogen and of total protein and albumin excretion did not differ statistically between the two groups.

At week 8 of this phase, total protein and albumin excretion decreased in the group treated with fosinopril (10 mg daily) but not in the placebo group (see Table 1). Creatinine clearance and urine urea nitrogen did not change significantly in either group. Serum ACE activity, measured 24 hours after the previous dose of fosinopril, decreased significantly in the fosinopril group but not in the placebo group. Systolic and diastolic blood pressure decreased in the fosinopril group but not in the placebo group. Protein excretion decreased significantly after 1 week of fosinopril therapy (Figure 1). At this time mean blood pressure decreased slightly in the fosinopril group [from 114 to 109 mm Hg, a decrease of 5.1 mm Hg (CI, –8.6 to –1.6 mm Hg)] and did not change in the placebo group (from 107 to 107 mm Hg, a change of –0.8 mm Hg [CI, –4.4 to 6.0 mm Hg]). After week 2, verapamil SR was given to patients whose diastolic blood pressure exceeded 95 mm Hg (four of the nine patients in the placebo group and 9 of the 17 patients in the fosinopril group). Because the proportion of patients in each group receiving verapamil SR was similar, this drug could not be directly implicated in the reduction in proteinuria that occurred in the fosinopril group.

View larger version (24K): [in this window] [in a new window]   Figure 1. Urine protein excretion. Upper panel. Changes in urine protein excretion throughout the randomized trial. By two-way ANOVA, the change in protein excretion from baseline differed statistically between the fosinopril group (\#9679;) and the placebo group ( (between-group, P < 0.01; within-group, P < 0.001). Middle panel. The effect of fosinopril on mean urine protein excretion, which was reversible when discontinued. Lower panel. The lack of significant effect of placebo on mean urine protein excretion throughout the study.

After 8 weeks in the study, the dose of fosinopril was increased from 10 to 20 mg every morning in 12 patients (see Methods). In these patients, no further reduction was observed in protein or albumin excretion (from 4.52 to 4.24 g/d, a decrease of 0.28 g/d [CI, –1.33 to 0.77 g/d] and from 2.85 to 2.93 g/d, a change of 0.08 g/d [CI, –0.80 to 0.96 g/d], respectively) between weeks 8 and 12 of the double-blind phase. Mean blood pressure decreased only slightly after the fosinopril dose was increased (from 111 to 108 mm Hg, a change of –3.6 mm Hg [CI, –7.6 to 0.5 mm Hg]). A further reduction in serum converting enzyme activity (from 11.1 to 6.0 µU/mL per min, a decrease of 4.9 µU/mL per min [CI, –7.6 to –2.3 µU/mL per min]) was documented in these 12 patients between weeks 8 and 12 of the double-blind phase.

When fosinopril was withdrawn (washout phase), protein excretion and albumin increased significantly back toward baseline values in the fosinopril group, whereas in the placebo group, protein and albumin excretion did not change significantly during this washout phase (see Figure 1). Throughout the randomized trial, the change in protein excretion from baseline differed statistically by two-way ANOVA between the fosinopril group and the placebo group (see Figure 1).

Lipids

Data on lipids are shown in Table 2 and Figure 2. During the baseline phase, serum total cholesterol and serum albumin as well as diet (total caloric intake and the relative percentages of dietary intake as fat, protein, and carbohydrate) were similar in the two groups (Table 2). No significant changes in any of the dietary variables were observed in either group.

View larger version (24K): [in this window] [in a new window]   Figure 2. Serum total cholesterol. Upper panel. Changes in serum total cholesterol throughout the double-blind phase. By two-way ANOVA, the change in serum total cholesterol from baseline differed statistically between the fosinopril group (\#9679;) and the placebo group ( (between-group, P < 0.05; within-group, P < 0.001). Middle panel. The effect of fosinopril on mean serum total cholesterol, which was reversible when discontinued. Lower panel. The lack of placebo effect on mean serum total cholesterol throughout the study.

