Proteinuria as a Risk Marker for the Progression of Chronic Kidney Disease in Patients on Predialysis Care and the Role of Angiotensin-Converting Enzyme Inhibitor/Angiotensin II Receptor Blocker Treatmentde Goeij M.C.M.a · Liem M.a · de Jager D.J.a · Voormolen N.a · Sijpkens Y.W.J.c · Rotmans J.I.b · Boeschoten E.W.d · Dekker F.W.a · Grootendorst D.C.a, e · Halbesma N.a · and the PREPARE-1 Study Group
Departments of aClinical Epidemiology and bNephrology, Leiden University Medical Center, Leiden, cDepartment of Internal Medicine, Bronovo Hospital, The Hague, dHans Mak Institute, Naarden, and eLinnaeus Research Institute, Kennemer Gasthuis, Haarlem, The Netherlands Corresponding Author
Background/Aims: Proteinuria is a risk marker for progression of chronic kidney disease (CKD) and treatment with an angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker (ACEi/ARB) is beneficial in these patients. However, little is known about proteinuria and ACEi/ARB treatment in patients on specialized predialysis care. Therefore, we investigated the association of urinary protein excretion (UPE) and ACEi/ARB treatment with renal function decline (RFD) and/or the start of renal replacement therapy (RRT) in patients on predialysis care. Methods: In the PREPARE-1 cohort, 547 incident predialysis patients (CKD stages IV–V), referred as part of the usual care to outpatient clinics of eight Dutch hospitals, were included (1999–2001) and followed until the start of RRT, mortality, or January 1, 2008. The main outcomes were rate of RFD, estimated as the slope of available eGFR measurements, and the start of RRT. Results: Patients with mild proteinuria (>0.3 to ≤1.0 g/24 h) had an adjusted additional RFD of 0.35 ml/min/1.73 m2/month (95% CI: 0.01; 0.68) and a higher rate of starting RRT [adjusted HR: 1.70 (1.05; 2.77)] compared with patients without proteinuria (≤0.3 g/24 h). With every consecutive UPE category (>1.0 to ≤3.0, >3.0 to ≤6.0, and >6.0 g/24 h), RFD accelerated and the start of RRT was earlier. Furthermore, patients starting (n = 16) or continuing (n = 133) treatment with ACEi/ARBs during predialysis care had a lower rate of starting RRT compared with patients not using treatment [n = 152, adjusted HR: 0.56 (0.29; 1.08) and 0.90 (0.68; 1.20), respectively]. Conclusion: In patients on predialysis care, we confirmed that proteinuria is a risk marker for the progression of CKD. Furthermore, no evidence was present that the use of ACEi/ARBs is deleterious.
Copyright © 2012 S. Karger AG, Basel
Proteinuria is an important risk marker for renal function decline (RFD) [1,2,3] and progression to chronic kidney disease (CKD) [4,5,6,7] in the general ‘healthy’ population. In patients with CKD (mainly stages I–IV), proteinuria, defined by a specific cutoff value, remains a risk marker for CKD progression [4,8,9,10,11,12,13,14]. Proteinuria deserves much attention because it is a ‘modifiable’ risk marker that can be treated with antihypertensive medication such as an angiotensin-converting enzyme inhibitor (ACEi) or angiotensin II receptor blocker (ARB).
Several trials have shown that treatment with ACEi/ARBs can delay the progression to end-stage renal disease and lower the risk of mortality in patients with CKD [15,16,17,18,19,20]. Brenner et al.  showed that treatment with an ARB results in a 28% risk reduction of progression to end-stage renal disease and 25% risk reduction of doubling of serum creatinine. For ACEi, the results vary from risk reductions of approximately 22–56% [16,17,18,19,20]. Based on this evidence, the current guidelines recommend to start treatment with ACEi/ARBs, independent of the exact stage of CKD, when the protein-to-creatinine ratio is ≥500–1,000 mg/g (equivalent to ≥0.5–1.0 g/24 h) . However, for the specific population of patients with CKD stages IV–V on specialized predialysis care, little evidence is available about whether proteinuria is still a risk marker for disease progression and whether a higher proteinuria level strengthens this effect (dose-response relation). Furthermore, it is unknown whether treatment with ACEi/ARBs might delay the need for renal replacement therapy (RRT) in this specific patient group on predialysis care.
