Skip Navigation


NDT Advance Access originally published online on July 5, 2006
Nephrology Dialysis Transplantation 2006 21(10):2800-2808; doi:10.1093/ndt/gfl342
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
21/10/2800    most recent
gfl342v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Wakai, K.
Right arrow Articles by Sakai, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wakai, K.
Right arrow Articles by Sakai, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2006]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

A scoring system to predict renal outcome in IgA nephropathy: from a nationwide prospective study

Kenji Wakai1,, Takashi Kawamura2, Masayuki Endoh3, Masayo Kojima4, Yasuhiko Tomino5, Akiko Tamakoshi6, Yoshiyuki Ohno6, Yutaka Inaba7 and Hideto Sakai3

1Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, Nagoya 2Kyoto University Health Service, Kyoto 3 Division of Nephrology and Metabolism, Department of Internal Medicine, Tokai University School of Medicine, Isehara 4Department of Health Promotion and Disease Prevention, Nagoya City University Graduate School of Medicine, Nagoya 5Division of Nephrology, Department of Internal Medicine, Juntendo University School of Medicine, Tokyo 6Department of Preventive Medicine/Biostatistics and Medical Decision Making, Nagoya University Graduate School of Medicine, Nagoya and 7Department of Epidemiology and Environmental Health, Juntendo University School of Medicine, Tokyo, Japan

Correspondence and offprint requests to: Kenji Wakai, MD, Division of Epidemiology and Prevention, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa-ku, Nagoya 464-8681, Japan. Email: wakai{at}aichi-cc.jp



   Abstract
Top
 Abstract
Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
Background. Immunoglobulin A (IgA) nephropathy is the most common form of glomerulonephritis in the world, and a substantial number of patients develop end-stage renal disease (ESRD). Although there are several prognostic indicators, it remains difficult to predict the renal outcome in individual patients.

Methods. A prospective cohort study was conducted in 97 clinical units in Japan from 1995 to 2002. We analysed the data from 2269 patients using proportional hazards models in order to determine the predictors of ESRD in IgA nephropathy and to develop a scoring system to estimate ESRD risk.

Results. During the follow-up (median, 77 months), 207 patients developed ESRD. Systolic hypertension, proteinuria, hypoproteinaemia, azotaemia and a high histological grade at initial renal biopsy were independently associated with the risk of ESRD. Mild haematuria predisposed patients to ESRD more than severe haematuria. A scoring system was developed to estimate the 7-year ESRD risk from eight clinical and pathological variables. Actually, this prognostic score accurately classified patients by risk: patients with estimates of 0.0–0.9, 1.0–4.9, 5.0–19.9, 20.0–49.9, and 50.0–100.0% had a 0.2, 2.4, 12.2, 40.2 and 80.8% of ESRD incidence over 7 years, respectively. The corresponding area under the receiver operating characteristic curve was 0.939 [95% confidence interval (CI), 0.921–0.958]. This score was verified in repetitions of the derivation-validation technique.

Conclusions. Although the quality of some data collected by the mail survey is limited and the influence of therapy could not be considered, this scoring system will serve as a useful prognostic tool for IgA nephropathy in clinical practice.

Keywords: end-stage renal disease; IgA nephropathy; kidney failure; prospective studies; renal prognosis



   Introduction
Top
 Abstract
 Introduction
Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 
Immunoglobulin A (IgA) nephropathy is the most common form of glomerulonephritis in the world today [1]. When Berger and Hinglais [2] described this disease as a new clinical entity in 1968, clinicians regarded it as a benign nephropathy. However, successive studies indicated that 6–43% of IgA nephropathy patients would develop end-stage renal disease (ESRD) over a period of 10 years [3,4].

Many investigators have tried to determine the prognostic indicators of this disease, which include elevated serum creatinine (sCr), heavy proteinuria, severe histological changes, hypertension, hypoproteinaemia, older age and male sex [4–7].

However, difficulties remain in predicting the renal outcomes in individual patients and determining those who need aggressive therapeutic intervention. This may be partly a result of the relatively small sample sizes used in previous studies [8]. We, therefore, set up a large-scale, nationwide prospective study in order to develop an IgA nephropathy prognostic score.



