Nephrol Dial Transplant (2002) 17: 1197–1203
Original Article
Proteinuria patterns and their association with subsequent end-stage renal disease in IgA nephropathy James V. Donadio1, Erik J. Bergstralh2, Joseph P. Grande3 and Diana M. Rademcher2 1
Division of Nephrology, Department of Internal Medicine, 2Section of Biostatistics and 3Division of Anatomic Pathology, Department of Laboratory Medicine and Pathology, Mayo Clinic and Foundation, Rochester, MN, USA
Abstract Background. Proteinuria (UP) )1.0 gu24 h at diagnosis is a well-known indicator of progressive renal disease in patients with IgA nephropathy (IgAN). To determine if persistent UP is a more sensitive marker for later progression of IgAN, the hypothesis was tested that the prior level and trend (slope) in UP for 1 year was better at predicting later end-stage renal disease (ESRD) (dialysis or transplant) than a current 24-h UP, serum creatinine (SC), SC slope, hypertension, or total glomerular histopathological score on index renal biopsy in an observational study of 154 high-risk patients enrolled in two clinical trials (IgAN 1, IgAN 2). Methods. Measurements of 24-h UP and SC were made at time 0, 6 weeks, 6 months and 1 year in all patients, who were then followed for an additional 5.76 years and 1.63 years in the two studies, respectively. The Cox proportional hazards model was used to identify predictors of ESRD following the 1-year visit. Results. Adjusting only for randomized treatment, nearly all UP variables (number of high readings, 1-year level, slopes), SC at 1 year, and SC trends (slopes) over the prior year were significantly associated with subsequent ESRD (all P values -0.05) in both studies. However, among the UP variables, the 1-year readings had the strongest association with ESRD in IgAN 1 (hazard ratio (HR), 95% CI, for a 1g increase: 1.5, 1.2,1.9), and the second strongest association (similar to UP trends) in IgAN 2 (1.4, 1.2,1.6). Males had lower rates of ESRD in both studies (IgAN 1 HR: 0.5, 0.2,1.2, Ps0.11; IgAN 2 HR: 0.2, 0.1,0.6, Ps0.002). In the multivariate analyses that examined all clinical and histological variables, 1-year levels of 24-h UP and SC, and female gender were independently associated with subsequent ESRD.
Correspondence and offprint requests to: James V. Donadio MD, Emeritus Staff, Mayo Clinic, 200 First Street South West, Stabile 722, Rochester, MN 55905–0001, USA. Email:
[email protected] #
Conclusion. In a high-risk patient with IgAN, the current 24-h UP and SC measurements are as good predictors of subsequent ESRD as UP and SC trends and levels over the prior year. Additionally, it appears that females have poorer outcomes than males. Keywords: IgA nephropathy; prognosis; proteinuria
Introduction At the time of renal biopsy diagnosis, proteinuria is a well-known indicator of progressive renal disease in patients with IgA nephropathy (IgAN). Among the risk factors examined in clinicopathological studies, increasing levels of urine protein (UP) )1.0 gu24 h is strongly associated with later development of renal failure [1]. Persistent proteinuria exceeding 1.0 gu24 h was an independent predictor of subsequent renal progression in a recently reported large series of patients with IgAN followed for 10 years or more [2]. In idiopathic membranous nephropathy, another glomerulopathy associated with high levels of UP, Pei et al. [3] showed that the highest sustained, 6-month period of proteinuria was an independent predictor of renal progression. This semiquantitative, predictive model was validated in an analysis of independent data from two other series of patients with membranous nephropathy [4]. In this study, it was also pointed out that, ideally, the defined time period of increased UP should be early enough in the clinical course of a patient to allow for a timely discussion between doctor and patient about prognosis, and for the physician to prescribe therapies that may slow progression of the renal disease. For example, the time from renal biopsy diagnosis to the start of the worst 6 months of proteinuria occurred between 1 and 2 years in 72–85% of the three groups of patients studied [4]. To determine if persistent proteinuria is an early, sensitive and independent marker for later progression of IgAN, the hypothesis was tested that the prior trend
2002 European Renal Association–European Dialysis and Transplant Association
1198
in UP for 1-year was better at predicting subsequent renal failure events than a current UP, serum creatinine, or other commonly used predictors (e.g. age, gender, blood pressure measurement of the day, or total glomerular histopathological score) in an observational study based on post-hoc analysis of two patient cohorts who participated in randomized clinical trials that tested the efficacy of omega-3 polyunsaturated fatty acids (v-3 PUFA) [5–7].