Indeed, the evidence for the effect of ACEi/ARBs on CKD progression is contradictory in patients on predialysis care who are in the transition phase between advanced CKD and dialysis. In clinical practice, not prescribing ACEi/ARBs or discontinuing ACEi/ARBs during predialysis care is common practice, especially in the elderly and patients with acute renal failure. This practice was recently supported by a study that showed a delayed onset of RRT after the discontinuation of ACEi/ARBs in advanced CKD patients (stages IV–V) . Furthermore, a debate is ongoing concerning the general applicability of the many trials showing a delayed progression to end-stage renal disease when treated with ACEi/ARBs in patients with mainly CKD stages I–IV .
Taking all these considerations into account, the first aim of this study was to investigate whether and from what level urinary protein excretion (UPE) is an independent risk marker for the progression of CKD, assessed as RFD and the start of RRT, in the specific population of patients on specialized predialysis care (advanced CKD stages IV–V). Our second aim was to investigate what the frequency is of different prescription choices of ACEi/ARB treatment at the start of predialysis care (yes or no treatment) and during predialysis care (no use, start, stop, or continue treatment) within the population of patients on predialysis care and whether treatment with ACEi/ARBs at the start and during predialysis care were associated with the progression to RRT.
Subjects and Methods
The PREPARE-1 study is a follow-up study in which consecutive incident adult patients with advanced CKD were included from outpatient clinics of eight Dutch hospitals when referred for predialysis care between 1999 and 2001. Referral to these outpatient clinics was indicated if patients were expected to need RRT within 1 year and if their estimated creatinine clearance was below 20 ml/min. Patients who spent less than 1 month on predialysis care or patients with prior RRT were excluded. The clinical course of predialysis patients was followed through the medical charts until the start of dialysis, transplantation, death, lost to follow-up, or January 1, 2008, whichever was earliest. Predefined data on demography, anthropometry, and clinical symptoms were extracted from medical charts at the start and end of follow-up. All available data concerning laboratory measurements during predialysis care were extracted from the hospital information systems. The use of medication and laboratory measurements during predialysis care were extracted until January 1, 2003. The study was approved by the institutional review boards of the participating hospitals.
A 24-hour urine sample was collected routinely at the first predialysis visit, and the amount of excreted protein (g) in this sample was measured according to the standard procedure applied in each outpatient clinic. Data on treatment with ACEi/ARBs was collected at the first predialysis visit and at the end of follow-up. All other variables used for the adjustment in multivariable analyses were routinely assessed at the first predialysis visit and measured according to the standard procedure applied in each outpatient clinic. The presence of proteinuria was defined as excreting more than 0.3 g/24 h protein in the urine, according to the guidelines . Glomerular filtration rate (eGFR) was estimated using the abbreviated Modification of Diet in Renal Disease formula, taking into account age, sex, race, and serum creatinine . Baseline serum creatinine, hemoglobin, and proteinuria measurements were defined as the measurement closest to the start of predialysis care, within 90 days before and 14–30 days after the start of predialysis care.
The study outcomes were RFD and the start of RRT during complete follow-up (until January 1, 2008). For the patients who were still in the study after January 1, 2003 (n = 75), no complete follow-up data on eGFR and medication use was available. To prevent loss of power, we also included these patients and made the assumption that RFD followed a linear pattern  and medication use remained stable after this date. It is plausible to assume that these 75 patients are ‘stable’ patients because none of them started dialysis within the first 2 years of predialysis care. In each individual patient, the rate of RFD was estimated as the slope of a linear regression model including all available eGFR measurements. eGFR measurements between 1 month prior to inclusion and 2 weeks before reaching an endpoint were used and at least two measurements had to be available to estimate the rate of RFD. Furthermore, the start of RRT was defined as starting dialysis or being transplanted during predialysis care.