   Subjects and methods
Top
 Abstract
 Introduction
 Subjects and methods
Results
 Discussion
 Acknowledgements
 References
 
Patients and follow-up
The Research Group on Progressive Renal Diseases and the Research Committee on the Epidemiology of Intractable Diseases, both organized by the former Ministry of Health and Welfare of Japan (currently the Ministry of Health, Labor and Welfare), conducted a nationwide survey on IgA nephropathy in January 1995 [9]. This survey identified 5324 patients with biopsy-proven IgA nephropathy, who had visited general physicians, nephrologists, paediatricians or urologists in Japan during 1994.

Follow-up was carried out at clinical units that had 10 or more IgA nephropathy patients in the survey. When more than 50 of them were identified in a unit, we randomly selected 50 to reduce the burden on the unit. We eventually designated 3409 patients as potential subjects for the follow-up study and undertook the first mail survey in May 1997. The response rate was 82.5% in a patient base. A second mail survey was conducted for 2350 cases in August 1999, and a third for 2285 cases in November 2002, with response rates of 95.7 and 93.3%, respectively. Excluding those who had died, who had developed ESRD before baseline, or whose essential baseline data were missing, we included 2269 eligible patients for the analyses from 97 clinical units. Figure 1 shows their distribution by sex and age. There are two peaks in age distribution: 15–24 and 40–49 years. The median follow-up period was 77 months (range, 1–94 months).


Figure 1
View larger version (25K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. Sex and age distribution of patients (1104 men and 1165 women) for analysis.

 
This investigation was approved by the Ethics Committee of the Kyoto University Graduate School of Medicine and the Ethics Committee of the Juntendo University School of Medicine.

Data collection
The baseline data of the patients were obtained by reviewing medical records in the nationwide survey in 1995. The data included sex, age, year of diagnostic renal biopsy, systolic and diastolic blood pressure, urine protein and blood, serum total protein and albumin and sCr. Proteinuria was semi-quantified with a standard urine dipstick with (–), (+–), (+), (++) and (+++) corresponding to <10, 10–29, 30–99, 100–299 and ≥300 mg/dl of urine albumin, respectively. Histological grade at initial renal biopsy was reassessed by pathologists or nephrologists in each participating hospital at the first follow-up survey using the new criteria from the Joint Committee of the Research Group on Progressive Renal Diseases and the Japanese Society of Nephrology (Table 1) [10]. This reassessment was needed because the classification criteria were established after our baseline survey. In the three follow-up surveys, information on outcomes such as death, ESRD and sCr was collected by mail.


View this table:
[in this window]
[in a new window]

  		Table 1. Criteria for histological grading from the Joint Committee of the Research Group on Progressive Renal Diseases (Ministry of Health and Welfare of Japan), and the Japanese Society of Nephrology [10]

 
Statistical analysis
Predictors of ESRD
The primary endpoint in this study was ESRD, which was defined as the initiation of dialysis therapy. The follow-up period for each patient was calculated in months from the data of the nationwide survey on ESRD, death or the last visit. Those who died without ESRD were treated as censored cases.

The 7-year cumulative incidence (risk) of ESRD was computed by the Kaplan–Meier method [11] according to demographic and clinical characteristics. The rate ratios (RRs) for ESRD were estimated by potential prognostic factors using proportional hazards models [12]. The independent effect of each variable was assessed by multivariate analysis. Diastolic blood pressure and serum albumin were excluded from the multivariate model because of their close relation with systolic blood pressure and serum total protein, respectively. To test for a linear trend, we coded each stratum of the variable as 0, 1, 2, ... and included it as a continuous variable in the proportional hazards model [13]. For variables that displayed ‘U-shape’ or ‘inverted U-shape’ relationships with ESRD risk, we also analysed non-linear associations using a quadratic model [14]. Subjects with missing values were omitted from the relevant analysis.

Construction of scoring system
A scoring system to predict ESRD in individual patients with IgA nephropathy was based on the proportional hazards model. It included sex, age and all the significant variables in the aforementioned multivariate analysis. In this model, a patient with covariate values X1, X2, ..., Xn has an expected renal survival rate (= 1 – cumulative incidence of ESRD) at time t, S (t), formulated as:


Formula

where S0 (t) is the baseline survivor function and ß1, ß2, ..., ßn are the coefficients estimated from the model [8,15]. The RRs are exponents of these ß coefficients.