In the fosinopril group, serum total cholesterol decreased statistically as early as week 4 of the double-blind phase, and this reduction was sustained throughout the double-blind phase (see Figure 2). During the washout phase, serum total cholesterol levels increased back toward baseline in the fosinopril group but not in the placebo group (see Figure 2). Throughout the randomized trial, the change in total serum cholesterol from baseline differed statistically by two-way ANOVA between the fosinopril group (15 patients) and the placebo group (7 patients) (see Figure 2). Serum total cholesterol was also reduced in the fosinopril group if all 17 patients (including 2 who were receiving lipid-lowering agents [see Methods]) were included in the analysis (from 6.5 to 6.0 mmol/L, a decrease of 0.51 mmol/L [CI, –0.91 to –0.12 mmol/L]). A positive correlation between the decrease in protein excretion and that in serum total cholesterol was found in the fosinopril group (r = 0.465, P < 0.001) but not in the placebo group (r = 0.032).

Plasma lp(a) protein had not changed significantly in either the placebo- or the fosinopril-treated group by week 8 of the double-blind phase. By the end of the double-blind phase (week 12) plasma lp(a) protein decreased in the fosinopril group (from 3.94 to 3.33 mg/dL, a reduction of 0.60 mg/dL [CI, –1.02 to –0.18 mg/dL]), whereas no significant change was seen in the placebo group (from 2.85 to 3.19 mg/dL, a change of 0.34 mg/dL [CI, –0.53 to 0.2 mg/dL]). Throughout the double-blind phase, however, no significant correlation was found between the decrease in protein excretion and the decrease in plasma lp(a) protein (r = 0.18).

Subset Analysis

In patients who received fosinopril alone (eight patients), protein excretion decreased from 6.10 to 3.57 g/d, a reduction of 32% (CI, 5% to 56%). In patients treated with fosinopril who had received additional therapy with verapamil SR (nine patients), protein excretion did not decrease to any greater extent (from 5.46 to 4.32 g/d, a reduction of 20% [CI, –22% to 60%]). In patients treated with fosinopril alone, mean blood pressure decreased from 106 to 93 mm Hg, a reduction of 13.3 mm Hg (CI, –19.8 to –7.3 mm Hg). In patients treated with both fosinopril and verapamil SR, blood pressure was higher than in the group treated with fosinopril alone and decreased from 121 to 111 mm Hg, a decrease of 10.1 mm Hg (CI, –20.8 to –0.3 mm Hg).

Because the presumed cause of the nephropathy in many of our patients was type II diabetes, we examined the effect of fosinopril on proteinuria in our 12 diabetic patients. In these patients, total protein excretion fell from 6.02 to 4.21 g/d, a reduction of 1.81 g/d (CI, –3.69 to –0.07 g/d), and albumin excretion decreased from 3.90 to 2.88 g/d, a reduction of 1.02 g/d (CI, –2.23 to –0.11 g/d). The reduction in proteinuria occurred without a significant change in serum glucose (from 9.7 to 9.1 mmol/L, a change of –0.59 mmol/L [CI, –2.83 to 1.64 mmol/L]). Plasma lp(a) protein decreased from 4.06 to 3.53 mg/dL, a change of –0.52 mg/dL (CI, –0.96 to –0.09 mmol/L), and triglycerides from 2.40 to 1.85 mmol/L, a change of –0.55 mmol/L (CI, –1.21 to –0.15 mmol/L). Total serum cholesterol decreased from 6.10 to 5.61 mmol/L, a change of –0.49 mmol/L (CI, –1.14 to 0.16 mmol/L), although this change did not achieve statistical significance.

The correlation found between the decrease in protein excretion and the decrease in serum total cholesterol prompted us to divide the patients treated with fosinopril into responders and nonresponders based on the change in protein excretion (more or less than a 30% reduction, respectively) observed at week 12 of the double-blind phase. At week 8 of this phase, the dose of fosinopril was increased to 20 mg once a day in five of nine responders and all nonresponders (see Methods). In the responders, the reduction in protein excretion ranged from –33% to –81%(mean change, –48%± 4%), and in the six nonresponders it ranged from a reduction of 26% to an increase of 127% (mean change, 11% ± 25%) (Table 3). Five of the nine responders and four of the six nonresponders were receiving verapamil SR.

Serum total cholesterol and LDL cholesterol decreased significantly in the responders but not in the nonresponders (Table 3). Lipoprotein(a) protein levels decreased in both subgroups, but the change achieved statistical significance only in the subgroup of responders (see Table 3). Systolic blood pressure decreased more in responders than in nonresponders, whereas diastolic blood pressure decreased about the same amount in both subgroups. Serum ACE activity decreased to about the same extent in both subgroups (see Table 3).