Continuous data were expressed as means ± SD and skewed data as medians [boundaries of interquartile range (IQR)]. UPE was divided into five categories: ≤0.3 (no proteinuria, reference group), >0.3 to ≤1.0, >1.0 to ≤3.0, >3.0 to ≤6.0, and >6.0 g/24 h. This classification is based on the definition of proteinuria (>0.3 g/24 h) and the upper limit of the treatment target goal of proteinuria (≤1.0 g/24 h) below which UPE at least should be reduced. Furthermore, to investigate the frequency of different ACEi/ARB prescription choices and the association of ACEi/ARB treatment with the start of RRT, we performed an analysis with ACEi/ARB treatment at the start of predialysis care (yes or no treatment) and the change of ACEi/ARB treatment during predialysis care (no use, start, stop, or continue treatment) as determinants. We also chose to investigate the change of ACEi/ARB treatment during predialysis care because it is known from clinical practice that in certain patients on predialysis care, often with the worst prognosis, treatment is not started or discontinued to preserve renal function. For the change of treatment during predialysis care, four categories were defined based on the prescription of ACEi/ARBs at the start of predialysis care and at the end of follow-up. The categories were coded as follows: (1) no use of treatment (reference group), ‘no’ at the start and ‘no’ at the end of follow-up; (2) start treatment, ‘no’ at the start and ‘yes’ at the end of follow-up; (3) stop treatment, ‘yes’ at the start and ‘no’ at the end of follow-up, and (4) continue treatment, ‘yes’ at the start and ‘yes’ at the end of follow-up. The baseline and treatment characteristics were presented for the total study population and for the population stratified by the five UPE categories.
A linear regression analysis was used to investigate the association of UPE, both continuously and in the defined five categories, with RFD. Multivariable analyses were used to adjust for the possible confounders age, sex, primary kidney disease, baseline eGFR, systolic blood pressure, hemoglobin level, presence of cardiovascular disease (angina pectoris, coronary disease, and/or myocardial infarction), and presence of diabetes mellitus. Systolic blood pressure and/or hemoglobin level were missing for 9 patients.
Cox proportional hazard regression analysis was used to assess the association of: (1) UPE, continuously and in the five defined categories, and (2) ACEi/ARB treatment, at the start (yes, no), and changes during predialysis care (no use, start, stop, continue), with the start of RRT. The hazard ratio of starting RRT was adjusted for the same confounders as used in the multivariable linear regression analyses described earlier. The time from the first predialysis visit until the start of RRT was used as follow-up time in the Cox proportional hazard regression model. Both ‘mortality’ and ‘lost to follow-up’ were censored events. Finally, we also investigated whether the association of ACEi/ARB treatment at the start or during predialysis care with the start of RRT is dependent on the level of proteinuria by performing analyses stratified by the five UPE categories.
Multiple sensitivity analyses were performed. First, we investigated whether the association of UPE with the progression of CKD is different after stratifying by treatment with ACEi/ARBs at the start of predialysis care. Second, all analyses with the start of RRT as outcome were repeated with follow-up time until January 1, 2003 instead of January 1, 2008. Third, a linear mixed model was used to validate our chosen method for estimating the rate of RFD. Fourth, systolic blood pressure, hemoglobin, UPE, and RFD were imputed (using 5 repetitions) in patients who had no prior RRT and received at least 1 month of predialysis care (n = 525) with a missing value at baseline (n = 17, n = 68, n = 112, and n = 53, respectively). Multiple imputation is a recommended technique where missing data for a patient are imputed by a value that is predicted by other known characteristics of this patient (i.e. demographic, anthropometric, and clinical characteristics, as well as the outcome) [27,28]. Multiple imputation may increase statistical power because all patients are included in the analyses. At the same time, it may reduce selection bias as patients with missing values often have poor prognosis and are therefore more prone to have poor outcomes. Data were analyzed with PASW/SPSS version 17.
In our predialysis cohort, 547 patients were included, of whom 525 patients had no prior RRT and received at least 1 month of predialysis care. At baseline, UPE was available for 413 patients and these patients were included in our statistical analyses. These 413 patients were slightly younger, had a higher systolic blood pressure, lower eGFR, and higher prevalence of cardiovascular disease and diabetes mellitus compared with patients without an available UPE at baseline (n = 112). At the start of predialysis care, 368 of the 413 patients (89%) had proteinuria (>0.3 g/24 h; table 1). Patients with proteinuria were younger, more often male, had a higher prevalence of diabetes mellitus, a lower eGFR, and higher systolic blood pressure. These differences became more pronounced with increasing UPE.