We stratified the patients by each prognostic factor and applied a proportional hazards model with dummy variables (0 or 1 for X1, X2, ..., Xn). The scores derived from the ß coefficients were smoothed by linear interpolation in each stratum of the variables and were then multiplied by 10 to simplify the calculation. The sum of all the scores for individual factors (total score) was 10 x (ß1X1 + ß2X2 + ··· + ßnXn), and a corresponding ESRD risk, 1 – S (t), was computed from the total score. The baseline survivor function, S0 (t), was estimated by the product-limit method using the PHREG procedure of the Statistical Analysis System [16].

Validation of scoring system
To examine the goodness of fit of the scoring system to the data, we divided the patients into five groups according to the predicted 7-year risk of ESRD, that is, minimum (0.0–0.9%), low (1.0–4.9%), moderate (5.0–19.9%), high (20.0–49.9%) and very high (50.0–100.0%). The renal survival curve of ESRD was then drawn in each group using the Kaplan–Meier method [11]. To further assess the utility of the score, we used the area under the receiver operating characteristic (ROC) curve [17] for the 7-year risk of ESRD. The area and its 95% confidence interval (CI) were estimated by the non-parametric method [17].

As an additional analysis [18], one-third (n = 756) of the subjects were randomly allocated to a validation sample and the remainder to a derivation sample. The prognostic score was developed in the derivation sample, and the actual 7-year cumulative incidence of ESRD was computed by the predicted risk in the validation sample [11]. The area under the ROC curve was also estimated in the validation group. Considering the sampling error, we repeated this procedure in 100 different validation sets. Smoothing of the scores was not done in this analysis.



   Results
Top
 Abstract
 Introduction
 Subjects and methods
 Results
Discussion
 Acknowledgements
 References
 
During the follow-up of 11 923.5 person-years, 207 patients with IgA nephropathy developed ESRD. Sixteen deaths without ESRD were also reported: five from circulatory diseases, five from cancer and six from unknown causes.

Table 2 summarizes the 7-year cumulative incidence and RR for ESRD by demographic and clinical factors. Patients with no or only mild proteinuria [urine protein of (–) or (+–)] had an appreciably lower risk of ESRD: their 7-year cumulative incidence was 0.7% (95% CI, 0.0–1.5%). In contrast, patients with excretional impairment at baseline had a much higher risk. Their 7-year cumulative incidences were 68.5% (95% CI, 59.2–77.8%) and 90.1% (82.4–97.8%) for sCr of 1.68–2.50 and ≥2.51 mg/dl (148–221 and ≥222 µmol/l), respectively. The proportion was as high as 25.9% (95% CI, 19.2–32.6%) even among those with mild azotaemia [sCr, 1.26–1.67 mg/dl (112–147 µmol/l)]. A univariate analysis with proportional hazards models revealed that earlier renal biopsy, systolic/diastolic hypertension, proteinuria, hypo-proteinaemia/albuminaemia, azotaemia and a higher histological grade at initial renal biopsy had a strong dose-dependent association with the risk of ESRD. Systolic hypertension, proteinuria, hypoproteinaemia, azotaemia and a higher histological grade remained significant predictors in the multivariate analysis (trend P < 0.05).


View this table:
[in this window]
[in a new window]

  		Table 2. Seven-year cumulative incidence and RRs for ESRD by demographic and clinical characteristics (at baseline except for histological grade)

 
Although male patients had a higher RR in the univariate analysis, the high risk disappeared when considering other factors. Patients in their 30s were at the lowest risk in the multivariate analysis, whereas an upward trend in RR with advancing age was found in the univariate model. Mild haematuria [<30 red blood cells per high-power field (RBC/HPF)] was associated with a higher risk compared with severe haematuria (≥30 RBC/HPF) in both the univariate and multivariate analyses. In the multivariate models, we found a significant non-linear association of haematuria with ESRD risk (P for non-linear association <0.0001) but none for age (P = 0.26).