Methods Patients In this review, the first group (IgAN 1) of 106 patients participated in a randomized, placebo-controlled, doubleblind trial [5,6], and the second group (IgAN 2) of 73 patients took part in a randomized, open-label, v-3 PUFA 2-dose comparative design [7]. The study period for the first trial extended from June 1988 to December 1993, and for the second trial from September 1995 to January 2000. In both studies, patients were followed for a minimum of 2 years unless they developed a study end point or became noncompliant. Long-term, off-study follow-up data for serum creatinine levels and end-stage renal disease (ESRD) was collected for IgAN 1.
J. V. Donadio et al.
Time from the 1-year trial visits in IgAN 1 and IgAN 2 to the end point of ESRD, was estimated using the Kaplan– Meier method [9]. Due to differing follow-up, the percentage (Kaplan-Meier) of patients free of ESRD and the standard error was summarized at 6 years (after the 1-year visit) for the IgAN 1 trial, and at 2 years (also after the 1-year visit) for the IgAN 2 trial. The univariate association of factors (analysed as a linear trend in the log hazard rate for continuous factors) with time to ESRD was assessed using the Cox proportional hazards model [10]. Results based on the Cox model are presented using hazard ratios (HR) and 95% confidence intervals. Multivariate analyses were done using stepwise regression (Cox model) analysis in an attempt to identify which variables were most important in predicting subsequent ESRD. In the first analysis only the UP variables (adjusted for randomized treatment group) were included. This was followed by an analysis including all variables (adjusted for treatment) including serum creatinine variables (1-year measurement and slope), presence of hypertension at 1 year, and total glomerular histopathological score on the index renal biopsy. All renal biopsies were performed at various intervals preceding the trials, and the assumption was made that there was no change in the pathology when the biopsy scores were applied to the Cox model. In a previous study of 148 patients with IgAN, total glomerular score, derived from the extent of mesangial cell proliferation, matrix increase, capillary loop narrowing or disruption, glomerular sclerosis, cellular crescents and fibrous adhesions, was an independent predictor of adverse outcome [1]. All tests were 2-sided at alpha level 0.05.
Study design As the two trials had differing enrolment criteria regarding UP levels (IgAN 1, P1.0 gu24 h; IgAN 2, any amount), differing treatments (IgAN 1, v-3 PUFA vs placebo; IgAN 2, low vs high-dose v-3 PUFA), and differing lengths of follow-up beyond 1 year (IgAN 1, 5.76 year, median; IgAN 2, 1.63 year, median), we felt it prudent to perform separate analyses for each trial. In order to allow for a reasonable number of prior UP readings and an adequate number of subsequent renal events, all analyses were based on the patients who were alive at 1 year without prior ESRD. All patients in both trials had four prior UP and serum creatinine measurements taken at scheduled visits (baseline, 6 weeks, 6 month and 1 year). One or two (average taken if two) 24-h UP determinations were made at each scheduled visit. Proteinuria levels were measured with Pyrogallol Red Molybdate Reagent (Waco Pure Chemicals Industries Ltd, Osaka, Japan) using a technique adapted from the method described by Yoshimoto et al. [8]. Missing values were imputed using within-patient linear regression analysis for 15 UP readings (in 15 patients) out of a total of 616 readings.
Statistical analysis The focus of the data analysis was to address the predictive ability of a current (1-year trial visit) UP measurement relative both to changes in proteinuria over the past year, and to the number (0–4) of high ()500, )1,000, )1,500, )2,000, )2,500 or )3,000 mgu24 h) UP levels measured over the past year. Changes in UP and serum creatinine were based on slopes from within patient linear regression analyses using the four prior readings. The linearity assumption seemed tenable for the relatively short period of 1 year.