|Table 1. Baseline characteristics of the total population and stratified by UPE categories|
At the start of predialysis care, 41% (n = 168) of the 413 patients were not treated and 59% (n = 245) were treated with ACEi/ARBs (fig. 1). Furthermore, of all patients, 4% (n = 16) started treatment and 27% (n = 112) stopped treatment with ACEi/ARBs during predialysis care. This means that 10% (16 of the 168) of the patients not treated at the start of predialysis care started treatment, and 46% (112 of the 245) of the patients treated at the start of predialysis care stopped ACEi/ARB treatment. The treatment characteristics did not differ much between the UPE categories (fig. 1). The frequency of patients starting ACEi/ARB treatment was somewhat lower in patients with proteinuria (>0.3 to ≤6.0 g/24 h) compared with patients without proteinuria. Furthermore, the frequency of patients continuing treatment was somewhat higher in patients with proteinuria.
|Fig. 1. Treatment with ACEi/ARBs during predialysis care. Treatment characteristics of the total population and stratified by UPE categories. On the y-axis the percentage of patients in each of the 4 ACEi/ARB treatment categories during predialysis care (legend) is presented. The 4 treatment categories with ACEi/ARBs are based on the use or no use of ACEi/ARBs at the start of predialysis care and at the end of follow-up (no use, start, stop, and continue). The results are given for the total population (all, x-axis) and stratified by UPE categories (≤0.3, >0.3 to ≤1.0, >1.0 to ≤3.0, >3.0 to ≤6.0, and >6.0 g/24 h, x-axis).|
Two or more eGFR measurements for estimating the rate of RFD, were available for 408 patients. In these patients, the median (IQR) number of available eGFR measurements was 13 (7–19) and the mean ± SD RFD was 0.43 ± 0.87 ml/min/1.73 m2/month. RFD accelerated with increasing UPE [adjusted additional decline: 0.04 ml/min/1.73 m2/month per g/24 h increase (95% CI: 0.01; 0.08); table 2]. The same trend was visible when UPE was analyzed in categories. Even patients with mild proteinuria (>0.3 to ≤1.0 g/24 h) already had an accelerated RFD [adjusted additional decline: 0.35 ml/min/1.73 m2/month (0.01; 0.68)] compared with patients without proteinuria (≤0.3 g/24 h, mean ± SD decline: 0.07 ± 0.84 ml/min/1.73 m2/month). Every consecutive higher UPE category showed an increasing additional decline compared with the reference category (patients without proteinuria, ≤0.3 g/24 h).
|Table 2. Association of UPE with RFD|
At the end of follow-up (January 1, 2008), 43% started hemodialysis and 37% peritoneal dialysis, 4% were transplanted, 11% died, and 5% were lost to follow-up. The median (IQR) follow-up time was 11.6 (4.7–22.4) months. For the consecutive UPE categories the follow-up time was 23.0 (10.9–42.5), 16.4 (6.0–29.2), 14.1 (4.8–24.1), 7.0 (4.2–15.0), and 6.9 months (3.2–17.7). The association of increasing UPE with the start of RRT showed a similar pattern as increasing UPE with RFD [adjusted HR: 1.06 (1.02; 1.10) per g/24 h increase; table 3]. Patients with mild proteinuria (>0.3 to ≤1.0 g/24 h) had a higher rate of starting RRT compared with patients without proteinuria [adjusted HR: 1.70 (1.05; 2.77)]. Every consecutive higher UPE category showed an even higher rate compared with the reference category.