Open-label Phase

In this phase, the effect of fosinopril on protein excretion and plasma lipids was examined over an extended period (6 months), during which time other antihypertensive medications were not allowed (see Figure 3). In the 16 patients enrolled in this phase, protein excretion decreased from 4.53 to 3.83 g/d, a reduction of 15% or 0.70 g/d (CI, –1.39 to –0.01 g/d) after only 1 week of treatment with fosinopril, 10 mg daily. After 7 weeks on this dose, all patients were advanced to 20 mg orally per day until the end of the study. At the end of month 3, serum total cholesterol, LDL cholesterol, and plasma lp(a) protein levels decreased only marginally despite a significant reduction in urine albumin excretion (see Figure 3). At the end of the study (6 months), protein and albumin excretion decreased further to 29% (P = 0.02) and to 33% (P = 0.01), respectively (see Figure 3). These changes in protein and albumin excretion were accompanied by a decrease in serum total cholesterol (from 6.37 to 5.54 mmol/L, a change of –0.84 mmol/L [CI, –1.59 to –0.08 mmol/L]), LDL cholesterol (from 4.40 to 3.72 mmol/L, a change of –0.68 mmol/L [CI, –1.33 to –0.03 mmol/L]), and plasma lp(a) protein (from 3.58 to 2.81 mg/dL, a change of –0.82 mg/dL [CI, –1.58 to –0.05 mg/dL]) (see Figure 3). Triglycerides decreased but not significantly (from 2.61 to 2.24 mmol/L, a change of –0.37 mmol/L [CI, –1.18 to 0.45 mmol/L]), whereas HDL cholesterol did not change significantly (from 0.99 to 0.95 mmol/L, a change of –0.03 mmol/L [CI, –0.25 to 0.19 mmol/L]). Serum albumin did not increase significantly from 38.1 to 39.7 g/L, a change of 1.6 g/L (CI, –0.1 to 3.4 g/L).

View larger version (22K): [in this window] [in a new window]   Figure 3. The long-term effect of fosinopril (open-label phase) on urine albumin excretion, serum total cholesterol, serum low-density lipoprotein cholesterol, and plasma lipoprotein(a). The asterisk denotes a statistically significant change from washout (placebo) phase (P < 0.05).

During the open-label phase, a positive correlation between the decrease in protein excretion and in total serum cholesterol (r = 0.53, P < 0.005) and in the decrease in protein excretion and in plasma lp(a) protein (r = 0.64, P < 0.001) were found. Serum glucose did not change significantly during this phase of the study (from 9.3 to 7.3 mmol/L, a change of –2.0 mmol/L [CI, –5.5 to 1.5 mmol/L]).

Fosinopril treatment resulted in a reduction in systolic blood pressure from 142 to 133 mm Hg (a change of –9 mm Hg [CI, –15 to –3 mm Hg]) but no significant change in diastolic blood pressure (from 85 to 82 mm Hg, a change of –2.7 mm Hg [CI, –8.2 to 2.8 mm Hg]). The minimal effect of fosinopril on blood pressure may have reflected, in part, that initial blood pressure had been under the effect of verapamil SR, which was withdrawn in 6 of the 16 patients when fosinopril was started. Serum ACE activity decreased significantly (from 13.6 to 6.9 µU/mL per min, a change of –6.8 µU/mL per min [CI, –12.1 to –1.5 µU/mL per min]). No significant changes were detected in the glomerular filtration rate (from 0.78 to 0.83 mL/s, mean change of 0.05 mL/s [CI, –0.19 to 0.28 mL/s]), urine urea nitrogen excretion (from 389 to 412 mmol/d, a mean change of 25.0 mmol/d [CI, –35.7 to 85.7 mmol/d]), or urine sodium excretion (from 148 to 167 mmol/d, a change of 18 mmol/d [CI, –20 to 55 mmol/d]).