Based on these analyses, we established a scoring system to estimate 4- and 7-year cumulative incidence rates of ESRD. Of the 2269 patients included in this study, 1754 (77.3%) had a complete data set needed for our system designing. Table 3 lists the scores of individual prognostic factors. The total score (sum of individual scores) can then be converted to the corresponding estimated risk using Table 4. The baseline survivor function, S0 (t), was estimated as 0.99955026 and 0.99887700 at 4 and 7 years, respectively. An illustrative example of how to apply this scoring system to patients with IgA nephropathy follows. The patient is a 59-year-old man with systolic blood pressure of 142 mmHg, proteinuria of (++), haematuria of ≥30 RBC/HPF, serum total protein of 6.6 g/dl, histological grade of IV at initial renal biopsy, and sCr of 1.35 mg/dl. As shown in Table 3, the scores for sex, age, systolic blood pressure, proteinuria, haematuria, serum total protein, histological grade and sCr are –2, –3, 3, 20, 3, 5, 9 and 11, respectively. Thus, the total score is calculated to be 46 [(–2) + (–3) + 3 + 20 + 3 + 5 + 9 + 11]. Using Table 4, this total score of 46 can be converted to a 4- and 7-year ESRD risk of 4.4 and 10.6%, respectively.


View this table:
[in this window]
[in a new window]

  		Table 3. Prognostic scores to estimate risk of ESRD by demographic and clinical factors

 

View this table:
[in this window]
[in a new window]

  		Table 4. Estimated 4- and 7-year risk of ESRD by total score

 
The actual renal survival (= 1 – cumulative incidence of ESRD) according to the estimated 7-year risk was plotted in Figure 2. The prognostic score successfully classified the patients by risk. Those with an estimated risk of 0.0–0.9% (score, –8 to 21), 1.0–4.9% (22–38), 5.0–19.9% (39–52), 20.0–49.9% (53–64) and 50.0–100.0% (65 or more) had an actual cumulative incidence of ESRD in 7 years of 0.2% (95% CI, 0.0–0.7%), 2.4% (0.8–3.9%), 12.2% (7.2–17.3%), 40.2% (27.7–52.6%) and 80.8% (73.3–88.3%), respectively. This showed good agreement between the estimated and observed risks.


Figure 2
View larger version (14K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Renal survival curves by the predicted 7-year risk of ESRD. Patients were categorized into five groups according to the estimated risk: minimum (0.0–0.9%, n = 769), low (1.0–4.9%, n = 556), moderate (5.0–19.9%, n = 230), high (20.0–49.9%, n = 72) and very high (50.0–100.0%, n = 127). Numbers of patients at risk were 1754, 1478, 1237, 961 and 817 at 0, 2, 4, 6 and 7 years, respectively.

 
The corresponding area under the ROC curve was 0.939 (95% CI, 0.921–0.958). The area remained almost the same when the prognostic scores were not smoothed by linear interpolation (0.947; 95% CI, 0.930–0.964).

The influence of therapy could not be taken into account in the scoring system, because data on treatment were not collected at baseline and only the history of use (yes or no) of six groups of drugs (antiplatelet agents, corticosteroids, immunodepressants, angiotensin-converting enzyme inhibitors, calcium channel blockers and fish oil, and others) [19,20] at the 1997 survey was available for most of the cases. The drugs had been used for 82.2, 35.3, 10.9, 29.2, 20.0 and 34.1% of the patients, respectively, until the 1997 survey. Additional multivariate analyses that included the history of drug use did not essentially alter the associations of clinical characteristics with the ESRD risk presented in Table 2, except for the attenuation of risk for systolic blood pressure (data not shown).

Even when the prognostic scores were developed using derivation samples randomly selected from all subjects, the estimated 7-year cumulative incidence of ESRD well-predicted the observed ones in the remaining validation sample. The median values of observed 7-year incidence were 0.5% [inter-quartile range (IQR), 0.0–0.8%], 2.2% (1.3–3.0%), 12.9% (10.0–16.2%), 39.5% (32.9–44.6%) and 75.3% (70.7–79.7%) in patients with an estimated risk of 0.0–0.9, 1.0–4.9, 5.0–19.9, 20.0–49.9 and 50.0–100.0%, respectively. The median of the corresponding area under the ROC curve (0.927; IQR, 0.913–0.941) was comparable with the area in the full data set (0.939).



   Discussion
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
Acknowledgements
 References
 
Based on a large-scale cohort study, we described the prognostic indicators for IgA nephropathy and developed a scoring system for estimating the ESRD risk. Systolic hypertension, proteinuria, haematuria, hypoproteinaemia, azotaemia and a high histological grade at initial renal biopsy were related to the risk, independent of other factors. The prognostic score successfully classified patients according to their ESRD risk and was verified by the analysis, dividing the subjects into derivation and validation samples.