|Table 3. Association of UPE with the start of RRT|
The median (IQR) follow-up time was 11.4 months (4.8–22.1) for patients treated and 12.5 months (4.2–23.2) for patients not treated with ACEi/ARBs at the start of predialysis care. The Cox proportional hazard regression model also showed that patients treated with ACEi/ARBs at the start of predialysis care had a somewhat higher rate of starting RRT compared with patients not treated with ACEi/ARBs [adjusted HR: 1.13 (0.89; 1.44); table 4]. For patients not using, starting, stopping, or continuing treatment during predialysis care, the follow-up time was 11.7 (4.1–21.7), 16.0 (4.6–34.7), 8.2 (4.5–20.6), and 11.8 (6.7–24.9) months, respectively. Patients starting treatment with ACEi/ARBs during predialysis care had a lower rate of starting RRT compared with patients not using treatment with ACEi/ARBs, which was not significant [adjusted HR: 0.56 (0.29; 1.08); table 4]. Patients continuing treatment with ACEi/ARBs during predialysis care also had a slightly lower rate of starting RRT compared with the patients not using treatment [adjusted HR: 0.90 (0.68; 1.20)]. Compared with the reference group, patients discontinuing the use of ACEi/ARBs during predialysis care had a higher rate of starting RRT. However, this difference was not significant [adjusted HR: 1.27 (0.95; 1.70)]. Furthermore, the association of ACEi/ARB treatment at the start or during predialysis care with the start of RRT varied through the UPE categories; however, no clear pattern was detected.
|Table 4. Association of treatment with ACEi/ARBs with the start of RRT|
The point estimates for the association of UPE with RFD and the start of RRT were higher in patients treated compared with patients not treated with ACEi/ARBs at the start of predialysis care. The point estimates for the association of UPE and the use of ACEi/ARBs with the start of RRT remained similar when using follow-up until January 1, 2003 for all patients. Only the confidence intervals for the hazard ratios became somewhat larger, probably because 75 patients had not reached an endpoint on this date yet. Using linear mixed models resulted in similar significant results for the association of UPE with RFD, with lower point estimates and smaller confidence intervals. Finally, multiple imputation resulted in similar results for the continuous and categorical association of UPE with RFD and the start of RRT.
In our cohort of patients on specialized predialysis care, 89% of the patients had proteinuria (>0.3 g/24 h) and almost half of the patients, irrespective of proteinuria, were not treated with ACEi/ARBs. Of all 413 patients, only 10% of the patients not treated with ACEi/ARBs at the start of predialysis care started the treatment, and 46% of the patients treated with ACEi/ARBs at the start of predialysis care discontinued the treatment during predialysis care. Increased UPE was associated with a faster progression of CKD, assessed as an accelerated RFD and an earlier start of RRT. Even patients with mild proteinuria (>0.3 to ≤1.0 g/24 h) already experienced a faster progression of CKD compared with patients without proteinuria (≤0.3 g/24 h) and every consecutive higher UPE category resulted in an even faster progression of CKD. Patients treated with ACEi/ARBs at the start of predialysis care started RRT somewhat earlier. In contrast, according to ACEi/ARB treatment during predialysis care, patients continuing or starting treatment with ACEi/ARBs during predialysis care, had a somewhat later start of RRT compared with patients not using ACEi/ARB treatment. Furthermore, patients discontinuing treatment had a considerable faster start of RRT compared with patients not using ACEi/ARB treatment during predialysis care. Therefore, the small increased rate of starting RRT in all patients on ACEi/ARB treatment at the start of predialysis care can be explained by the increased rate and shorter median follow-up time in patients discontinuing treatment with ACEi/ARBs during predialysis care as seen in our analyses. These results may indicate that in the clinic, treatment with ACEi/ARBs is especially discontinued in patients with a bad prognosis.
Our findings of a dose-response relation between increasing UPE and a faster progression of CKD, assessed as RFD and the start of RRT, has not been shown previously in patients with advanced CKD (stages IV–V). A faster progression was already present in patients with mild proteinuria (>0.3 to ≤1.0 g/24 h). Other observational studies have investigated proteinuria as a risk marker for CKD progression [4,8,10,11,12,13,14]. However, these studies included patients in all stages of CKD and/or proteinuria was defined only as a dichotomous variable. The study of Obi et al.  resembles our study the most, as this study was performed in a cohort of patients with CKD stages III–V referred to nephrologists. They showed that patients with overt proteinuria, defined as a urinary protein/creatinine ratio ≥1 g protein/g creatinine or a urine dipstick ≥2+ (both equivalent to ≥1.0 g/24 h), started RRT earlier compared with patients without overt proteinuria [adjusted HR: 4.97 (95% CI: 2.23–11.1)]. Besides these observational studies, a meta-analysis by Jafar et al.  showed that the beneficial effect of ACEis/ARBs on the progression of CKD, defined as the doubling of serum creatinine or progression to end-stage renal disease, acts in part via the lowering of proteinuria. This finding may indicate that lowering proteinuria slows down the progression of CKD which is consistent with the dose-response relation we found between increasing UPE and a faster progression of CKD.