Results There were 91 patients in IgAN 1 (74% male) and 63 patients in IgAN 2 (83% male) who were alive after 1-year of follow-up and had not reached ESRD in their respective trials. There were 18 ESRD events after 1 year in IgAN 1, and 14 ESRD events after 1 year in IgAN 2. There was a wide distribution of 24-h UP levels and slopes over the prior year in both study cohorts with the most favourable changes, i.e. a reduction of 500 mgu24 h or greater (slope) occurring in the IgAN 1 group (Table 1). Sixty to ninety per cent of patients in both groups had had 75–100% (3 or 4) of their UP readings above the 500 and 1,000 mgu24 h levels during the previous year (Table 2). The percentages of patients having 3–4 readings declined with an increase in UP levels. Seventy-seven out of the 91 patients (85%) in IgAN 1, and 25u63 patients (40%) in IgAN 2 had serum creatinine -2 mgudl at 1 year. Slopes in serum creatinine were more positive, i.e. P0.30 mgudluyear, in more IgAN 2 patients, indicating worsening renal function, compared with IgAN 1 (Table 1). This reflects a better baseline renal function in IgAN 1 patients [5]. Due to the study design, eligibility serum creatinine entry levels in IgAN 1 had to be -3.0 mgudl [5], and in IgAN 2 eligible serum creatinine levels ranged from 1.5 to 4.9 mgudl [7]. The presence of uncontrolled hypertension at 1 year and total glomerular histopathological scores were similar in the two groups (Table 1).
Predictors of ESRD in IgA nephropathy
1199
Table 1. Urine protein, serum creatinine, hypertension and glomerular score in patients with a 1-year visit and no prior ESRD Study Variable
Urine protein (mgu24 h) 1-year O500 501–1500 1501–3000 )3000 Slope over prior year - 500 500 – 0 0 – 499 P500 Serum creatinine (mgudl) 1-year -2 2.0 –2.9 P3.0 Slope over prior year -0 0.0–0.29 P0.30 Hypertension at 1 yeara Yes No Glomerular score at biopsy 1– 4 5 –7 8 –11 Unknown
IgAN 1 (ns91) number (%)
IgAN 2 (ns63) number (%)
11 36 28 16
(12) (40) (31) (18)
17 16 14 16
(27) (25) (22) (25)
38 21 19 13
(42) (23) (21) (14)
12 22 13 16
(19) (35) (21) (25)
77 (85) 11 (12) 3 (3)
25 (40) 23 (36) 15 (24)
34 (37) 45 (49) 12 (13)
23 (36) 22 (35) 18 (29)
38 (42) 53 (58)
29 (46) 34 (54)
29 34 28 0
23 23 11 6
(32) (37) (31) (0)
(37) (37) (17) (10)
a Hypertension was defined as a systolic pressure of P140 mmHg or a diastolic pressure of P85 mmHg, regardless of therapy.
