Albuminuria, Proteinuria, and Graft Survival in Kidney Transplantation

Abstract

Nauta et al1Nauta F.L. Bakker S.J.L. van Oeveren W. et al.Albuminuria, proteinuria, and novel urine biomarkers as predictors of long-term allograft outcomes in kidney transplantation.Am J Kidney Dis. 2011; 57: 733-743Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar reported important new information regarding the predictive value of urine biomarkers in kidney transplantation. The investigators also claimed that albuminuria provided better predictions than proteinuria based on ROC (receiver operating characteristic) curves. However, albuminuria and proteinuria were measured at a random time after transplant (25th-75th percentile, 2.6-11.4 years). To uniformly define time relative to transplant, time origins should be related closely to the date of transplant and should be the same for all patients.2Andersen P.K. Keiding N. Survival analysis, overview.in: Armitage P. Colton T. Encyclopedia of Biostatistics. Vol 6. John Wiley, Chichester, England1998: 4452-4461Google Scholar Otherwise, sensitivity, specificity, and ROC curves should not be estimated, especially in patients with variable follow-up.3Collett D. Modelling Survival Data in Medical Research. Chapman & Hall, London, UK1994Crossref Google Scholar Nonetheless, performance of the multivariate models using albuminuria or proteinuria can be compared with each other using appropriate methods, such as the Akaike information criterion (AIC).4Tangri N. Stevens L.A. Griffith J. et al.A predictive model for progression of chronic kidney disease to kidney failure.JAMA. 2011; 305: 1553-1559Crossref PubMed Scopus (725) Google Scholar As noted by the investigators, we previously published a study showing that both 24-hour proteinuria and albuminuria measured at the same time after kidney transplant were predictive of transplant loss.5Halimi J.M. Laouad I. Buchler M. et al.Early low-grade proteinuria: causes, short-term evolution and long-term consequences in renal transplantation.Am J Transplant. 2005; 5: 2281-2288Crossref PubMed Scopus (116) Google Scholar After the publication of Nauta et al,1Nauta F.L. Bakker S.J.L. van Oeveren W. et al.Albuminuria, proteinuria, and novel urine biomarkers as predictors of long-term allograft outcomes in kidney transplantation.Am J Kidney Dis. 2011; 57: 733-743Abstract Full Text Full Text PDF PubMed Scopus (70) Google Scholar we decided to use Cox models and AIC in 674 consecutive patients who had both albuminuria and proteinuria measured at 12 months after transplant (median follow-up, 4.6 [range, 1.1-23.5] years; 32 transplant losses during follow-up). As listed in Table 1, models using albuminuria were not more accurate than models using proteinuria to predict transplant loss. More importantly for patients, short-term change in proteinuria also is an excellent predictor of transplant loss (Table 1).6Halimi J.M. Buchler M. Al-Najjar A. et al.Urinary albumin excretion and the risk of graft loss and death in proteinuric and non-proteinuric renal transplant recipients.Am J Transplant. 2007; 7: 618-625Crossref PubMed Scopus (90) Google ScholarTable 1Albuminuria, Proteinuria, and Risk of Transplant LossParameterHR (95% CI)PAICaLow values for AIC indicate better models.Univariate Albuminuria at 12 mo1.90 (1.58-2.29)<0.001354 Proteinuria at 12 mo1.56 (1.36-1.78)<0.001354 Change in proteinuriabFrom 3-12 months.1.65 (1.40-1.94)<0.001343MultivariatecAdjustment for serum creatinine level, age, body weight, and acute rejection. Albuminuria at 12 mo2.09 (1.70-2.57)<0.001344 Proteinuria at 12 mo1.67 (1.44-1.94)<0.001343 Change in proteinuriabFrom 3-12 months.1.77 (1.48-2.12)<0.001332Note: HR is per 1-g/d increase.Abbreviations: AIC, Akaike information criterion; CI, confidence interval; HR, hazard ratio.a Low values for AIC indicate better models.b From 3-12 months.c Adjustment for serum creatinine level, age, body weight, and acute rejection. Open table in a new tab Note: HR is per 1-g/d increase. Abbreviations: AIC, Akaike information criterion; CI, confidence interval; HR, hazard ratio. Financial Disclosure: The author declares that he has no relevant financial interests. Albuminuria, Proteinuria, and Novel Urine Biomarkers as Predictors of Long-term Allograft Outcomes in Kidney Transplant RecipientsAmerican Journal of Kidney DiseasesVol. 57Issue 5PreviewProteinuria is an established marker of decreased kidney function after kidney transplant. It recently has been suggested that albuminuria might be a more reliable marker. Although albuminuria often is regarded as a marker of glomerular damage, because chronic renal allograft damage is believed to be predominantly an interstitial process, albuminuria in this case might reflect tubular damage. Accordingly, we investigated the value of albuminuria, proteinuria, and tubular damage markers (KIM-1 [kidney injury molecule 1], NAG [N-acetyl-β-d-glucosaminidase], NGAL [neutrophil gelatinase-associated lipocalin], and H-FABP [heart fatty acid binding protein]) in predicting graft outcome in kidney transplant recipients. Full-Text PDF In Reply to 'Albuminuria, Proteinuria, and Graft Survival in Kidney Transplantation'American Journal of Kidney DiseasesVol. 58Issue 6PreviewWe acknowledge that, as discussed by Halimi,1 the use of ROC (receiver operating characteristic) curves can be problematic when differences in follow-up time exist among patients. Moreover, the competing risk of mortality can introduce bias. However, it should be noted that in our study, Cox analyses showed results similar to our ROC analyses.2 Of note, Cox analyses are not influenced by differences in follow-up time among patients. Furthermore, the cumulative incidence plot of transplant loss was nearly identical to the complement of the Kaplan-Meier (KM) plot (1 – KM). Full-Text PDF