In our cohort, 59% of the patients (245 of the 413 patients) were treated with ACEi/ARBs (most of them used ACEi) and this finding is in line with other studies [29,30]. Furthermore, we found that 112 of the 245 patients (46%) who were treated with ACEi/ARBs at the start of predialysis care discontinued the use of ACEi/ARBs during predialysis care. For the 168 patients not treated with ACEi/ARBs at the start of predialysis care, only 16 patients (10%) started treatment. These findings are in line with the clinical practice because in many predialysis patients treatment with ACEi/ARBs is not started or discontinued to preserve the patients’ renal function or because of hyperkalemia. Recently, this practice became more evidence based, as Ahmed et al.  showed that discontinuing the use of ACEi/ARBs delayed the onset of RRT in patients with CKD stages IV–V. However, their patient population was very specific, consisting of elderly patients (mean ± SD: 73.3 ± 1.8 years) with relatively high eGFR (16.38 ± 1 ml/min/1.73 m2). Unfortunately, in our cohort we could not investigate whether kidney function improves after discontinuation of ACEi/ARB treatment because we do not know the exact date of discontinuation. Therefore, future studies should focus on this possible beneficial effect of not prescribing ACEi/ARB treatment in patients on predialysis care.
Our study furthermore showed that ACEi/ARB treatment at the start of predialysis care was associated with a somewhat faster progression to RRT [adjusted HR: 1.13 (95% CI: 0.89; 1.44)]. This finding is in line with the study of Ahmed et al. , who showed that treatment with ACEi/ARBs in predialysis patients is detrimental. However, this possible detrimental effect of ACEi/ARBs disappeared when we performed an analysis taking into account the treatment of ACEi/ARBs during predialysis care (no use, start, stop, or continue treatment). Patients who discontinue ACEi/ARB treatment during predialysis care started RRT earlier than patients not using treatment, which explained the detrimental effect found at baseline. However, patients continuing or starting treatment with ACEi/ARBs during predialysis care have a somewhat slower progression to RRT compared with patients not using treatment with ACEi/ARBs. This finding is in line with several trials performed in CKD patients [15,16,17,18,19,20], but in contrast to the study of Ahmed et al. . An explanation for this contradictory finding could be that in the study of Ahmed et al. , all included patients discontinued ACEi/ARB treatment and in our study, ACEi/ARB treatment was only discontinued in patients with clinical indications to stop, such as hyperkalemia, renal dysfunction, and/or angioedema .
Our results representing the association of ACEi/ARB treatment with the start of RRT are therefore contradictory. We found a very small detrimental effect of ACEi/ARB treatment, but this effect became somewhat beneficial when we excluded patients that discontinue the treatment. Because of this current clinical practice to stop or not start ACEi/ARB treatment in some predialysis patients to preserve renal function, we could get problems with confounding by indication. Therefore, it is difficult to prove whether our findings of a somewhat beneficial effect of treatment with ACEi/ARBs during predialysis care is due to the absence of clinical indications to stop the treatment, often related to a better prognosis, or due to a real beneficial effect of ACEi/ARBs. The combination of our observational results showing that the clinical practice is to discontinue ACEi/ARB treatment in specific predialysis patients and the fact that the patients in the study of Ahmed et al.  were mainly elderly patients with a relatively high eGFR may support the thought that ACEi/ARB treatment is not beneficial for a specific subgroup of predialysis patients. However, future studies in predialysis patients should elaborate whether discontinuation of ACEi/ARBs is indeed beneficial, and if so, for which specific subgroups.