Table 2. Patients with 75–100% of UP above specified levels Study Variable Urine protein cut-off point (mgu24 h)
IgAN 1 (ns91) %a
IgAN 2 (ns63) %a
500 1000 1500 2000 2500 3000
90 67 47 30 24 12
79 60 38 29 24 19
95% CI, for a 1 g increase: 1.5, 1.2,1.9) and IgAN 2 (1.4, 1.2,1.6) (Table 3). Higher frequencies of readings above all pre-selected UP cut-off points were also significantly associated with later ESRD (Table 3). Serum creatinine measurements at 1 year (Figures 3A and B) and slopes over the prior 1 year (Figures 4A and B) were strongly associated with the subsequent development of ESRD after the 1-year visit in both study groups (Table 3). Hypertension and ACE inhibitor use at 1 year were not associated with later ESRD, and higher total glomerular scores were associated with ESRD events in IgAN 1 only. Six-year ESRD-free survival was 93, 75 and 62%, for glomerular scores of 1–4, 5–7 and 8–11, respectively (P-0.001). Males had lower rates of ESRD in both studies (IgAN 1 HR: 0.5, 0.2,1.2, Ps0.11; IgAN 2 HR: 0.2, 0.1,0.6, Ps0.002) (Table 3). At 2 years, ESRD-free survival was 84"6% for males and 46"17% for females in IgAN 2. In the multivariate analysis for IgAN 1, adjusting only for treatment, the strongest predictor (x2 to enter stepwise model) of subsequent ESRD was the serum creatinine value at 1 year, followed by the serum creatinine slope and the UP value at 1 year (Table 3). Among the UP variables, the only significant independent (stepwise regression) predictor of ESRD was the UP value at 1 year. Using all the variables listed in Table 3, serum creatinine concentration at 1 year, UP measurement at 1 year and gender (females at higher risk) were selected as independent predictors of ESRD (all three P-values -0.001). In IgAN 2, adjusting only for treatment, the strongest predictor of subsequent ESRD was the serum creatinine slope, followed by the serum creatinine concentration at 1 year, and the UP value at 1 year (Table 3). Among UP variables, the only significant independent (stepwise regression) predictor of ESRD was the UP value at 1 year, which is consistent with the findings in IgAN 1. Using all the variables listed in Table 3, serum creatinine slope (P-0.001), UP slope (Ps0.002) and female gender (Ps0.001) were selected as independent predictors of ESRD (model x2s49.7). A model including gender, UP at 1 year and serum creatinine at 1 year (the model selected for IgAN 1) had an almost identical predictive power (model x2s46.4). A pooled analysis combining IgAN 1 and IgAN 2 found similar results.
a
Tabled values are the percentage of patients with 75% (3u4) or 100% (4u4) of UP readings above the indicated levels during the previous year.
Angiotensin-converting enzyme (ACE) inhibitor use at 1 year was documented in 53% of IgAN 1 patients and 78% of IgAN 2 patients. Urine protein levels measured at 1 year (Figures 1A and B) and slopes over the prior 1 year (Figures 2A and B) were significantly associated with subsequent ESRD following the 1-year visit in both IgAN 1 (HR,
Discussion A current assessment of 24-h UP excretion and serum creatinine concentration gave either better or equivalent prognostic information over UP trends on the subsequent development of ESRD in separate analyses of two cohorts of patients with IgAN who were enrolled in prospective clinical trials that tested the efficacy of v-3 PUFA and involved long-term follow-up [5–7]. The one-time serum creatinine and UP
1200
J. V. Donadio et al.
Fig. 1. 24-h UP levels at 1 year were significantly associated with subsequent ESRD after 1 year on study in (A) IgAN 1 (P-0.001, linear trend) and (B) IgAN 2 (P-0.001, linear trend).
Fig. 2. 24-h UP trends (slopes) over the prior year were significantly associated with later ESRD after 1-year of study in (A) IgAN 1 (Ps0.007, linear trend) and (B) IgAN 2 (Ps0.01, linear trend).
values at 1-year study visits in both study groups were independently associated with subsequent renal failure outcomes. In multivariate analysis, females patients appeared to have poorer outcomes in both cohorts. Total glomerular histopathological scores of index renal biopsies were univariately significant for one cohort.
In the current study, we directly compare the relative value of a single 24-h UP determination vs trends in UP excretion over time in ascertaining the risk of renal disease progression. Although several recent studies have established that baseline proteinuria and serum creatinine are strong predictors of subsequent GFR decline and progression to ESRD [11], the predictive
Predictors of ESRD in IgA nephropathy
1201
2
Table 3. HR and x of individual predictors of ESRD after 1 year of study adjusted for treatment effects using the Cox model Study Variable
Number of UPs )500c Number of UPs )1000 Number of UPs )1500 Number of UPs )2000 Number of UPs )2500 Number of UPs )3000 UP slope (1guyear increase) UP value at 1 year (1g increase) SC slope (0.1 mgudluyear increase) SC value at 1 year (0.1 mgudl increase) Glomerular score at biopsy Hypertension at 1 year ACE therapy at 1 year Male gender Age at 1 year visit (10 year increase)
IgAN 1 (ns91)
IgAN 2 (ns63)
Chi-squarea
HR (95% CI)b
Chi-squarea
HR (95% CI)b
3.4 3.7 5.3 6.4 4.9 8.7 3.3 11.8 17.1 26.8 6.1 0.6 0.6 2.6 1.8
NA 1.8 (0.9–3.4) 1.6 (1.0–2.3) 1.5 (1.1–2.1) 1.4 (1.0–1.9) 1.6 (1.2–2.3) 1.4 (1.0–2.0) 1.5 (1.2–1.9) 1.4 (1.2–1.6) 1.2 (1.1–1.3) 1.3 (1.1–1.6) 0.7 (0.3–1.8) 0.7 (0.8–1.8) 0.5 (0.2–1.2) 0.8 (0.5–1.1.)