Discussion

Top Methods Results Discussion Author & Article Info References

Our results showed that serum total cholesterol and plasma lp(a) protein levels decreased significantly in patients with overt proteinuria treated with fosinopril but not in those treated with placebo. The finding of lowered plasma lp(a) protein is of particular interest because Ip(a) has atherogenic and thrombogenic potential and also because lowering of plasma lp(a) levels by drug therapy has been difficult to achieve [5]. Other antilipemic agents, including the widely used 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, have failed to reduce lp(a) [24] and, in fact, lp(a) may increase in as many as 30% of patients receiving these agents [25]. Our patients ingested a constant diet in terms of total caloric intake and percentage of diet composed of fat, protein, and carbohydrate. Therefore, the observed amelioration in lipid abnormalities could not be attributed to dietary changes.

The finding that a reduction in urinary protein and albumin excretion was accompanied by a reduction in serum total cholesterol extends to the clinical setting of recent experimental work, which has shown that serum cholesterol decreased in rats with experimentally induced nephrotic syndrome treated with enalapril [26]. Previous studies in patients given ACE inhibitors had shown a reduction in proteinuria, but a decrease in associated lipid abnormalities was not generally reported. Heavy proteinuria is usually accompanied by hyperlipidemia characterized by increased total and LDL cholesterol [8, 27]. In our patients, total cholesterol and LDL cholesterol were moderately increased and HDL cholesterol was reduced, a pattern of increased risk for atherosclerosis. An accelerated rate of atherosclerosis in patients with overt proteinuria may explain the increased incidence of ischemic heart disease and stroke observed in such patients [3]. Lowering serum lipids with lovastatin and other antilipemic agents has also been shown to slow the progression of renal disease in experimental animals [7-10]. Taken together, these observations emphasize the increasing interest in correcting lipid abnormalities associated with renal disease.

In two recent studies in humans, lovastatin was found to effectively lower cholesterol in patients with the nephrotic syndrome but, unsurprisingly, not to reduce proteinuria [28, 29]. The attractiveness of using ACE inhibitors for patients with renal disease stems from the fact that these agents not only may improve the lipid profile at least in proteinuric patients, as shown in this study, but are also effective in lowering blood pressure and protein excretion and possibly in retarding the progression of renal disease [11-13, 30]. We find it noteworthy that in our study serum total cholesterol could be reduced in the face of only a partial reduction in proteinuria. The lipid profile of patients whose protein excretion was not reduced was not significantly improved. The cholesterol and lp(a) protein-lowering effects of fosinopril (about 12% and 15% reduction from baseline, respectively) in responders can be viewed as an added attribute of potential clinical significance and reveals an objective benefit associated with lowering protein excretion. Our patients had moderate hyperlipidemia, whereas alterations in the lipid profile are often more severe in nephrotic patients [8, 26]. Recent reports showing high levels of lp(a) in nephrotic patients involved patients with more severe dyslipidemia than those we reported [4]. It should be noted that the "normal" range for lp(a) protein has not been definitively established [5], raising the possibility that the mean lp(a) protein level of our patients, although within the normal range, may have been higher than that of the normal population. We feel it is likely that greater reductions in serum total cholesterol and lp(a) protein can be achieved in individuals with the nephrotic syndrome associated with more severe dyslipidemia.

Our data also provide some insight into the mechanisms underlying the hyperlipidemia of the nephrotic syndrome. In hypoalbuminemic patients, hyperlipidemia may result from reduced oncotic pressure, and albumin infusions have been shown to lower serum cholesterol [31, 32]. Infusion of dextran, a non-albumin, oncotically active substance, also has been shown to reduce blood lipid concentrations in both analbuminemic rats and nephrotic rats [32]. These results suggested that the reduction in serum oncotic pressure, and not specifically the absence of albumin from the serum, causes hyperlipidemia [31, 32]. A strong inverse correlation between serum oncotic pressure and serum lipids has been reported in patients with the nephrotic syndrome [31]. Such correlations, however, may reflect, at least in part, a covariance with other causative factors as recently discussed by Keane and Kasiske [8]. Our patients were not frankly hypoalbuminemic, and serum cholesterol decreased at a time when serum albumin had not increased significantly (see Table 2). Finally, we found a positive correlation between the decrease in protein excretion and the decrease in both total serum cholesterol and plasma lp[a]. The latter, however, was found only during prolonged therapy with fosinopril, suggesting that the decrease in lp(a) protein requires more time to become apparent than does the change in serum total cholesterol. We cannot be sure whether the decrease in lp(a) protein is related to the reduction in proteinuria or to some other action of fosinopril.