Our study has some potential limitations. First, in the general population, 36–65% of the patients with an eGFR below 15 ml/min/1.73 m2 are not treated by a nephrologist . Therefore, our results may not be generalizable to all patients with an eGFR below 15 ml/min/1.73 m2, but only to those who receive specialized predialysis care. However, for clinical practice our cohort is a highly representative and relevant population because this is the patient population seen and treated by nephrologists. Second, for estimating the rate of RFD with available eGFR measurements, we assumed that RFD follows a linear pattern in the advanced stages of CKD. It has been shown previously that linearity of the course in eGFR is a reliable assumption, although on theoretical grounds an exponential decline could be present over a longer period of time . Third, our finding that treatment with ACEi/ARBs during predialysis care resulted in a slower progression of CKD should be interpreted with caution due to the observational character of our study and thereby possible confounding by indication. To draw confirmative conclusions about the effect of a treatment, a randomized controlled trial is the most optimal study design to ensure an equal prognosis between the two treatment groups. However, the results in our study did not change essentially after the adjustment for clinical indications (high blood pressure and proteinuria) and contraindications (hyperkalemia and edema; data not shown) considered in the decision to prescribe antihypertensive medication. Unfortunately, we cannot adjust for RFD before the start of predialysis care, the third main indication. No RFD could be calculated because the number of available creatinine levels prior to 1 month before the start of predialysis care was too scarce in our cohort. However, if patients treated with ACEi/ARBs indeed had a faster RFD before the start of predialysis care, our observed associations might even be stronger. Fourth, the prescription of ACEi/ARBs was only assessed at the start of predialysis care and at the end of follow-up. The treatment regimen in the intervening time and the actual compliance to the medication is unknown and could not be accounted for. However, it is unlikely that patients stop and restart treatment with ACEi/ARBs during predialysis care. Besides this, we do not know the exact date of discontinuation of ACEi/ARBs, which could be just before the start of dialysis. Furthermore, we were not able to perform an analysis in which the single treatment with ACEis or ARBs and the combined treatment with ACEis and ARBs was investigated due to the low frequency of patients using ARBs.
Our results indicate that during specialized predialysis care, proteinuria can be used as a risk marker, independent of other risk factors, for the progression of CKD. High-risk predialysis patients can be identified, treated more strictly, and timely preparation for dialysis can be assured. Furthermore, in our predialysis cohort, a large part of the patients were not treated with ACEi/ARBs to lower proteinuria and slow down the progression of CKD. More studies are necessary to elaborate on the clinical indications for discontinuation of ACEi/ARB treatment and the clinical course after discontinuation. Moreover, patients starting or continuing treatment with ACEi/ARBs during predialysis care started RRT somewhat later, indicating no deleterious effect of ACEi/ARB treatment. In conclusion, the results from this study provide evidence for using the proteinuria level as a strong and independent risk marker for the progression of CKD in patients on specialized predialysis care.
Trial nurses, data managers, and students from the Hans Mak Institute are gratefully acknowledged for their assistance in data collection. Furthermore, we thank all the laboratory information system managers who invested time and effort to supply laboratory data, and all supporting staff who helped tracing records of (eventually) every patient. This study was supported by an unrestricted grant from Amgen BV.
The PREPARE-1 study is an independent academic study designed and carried out by the Department of Clinical Epidemiology from the Leiden University Medical Center in collaboration with the Hans Mak Institute (Naarden) and the participating hospitals. This study was funded by an unrestricted grant from Amgen BV (the Netherlands and Switzerland). Amgen BV was not involved in study design, collection of data, statistical analyses, interpretation of data, writing of the manuscript, or in the decision to submit the paper for publication. None of the authors have declared a conflict of interest.
Moniek C.M. de Goeij
Department of Clinical Epidemiology, Leiden University Medical Center
NL–2333 ZA Leiden (The Netherlands)
The PREPARE-1 study group consists of P. Gerlag, Maxima Medical Centre, Veldhoven; C.J. Doorenbos, Deventer Ziekenhuizen, Deventer; K. Jie, Groene Hart Hospital, Gouda; A. Schrander-van der Meer, Rijnland Ziekenhuis, Leiderdorp, and C. Verburgh, Kennemer Gasthuis, Haarlem, The Netherlands.
Received: April 26, 2012
Accepted: August 3, 2012
Published online: October 30, 2012
Number of Print Pages : 10
Number of Figures : 1, Number of Tables : 4, Number of References : 32
Nephron Clinical Practice
Vol. 121, No. 1-2, Year 2012 (Cover Date: December 2012)
Journal Editor: McIntyre C. (Derby)
ISSN: 1660-2110 (Print), eISSN: 1660-2110 (Online)
For additional information: http://www.karger.com/NEC