5.8 7.3 6.2 8.7 8.3 8.3 7.8 21.8 56.0 35.8 0.1 0.7 0.0 10.4 1.6
NA 2.3 (1.1–5.1) 1.6 (1.1–2.3) 1.6 (1.1–2.1) 1.5 (1.1–2.1) 1.5 (1.1–2.0) 1.5 (1.1–1.9) 1.4 (1.2–1.6) 1.3 (1.2–1.5) 1.1 (1.07–1.2) 1.0 (0.8–1.4) 1.6 (0.5–4.6) 1.0 (0.3–3.6) 0.2 (0.1–0.6) 0.8 (0.5–1.2)
a Score x2 statistics for addition of the variable to the Cox model containing only treatment. x2 values above 3.84 are significant at P-0.05. All variables were analysed as continuous predictors, except hypertension, ACE therapy and male gender; 0, no; 1, yes. b HR (95% confidence interval) for effect of indicated change (unit increase unless otherwise noted) in the variable on the subsequent risk of ESRD. NA, HR not estimable due to division by zero. c Number of UP readings (0–4) above the indicated cut-off point (mgu24 h) in the previous year.
Fig. 3. Serum creatinine concentrations at 1-year were significantly associated with subsequent ESRD in (A) IgAN 1 (P-0.001, linear trend) and (B) IgAN 2 (P-0.001, linear trend).
value of a single baseline UP excretion determination vs trends in proteinuria obtained by repeated measurements over time has not been previously assessed. In a high-risk patient with IgAN, this study shows that knowing a 1-year history of proteinuria, which is often not available for the clinician’s review, is no more useful a clinical predictor of subsequent adverse renal outcome than is having the clinical information of
proteinuria and renal function of a single day. These findings are based on prospective studies in which patients had four scheduled study visits over the first year of study enrolment. In practice, many patients will not have had this systematic evaluation of their clinical course. Thus, the practical value of the findings in the present study indicates that a current 24-h measurement of proteinuria and serum creatinine
1202
J. V. Donadio et al.
Fig. 4. Serum creatinine trends (slopes) over the prior year were significantly associated with subsequent ESRD in (A) IgAN 1 (P-0.001, linear trend) and (B) IgAN 2 (P-0.001, linear trend).