The critical role of proteinuria in causing hyperlipidemia is shown by the recent study of Davies and colleagues in nephrotic rats [26]. These investigators showed that when proteinuria was reduced by enalapril, despite an unchanged rate of albumin synthesis, plasma albumin concentration increased and serum cholesterol decreased [26]. It seems reasonable, therefore, to conclude from our data, and by inference from the available experimental data, that the lipid-lowering effect of fosinopril is secondary to the reduction in proteinuria. A factor causing hyperlipidemia in nephrotic patients that could have been influenced favorably by reducing proteinuria is the attenuation of the loss of a liporegulatory substance in the urine. Such a substance has been implicated in facilitating removal of lipids from blood, a mechanism that may be impaired in the nephrotic syndrome [8, 33, 34].

The mean protein intake of our patients, estimated from urea nitrogen excretion, was in the normal range (about 80 g daily) and was maintained throughout the study as judged by dietary recall and 24-hour urea nitrogen excretion. Consequently, decreased dietary protein intake could not account for the observed decrease in protein excretion. In patients with renal disease, dietary protein restriction may retard disease progression but does not reduce serum lipids [35]. Previous studies using ACE inhibitors or calcium entry blockers to treat hypertension in patients with renal disease have not reported an antilipemic action, and the use of diuretics may be accompanied with an increase in serum total cholesterol.

Our study results also confirm previous reports of a reduction in proteinuria using ACE inhibitors [16-18]. Because verapamil SR was given to approximately the same proportion of patients in the placebo and fosinopril groups, this drug was probably not involved in the antiproteinuric and antilipemic effects observed. Moreover, a sustained antiproteinuric and lipid-lowering effect of fosinopril was documented over an extended period (open-label phase) when verapamil SR was not given to any of the study patients (see Figure 3). The reduction in systemic arterial blood pressure associated with fosinopril administration, although modest, clearly needs to be considered as a possible mechanism involved in the reduction of protein excretion. A role for a reduction in systemic blood pressure in the antiproteinuric effect of fosinopril, at least a permissive one, can be invoked on the basis of the observation that blood pressure decreased more in responders than in nonresponders (see Table 3). During the extended phase, however, fosinopril resulted in lowering of protein excretion even though systolic blood pressure declined only modestly whereas diastolic blood pressure did not change significantly. This observation is in contrast with the reduction in proteinuria that has been reported in previous studies using different classes of antihypertensive agents. In these studies, which used either ACE inhibitors [17, 36] or calcium entry blockers [36, 37], protein excretion was lowered in association with a substantial reduction in blood pressure.

The possibility that the underlying disease could be an important determinant in the proteinuric response and, in turn, the lipid-lowering response to ACE inhibitors deserves consideration. The patient sample we studied, however, is too small to form conclusions in this regard. It is nevertheless interesting that six of nine "responders" and five of six "nonresponders" were diabetic. This finding suggests that one cannot predict a priori whether individuals with type II diabetes would respond to ACE inhibitors by lowered proteinuria and serum lipids. As a group, however, our diabetic patients had a significant reduction in proteinuria and a decrease in serum triglyceride levels. Plasma lp(a) protein levels also decreased significantly in our diabetic patients. Glycemic control, assessed only by fasting blood glucose levels, remained essentially unchanged throughout the study phases, suggesting that the improvement in the lipid profile was related to fosinopril administration rather than to glycemic control. The choice of ACE inhibitors to treat hypertensive diabetic patients has attracted particular attention [38]. In this regard, our findings further suggest a beneficial effect of these agents in terms of proteinuria and lipid dysfunction for patients with type II diabetes who develop nephropathy and hyperlipidemia.

In summary, our data show that the ACE inhibitor, fosinopril sodium, causes a reduction in serum cholesterol and plasma lp(a) protein in association with a reduction in protein excretion. The finding that the lipid-lowering effect of fosinopril was preferentially observed in those patients who had a reduction in proteinuria suggests that attenuation of urinary protein loss is a mechanism involved in the antilipemic action. Our finding of amelioration of lipid abnormalities in the face of only partial reductions in proteinuria underscores the importance of therapeutic interventions geared toward the reduction of overt proteinuria.