To the best of our knowledge, this is the largest follow-up study to date on the renal outcome of IgA nephropathy. The present investigation enrolled many IgA nephropathy patients at various clinical stages, which enabled us to quantify ESRD risk by prognostic factors in detail. Whereas most studies to date followed IgA nephropathy patients from the time of initial renal biopsy or diagnosis [4], our study included patients at different stages. Therefore, the prognostic score can be applied appropriately whenever needed. Use of clinical data subsequent to renal biopsy may improve the prediction of renal outcome [4].

As expected from previous studies [3–7], hypertension, proteinuria, hypoproteinaemia, higher sCr and more severe morphological changes were independent risk factors for ESRD. Mild haematuria (<30 RBC/HPF) was related to a higher risk of ESRD than the more severe type (≥30 RBC/HPF). This supports the clinical impression that gross haematuria is a good sign [8], although the association still remains controversial [4] and its mechanism not yet fully understood. The reproducibility of severe haematuria was not so high as that of proteinuria. The potential misclassification, if any, will attenuate the actual difference in ESRD risk between the groups of mild and severe haematuria. Considering repeated measures of haematuria may serve to further clarify the poor prognosis associated with mild haematuria.

The lower the glomerular filtration rate (GFR) of a patient reflected in higher sCr, the sooner the patient will reach ESRD if the rate of decline in GFR is the same.

Male patients were at an increased risk of ESRD in the univariate analysis, but that finding disappeared in the multivariate model. The sCr level is essentially higher in men than in women due to the greater production of creatinine in males [21]. Nevertheless, the decision when to start haemodialysis is made mainly based on the sCr level irrespective of sex. Thus, male sex would seemingly be a risk factor for ESRD. In fact, it was not a risk factor after adjustment for sCr, but the elevated risk among men appeared again in the proportional hazards model excluding 1/sCr (RR, 1.49; 95% CI, 1.05–2.10).

A U-shaped association was found between age and ESRD risk after adjustment for other prognostic factors, though the non-linear association was not statistically significant. Age was positively associated with the ESRD risk in previous univariate analyses [4], whereas negative associations were observed in preceding multivariate ones [7,22]. Most of these studies treated age as a single continuous variable in the multivariate model. Our 10-year stratification of age revealed a possible non-linear relationship.

We proposed a new scoring system for the prediction of renal outcome in individual IgA nephropathy patients. Compared with several previous studies providing prognostic measures according to clinicopathological factors, our system has some advantages in terms of background data and target outcome. First, it is based on the data from a much larger sample than the one in a previous study by Beukhof et al. [8]. Second, it predicts the risk of ESRD, rather than that of surrogate endpoints such as an increase in sCr [15]. Although D’Amico and coworkers [23] developed a scoring index using proteinuria and three histological variables, and Frimat and colleagues [24] developed a classification using sCr and 24-h proteinuria, both of which correlated well with renal survival [23,24], the ESRD risk was not quantified in these two studies. Our scoring system can quantitatively estimate the ESRD risk in IgA nephropathy patients, using clinical and pathological information collected in routine medical practice.

As shown in Figure 2 and the ROC analysis, our prognostic score works well; even when it is based on randomly selected derivation samples, the score predicted the ESRD risk in the remaining validation sample as accurately as the score derived from the whole dataset. Thus, our estimates among all subjects were fully justified.

Some limitations should also be kept in mind when utilizing the score tables. First, our prognostic score is not able to predict ESRD risk for longer than 7 years (we are conducting a further survey to extend the follow-up period for a longer-term prediction). Second, the endpoint of the present study was focused on ESRD. Although this follow-up study collected sCr values as an outcome in addition to ESRD and deaths, the analytical methods for the change in sCr level over time greatly differ from those for ESRD. We, therefore, will report analyses for the change in sCr values as an outcome after the ongoing extended survey is completed. Third, we did not address the effect of therapy because data on treatment were not obtained at baseline. Further investigations focusing on treatment may be warranted.