will provide the best information required to assess a clinical course and prescribe appropriate therapies. These findings are at variance with those of Kobayashi et al. [2] who reported that persistent proteinuria exceeding 1,000 mgu24 h was the most reliable, independent predictor of long-term prognosis in Japanese patients with IgAN followed for 10 years or more. Detailed information concerning the methodology and analysis of a number of clinical variables related to outcome is not available as this study was published in abstract form only. There are no other prospective longitudinal studies examining persistent proteinuria as a risk factor in IgAN; only retrospective clinicopathological studies showing that baseline proteinuria is associated with later development of adverse renal outcomes [1]. It is important to define the clinical setting in which persistent proteinuria occurs. A protein excretion of less than 4 guday is seen in both interstitial and glomerular diseases, whereas an excretion rate greater than 4 guday is almost invariably the result of glomerular disease. In general terms, in glomerulonephritis associated with the nephrotic syndrome, a progressive decline in proteinuria to less than 2 guday (or less) is associated with a favourable prognostic outlook, whether the reduction occurs spontaneously or in response to treatment [12]. In idiopathic membranous nephropathy, another primary glomerular disease associated usually with nephrotic-range proteinuria, Pei et al. [3] constructed a predictive model for renal progression using a 6-month period of the highest amount of UP as the best predictor. Membranous nephropathy differs from IgAN in that it is associated with more massive
proteinuria, whereas in IgAN, UP excretion averages between 1000 and 3000 mgu24 h and is quite variable, as reported in most study cohorts including the two study groups that were examined in the present study [1,5–7]. It is not surprising, therefore, that the longitudinal determination of proteinuria that was carried out in the present study did not have a strong, independent predictive value in estimating later progressive renal failure. However, a reduction in proteinuria is considered by many investigators to be the hallmark for determining treatment effectiveness in preserving renal function in non-diabetic renal disease. This is exemplified in a recent meta-analysis of 11 randomized clinical trials testing the efficacy of ACE inhibitors in non-diabetic renal disease [13]. This study showed that higher levels of baseline and current levels of UP were associated with progressive renal disease in both ACE inhibitor and control groups. Patients were not analysed according to the causes of their renal diseases. Therapeutic interventions using a variety of pharmacological agents in IgAN have had varying effects on UP [14]. In three randomized clinical trials [15–17] and a large cohort retrospective study [18], a variety of ACE inhibitors has been shown to modestly lower UP in patients with IgAN. However, no improvement in renal function was demonstrated. Angiotensin receptor antagonists and ACE inhibitors reduced proteinuria equally in two short-term trials [19,20], and the combination was additive in one of these studies [20]. In a randomized clinical trial, treatment with corticosteroids was shown to lower proteinuria 50% after 6 months and to reduce renal progression by 36% after 5 years [21]. In three randomized trials testing the
Predictors of ESRD in IgA nephropathy
efficacy of v-3 PUFA there was an inconsistent reduction in proteinuria [5–7,22,23] raising concerns about the long-term efficacy of v-3 PUFA [14]. The present observational study was carried out to determine the value of UP trends as a predictor of later adverse renal outcome and was based on two study cohorts that involved clinical trials testing efficacy of v-3 PUFA. In the first placebo-controlled, randomized, 2-year trial (IgAN 1), despite the differences in renal end points favouring the v-3 PUFAtreated group, the overall reduction in UP (based on slopes) was modest and not significantly different between the v-3 PUFA and placebo groups, nor between normotensive and hypertensive patients, the latter having been treated primarily with an ACE inhibitor [5,6]. However, the magnitude of the reduction in proteinuria was similar to that achieved in studies reporting the effects of ACE inhibitors in patients with IgAN [15–18], the majority of whom had UP levels in the sub-nephrotic range, i.e. between 1000 and 3000 mgu24 h, as was the case with our patients. In the second study, an open-label, v-3 PUFA two-dose, comparative trial (IgAN 2), there was a modest decline in proteinuria over time and evidence of a slowing in the rate of renal function loss in high-risk patients with moderately advanced disease [7]. In high-risk patients with IgAN, the predictive power of UP excretion at one point in time is greater than any potential effects of therapeutic intervention, whether ACE inhibition or v-3 PUFA. Acknowledgements. The authors acknowledge Cherish Grabau for secretarial support. This work was supported in part by a research grant from Mayo Foundation, Rochester, MN, USA.
References 1. Radford MG Jr, Donadio JV Jr, Bergstralh EJ, Grande JP. Predicting renal outcome in IgA nephropathy. J Am Soc Nephrol 1997; 8: 199–207 2. Kobayashi Y, Hiki Y, Sano T et al. Prognostic significance of persistent massive proteinuria in IgA nephropathy: 10-year follow-up study of 366 cases. J Am Soc Nephrol 2000; 11: 67A (abstract) 3. Pei Y, Cattran DC, Greenwood C. Predicting chronic renal insufficiency in idiopathic membranous glomerulonephritis. Kidney Int 1992; 42: 960–966 4. Cattran DC, Pei Y, Greenwood CMT, Ponticelli C, Passerini P, Honkanen E. Validation of a predictive model of idiopathic membranous nephropathy: its clinical and research implications. Kidney Int 1997; 51: 901–907 5. Donadio JV Jr, Bergstralh EJ, Offord KP, Spencer DC, Holley KE. A controlled trial of fish oil in IgA nephropathy. Mayo Nephrology Collaborative Group. N Engl J Med 1994; 331: 1194–1199
1203 6. Donadio JV Jr, Grande JP, Bergstralh EJ, Dart RA, Larson TS, Spencer DC. The long-term outcome of patients with IgA nephropathy treated with fish oil in a controlled trial. Mayo Nephrology Collaborative Group. J Am Soc Nephrol 1999; 10: 1772–1777 7. Donadio JV Jr, Larson TS, Bergstralh EJ, Grande JP. A randomized trial of high-dose compared with low-dose v-3 fatty acids in severe IgA nephropathy. Mayo Nephrology Collaborative Group. J Am Soc Nephrol 2001; 12: 791–799 8. Yoshimoto M, Tsukahara H, Saito M et al. Evaluation of variability of proteinuria indices. Pediatr Nephrol 1990; 4: 136–139 9. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457–481 10. Cox DR. Regression models and life-tables (with discussion). J R Stat Soc [Ser B] 1972; 34: 187–220 11. Ruggenenti P, Perna A, Mosconi L, Pisosi R, Remuzzi G. Urinary protein excretion rate is the best independent predictor of ESRF in non-diabetic proteinuric chronic nephropathies. Kidney Int 1998; 53: 1209–1216 12. Donadio JV Jr, Torres VE, Velosa JA et al. Idiopathic membranous nephropathy: the natural history of untreated patients. Kidney Int 1988; 33: 708–715 13. Jafar TH, Stark PC, Schmid CH et al. Proteinuria as a modifiable risk factor for the progression of non-diabetic renal disease. AIPRD Study Group. Kidney Int 2001; 60: 1131–1140 14. Julian BA. Treatment of IgA nephropathy. Semin Nephrol 2000; 20: 277–285 15. Maschio G, Cagnoli L, Claroni F et al. ACE inhibition reduces proteinuria in normotensive patients with IgA nephropathy: a multicentre, randomized, placebo-controlled study. Nephrol Dial Transpl 1994; 9: 265–269 16. Bannister KM, Weaver A, Clarkson R, Woodroffe AJ. Effect of angiotensin-converting enzyme and calcium channel inhibition on progression in IgA nephropathy. Contrib Nephrol 1995; 111: 184–193 17. Cheng IKP, Fang GX, Wong MC, Ji YL, Chan KW, Yeung HWD. A randomized prospective comparison of nadolol, captopril with or without ticlopidine on disease progression in IgA nephropathy. Nephrology 1998; 4: 19–26 18. Cattran DC, Greenwood C, Ritchie S. Long-term benefits of angiotensin-converting enzyme inhibitor therapy in patients with severe immunoglobulin A nephropathy: a comparison to patients receiving treatment with other antihypertensize agents and to patients receiving no therapy. Am J Kidney Dis 1994; 23: 247–254 19. Perico N, Remuzzi A, Sangalli F et al. The antiproteinuric effect of angiotensin antagonism in human IgA nephropathy is potentiated by indomethacin. J Am Soc Nephrol 1998; 9: 2308–2317 20. Russo D, Pisani A, Balletta MM et al. Additive antiproteinuric effect of converting enzyme inhibitor and losartan in normotensive patients with IgA nephropathy. Am J Kidney Dis 1999; 33: 851–856 21. Pozzi C, Bolasco PG, Fogazzi GB et al. Corticosteroids in IgA nephropathy: a randomized controlled trial. Lancet 1999; 353: 883–887 22. Bennett WM, Walker RG, Kincaid-Smith P. Treatment of IgA nephropathy with eicosapentaenoic acid (EPA): a two-year prospective trial. Clin Nephrol 1989; 31: 128–131 23. Pettersson EE, Rekola S, Berglund L et al. Treatment of IgA nephropathy with v-3 polyunsaturated fatty acids: a prospective, double-blind, randomized study. Clin Nephrol 1994; 41: 183–190
Received for publication: 13.8.01 Accepted in revised form: 31.1.02