Another major methodological issue may be related to the limitation of the mail survey as a method to obtain laboratory data and to the quality of some data; about a quarter of the subjects did not have a complete set of data for the scoring system, and we could not check the quality of laboratory data from the participating institutions. The histological grading of biopsy specimens was done in each hospital and the inter-institutional variation in classification may have resulted in the relatively small contribution of histological grade to the prognostic score (Table 3). Because the data at baseline on 24-h urine excretion of protein were not available for two-thirds of the subjects and those on urinary protein to creatinine ratio were not collected, we had to assess proteinuria with a dipstick. The semi-quantified proteinuria was reasonably reproducible; 81.5% of patients were classified into the same or adjacent categories [patients were grouped into (–), (+–), (+), (++) and (+++) of proteinuria] in the two tests 2 years apart (at baseline and at the 1997 survey). The dipstick proteinuria was also rather strongly correlated with the 24-h urinary excretion of protein among patients with the relevant data (Spearman's correlation coefficient, 0.77 at baseline). The misclassification by the dipstick assessment, however, might have attenuated the association between proteinuria and ESRD risk. The use of dipstick did not allow us to estimate the risk for the amount of persistent proteinuria over time as suggested by a previous report [25]. Collecting more detailed laboratory data (e.g. creatinine clearance) will add to the accuracy of risk prediction by the scoring system.

Finally, our scoring system was developed among Japanese patients. It would be applicable to Western populations as well, since renal survival rates by prognostic factors are reasonably comparable between the present subjects and patients in Western countries [5,7,22,26]. To tailor the scoring system for populations other than Japanese, however, it may be warranted to adjust sCr for body weight [21] and racial groups [27] because the sCr level was the most important factor to predict ESRD risk as reflected in the wide range of prognostic scores assigned to this parameter. Additional validation studies may be helpful in optimizing this scoring system for populations of different races.

In summary, the present study found that hypertension, proteinuria, haematuria (particularly mild type), hypoproteinaemia, azotaemia and advanced histological change independently increased the risk of ESRD in IgA nephropathy patients. The ESRD-prediction score based on a multivariate model was sufficiently valid and will serve as a useful tool for IgA nephropathy in clinical practice.



   Acknowledgements
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
References
 
The authors express their sincere appreciation to the physicians who participated in this study. We also wish to thank Hidemi Hattori and Yuko Watanabe for their technical support.

 This study was supported in part by Grants-in-Aid for the Research Group on Progressive Renal Diseases and the Research Committee on the Epidemiology of Intractable Diseases from the former Ministry of Health and Welfare of Japan (currently the Ministry of Health, Labor and Welfare).

Conflict of interest statement. None declared.



   References
Top
 Abstract
 Introduction
 Subjects and methods
 Results
 Discussion
 Acknowledgements
 References
 

  1. D'Amico G. (1987) The commonest glomerulonephritis in the world: IgA nephropathy. Q J Med 64:709–727.[Web of Science][Medline]
  2. Berger J and Hinglais N. (1968) Les dépôts intercapillaires d’IgA-IgG. J Urol Nephrol 74:694–695.
  3. Koyama A, Igarashi M, Kobayashi M. Members and Coworkers of the Research Group on Progressive Renal Diseases. (1997) Natural history and risk factors for immunoglobulin A nephropathy in Japan. Am J Kidney Dis 29:526–532.[Web of Science][Medline]
  4. D'Amico G. (2004) Natural history of idiopathic IgA nephropathy and factors predictive of disease outcome. Semin Nephrol 24:179–196.[CrossRef][Web of Science][Medline]
  5. Johnston PA, Brown JS, Braumholtz DA, Davison AM. (1992) Clinico-pathological correlations and long-term follow-up of 253 United Kingdom patients with IgA nephropathy: a report from the MRC Glomerulonephritis Registry. Q J Med 84:619–627.[Web of Science][Medline]
  6. Yagame M, Suzuki D, Jinde K, et al. (1996) Value of pathological grading in prediction of renal survival in IgA nephropathy. Nephrology 2:107–117.
  7. Vikse BE, Bostad L, Aasarød K, Lysebo DE, Iversen BM. (2002) Prognostic factors in mesangioproliferative glomerulonephritis. Nephrol Dial Transplant 17:1603–1613.[Abstract/Free Full Text]
  8. Beukhof JR, Kardaun O, Schaafsma W, et al. (1986) Toward individual prognosis of IgA nephropathy. Kidney Int 29:549–556.[Web of Science][Medline]
  9. Ohno Y, Kawamura T, Tamakoshi A, et al. (1996) Epidemiology of intractable diseases in Japan. J Epidemiol 6 [Suppl]:S99–S109.
  10. Sakai H, Abe K, Kobayashi Y, et al. (1995) Clinical guidelines of IgA nephropathy. Jpn J Nephrol 37:417–421.
  11. Kaplan EL and Meier P. (1958) Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457–481.[CrossRef][Web of Science]
  12. Cox DR. (1972) Regression models and life-tables (with discussions). J R Stat Soc B 34:187–220.
  13. Bruemmer B, White E, Vaughan TL, Cheney CL. (1996) Nutrient intake in relation to bladder cancer among middle-aged men and women. Am J Epidemiol 144:485–495.[Abstract/Free Full Text]
  14. Khaw KT, Bingham S, Welch A, et al. (2001) Relation between plasma ascorbic acid and mortality in men and women in EPIC-Norfolk prospective study: a prospective population study. European Prospective Investigation into Cancer and Nutrition. Lancet 357:657–663.[CrossRef][Web of Science][Medline]
  15. Alamartine E, Sabatier JC, Guerin C, Berliet JM, Berthoux F. (1991) Prognostic factors in mesangial IgA glomerulonephritis: an extensive study with univariate and multivariate analyses. Am J Kidney Dis 18:12–19.[Web of Science][Medline]
  16. The PHREG procedure. SAS/STAT User's Guide, Version 8 (1999) (SAS Institute Inc., Cary, NC) pp. 2569–2658.
  17. Hanley JA and McNeil BJ. (1982) The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 143:29–36.[Abstract/Free Full Text]
  18. Blute ML, Bergstralh EJ, Iocca A, Scherer B, Zincke H. (2001) Use of gleason score, prostate specific antigen, seminal vesicle and margin status to predict biochemical failure after radical prostatectomy. J Urol 165:119–125.[CrossRef][Web of Science][Medline]
  19. Nolin L and Courteau M. (1999) Management of IgA nephropathy: evidence-based recommendations. Kidney Int 55 [Suppl 70]:S56–S62.
  20. Julian BA. (2000) Treatment of IgA nephropathy. Semin Nephrol 20:277–285.[Web of Science][Medline]
  21. Cockcroft DW and Gault MH. (1976) Prediction of creatinine clearance from serum creatinine. Nephron 16:31–41.[Web of Science][Medline]
  22. Radford MG Jr, Donadio JV Jr, Bergstralh EJ, Grande JP. (1997) Predicting renal outcome in IgA nephropathy. J Am Soc Nephrol 8:199–207.[Abstract]
  23. D'Amico G, Colasanti G, Barbiano di Belgioioso G, et al. (1987) Long-term follow-up of IgA mesangial nephropathy: clinico-histological study in 374 patients. Semin Nephrol 7:355–358.[Web of Science][Medline]
  24. Frimat L, Briançon S, Hestin D, et al. (1997) IgA nephropathy: prognostic classification of end-stage renal failure. Nephrol Dial Transplant 12:2569–2575.[Abstract/Free Full Text]
  25. Bartosik LP, Lajoie G, Sugar L, Cattran DC. (2001) Predicting progression in IgA nephropathy. Am J Kidney Dis 38:728–735.[Web of Science][Medline]
  26. Ibels LS, Gyory AZ, Caterson RJ, et al. (1997) Primary IgA nephropathy: natural history and factors of importance in the progression of renal impairment. Kidney Int 52 [Suppl 61]:S67–S70.
  27. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. (1999) A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 130:461–470.[Abstract/Free Full Text]
Received for publication: 3.11.05
Accepted in revised form: 17. 5.06


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Nephrol Dial TransplantHome page
M. Goto, K. Wakai, T. Kawamura, M. Ando, M. Endoh, and Y. Tomino
A scoring system to predict renal outcome in IgA nephropathy: a nationwide 10-year prospective cohort study
Nephrol. Dial. Transplant., June 10, 2009; (2009) gfp273v1.
[Abstract] [Full Text] [PDF]

Home page
 Nephrol Dial TransplantHome page
M. Goto, T. Kawamura, K. Wakai, M. Ando, M. Endoh, and Y. Tomino
Risk stratification for progression of IgA nephropathy using a decision tree induction algorithm
Nephrol. Dial. Transplant., April 1, 2009; 24(4): 1242 - 1247.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
21/10/2800    most recent
gfl342v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (4)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Wakai, K.
Right arrow Articles by Sakai, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wakai, K.
Right arrow Articles by Sakai, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?